diff --git a/dist/mercury.js b/dist/mercury.js
index 9f14b701..0d48f8dd 100644
--- a/dist/mercury.js
+++ b/dist/mercury.js
@@ -1498,6 +1498,8 @@ function removeAllButWhitelist($article) {
return acc;
}, {});
});
+
+ return $article;
}
// function removeAttrs(article, $) {
@@ -1508,9 +1510,11 @@ function removeAllButWhitelist($article) {
// Remove attributes like style or align
function cleanAttributes($article) {
- removeAllButWhitelist($article);
-
- return $article;
+ // Grabbing the parent because at this point
+ // $article will be wrapped in a div which will
+ // have a score set on it.
+ console.log('HMM', $article.parent().length);
+ return removeAllButWhitelist($article.parent().length ? $article.parent() : $article);
}
function removeEmpty($article, $) {
@@ -2603,9 +2607,6 @@ function extractCleanNode(article, _ref) {
// Make links absolute
makeLinksAbsolute(article, $, url);
- // Remove unnecessary attributes
- cleanAttributes(article);
-
// We used to clean UL's and OL's here, but it was leading to
// too many in-article lists being removed. Consider a better
// way to detect menus particularly and remove them.
@@ -2615,6 +2616,9 @@ function extractCleanNode(article, _ref) {
// Remove empty paragraph nodes
removeEmpty(article, $);
+ // Remove unnecessary attributes
+ cleanAttributes(article, $);
+
return article;
}
diff --git a/dist/mercury.js.map b/dist/mercury.js.map
index 86974b2b..01599d8e 100644
--- a/dist/mercury.js.map
+++ b/dist/mercury.js.map
@@ -1 +1 @@
-{"version":3,"file":null,"sources":["../src/utils/range.js","../src/utils/validate-url.js","../src/utils/errors.js","../src/resource/utils/constants.js","../src/resource/utils/fetch-resource.js","../src/resource/utils/dom/normalize-meta-tags.js","../src/resource/utils/dom/constants.js","../src/resource/utils/dom/convert-lazy-loaded-images.js","../src/resource/utils/dom/clean.js","../src/resource/index.js","../src/extractors/custom/nymag.com/index.js","../src/extractors/custom/blogspot.com/index.js","../src/extractors/custom/wikipedia.org/index.js","../src/extractors/custom/twitter.com/index.js","../src/extractors/custom/www.nytimes.com/index.js","../src/extractors/custom/www.theatlantic.com/index.js","../src/extractors/custom/www.newyorker.com/index.js","../src/extractors/custom/www.wired.com/index.js","../src/extractors/custom/www.msn.com/index.js","../src/extractors/custom/www.yahoo.com/index.js","../src/extractors/custom/www.buzzfeed.com/index.js","../src/extractors/custom/fandom.wikia.com/index.js","../src/extractors/custom/www.littlethings.com/index.js","../src/extractors/custom/www.politico.com/index.js","../src/extractors/custom/deadspin.com/index.js","../src/extractors/custom/www.broadwayworld.com/index.js","../src/extractors/custom/www.apartmenttherapy.com/index.js","../src/extractors/all.js","../src/utils/dom/constants.js","../src/utils/dom/strip-unlikely-candidates.js","../src/utils/dom/brs-to-ps.js","../src/utils/dom/paragraphize.js","../src/utils/dom/convert-to-paragraphs.js","../src/utils/dom/convert-node-to.js","../src/utils/dom/clean-images.js","../src/utils/dom/mark-to-keep.js","../src/utils/dom/strip-junk-tags.js","../src/utils/dom/clean-h-ones.js","../src/utils/dom/clean-attributes.js","../src/utils/dom/remove-empty.js","../src/extractors/generic/content/scoring/constants.js","../src/extractors/generic/content/scoring/get-weight.js","../src/extractors/generic/content/scoring/get-score.js","../src/extractors/generic/content/scoring/score-commas.js","../src/extractors/generic/content/scoring/score-length.js","../src/extractors/generic/content/scoring/score-paragraph.js","../src/extractors/generic/content/scoring/set-score.js","../src/extractors/generic/content/scoring/add-score.js","../src/extractors/generic/content/scoring/add-to-parent.js","../src/extractors/generic/content/scoring/get-or-init-score.js","../src/extractors/generic/content/scoring/score-node.js","../src/extractors/generic/content/scoring/score-content.js","../src/utils/text/normalize-spaces.js","../src/utils/text/extract-from-url.js","../src/utils/text/constants.js","../src/utils/text/page-num-from-url.js","../src/utils/text/remove-anchor.js","../src/utils/text/article-base-url.js","../src/utils/text/has-sentence-end.js","../src/utils/text/excerpt-content.js","../src/extractors/generic/content/scoring/merge-siblings.js","../src/extractors/generic/content/scoring/find-top-candidate.js","../src/utils/dom/clean-tags.js","../src/utils/dom/clean-headers.js","../src/utils/dom/rewrite-top-level.js","../src/utils/dom/make-links-absolute.js","../src/utils/dom/link-density.js","../src/utils/dom/extract-from-meta.js","../src/utils/dom/extract-from-selectors.js","../src/utils/dom/strip-tags.js","../src/utils/dom/within-comment.js","../src/utils/dom/node-is-sufficient.js","../src/utils/dom/is-wordpress.js","../src/cleaners/constants.js","../src/cleaners/author.js","../src/cleaners/lead-image-url.js","../src/cleaners/dek.js","../src/cleaners/date-published.js","../src/cleaners/content.js","../src/cleaners/title.js","../src/cleaners/resolve-split-title.js","../src/cleaners/index.js","../src/extractors/generic/content/extract-best-node.js","../src/extractors/generic/content/extractor.js","../src/extractors/generic/title/constants.js","../src/extractors/generic/title/extractor.js","../src/extractors/generic/author/constants.js","../src/extractors/generic/author/extractor.js","../src/extractors/generic/date-published/constants.js","../src/extractors/generic/date-published/extractor.js","../src/extractors/generic/dek/extractor.js","../src/extractors/generic/lead-image-url/constants.js","../src/extractors/generic/lead-image-url/score-image.js","../src/extractors/generic/lead-image-url/extractor.js","../src/extractors/generic/next-page-url/scoring/utils/score-similarity.js","../src/extractors/generic/next-page-url/scoring/utils/score-link-text.js","../src/extractors/generic/next-page-url/scoring/utils/score-page-in-link.js","../src/extractors/generic/next-page-url/scoring/constants.js","../src/extractors/generic/next-page-url/scoring/utils/score-extraneous-links.js","../src/extractors/generic/next-page-url/scoring/utils/score-by-parents.js","../src/extractors/generic/next-page-url/scoring/utils/score-prev-link.js","../src/extractors/generic/next-page-url/scoring/utils/should-score.js","../src/extractors/generic/next-page-url/scoring/utils/score-base-url.js","../src/extractors/generic/next-page-url/scoring/utils/score-next-link-text.js","../src/extractors/generic/next-page-url/scoring/utils/score-cap-links.js","../src/extractors/generic/next-page-url/scoring/score-links.js","../src/extractors/generic/next-page-url/extractor.js","../src/extractors/generic/url/constants.js","../src/extractors/generic/url/extractor.js","../src/extractors/generic/excerpt/constants.js","../src/extractors/generic/excerpt/extractor.js","../src/extractors/generic/word-count/extractor.js","../src/extractors/generic/index.js","../src/extractors/get-extractor.js","../src/extractors/root-extractor.js","../src/extractors/collect-all-pages.js","../src/mercury.js"],"sourcesContent":["export default function* range(start = 1, end = 1) {\n while (start <= end) {\n yield start += 1;\n }\n}\n","// extremely simple url validation as a first step\nexport default function validateUrl({ hostname }) {\n // If this isn't a valid url, return an error message\n return !!hostname;\n}\n","const Errors = {\n badUrl: {\n error: true,\n messages: 'The url parameter passed does not look like a valid URL. Please check your data and try again.',\n },\n};\n\nexport default Errors;\n","export const REQUEST_HEADERS = {\n 'User-Agent': 'Readability - http://readability.com/about/',\n};\n\n// The number of milliseconds to attempt to fetch a resource before timing out.\nexport const FETCH_TIMEOUT = 10000;\n\n// Content types that we do not extract content from\nconst BAD_CONTENT_TYPES = [\n 'audio/mpeg',\n 'image/gif',\n 'image/jpeg',\n 'image/jpg',\n];\n\nexport const BAD_CONTENT_TYPES_RE = new RegExp(`^(${BAD_CONTENT_TYPES.join('|')})$`, 'i');\n\n\n// Use this setting as the maximum size an article can be\n// for us to attempt parsing. Defaults to 5 MB.\nexport const MAX_CONTENT_LENGTH = 5242880;\n\n// Turn the global proxy on or off\n// Proxying is not currently enabled in Python source\n// so not implementing logic in port.\nexport const PROXY_DOMAINS = false;\nexport const REQUESTS_PROXIES = {\n http: 'http://38.98.105.139:33333',\n https: 'http://38.98.105.139:33333',\n};\n\nexport const DOMAINS_TO_PROXY = [\n 'nih.gov',\n 'gutenberg.org',\n];\n","import URL from 'url';\nimport request from 'request';\nimport { Errors } from 'utils';\n\nimport {\n REQUEST_HEADERS,\n FETCH_TIMEOUT,\n BAD_CONTENT_TYPES_RE,\n MAX_CONTENT_LENGTH,\n} from './constants';\n\nfunction get(options) {\n return new Promise((resolve, reject) => {\n request(options, (err, response, body) => {\n if (err) {\n reject(err);\n } else {\n resolve({ body, response });\n }\n });\n });\n}\n\n// Evaluate a response to ensure it's something we should be keeping.\n// This does not validate in the sense of a response being 200 level or\n// not. Validation here means that we haven't found reason to bail from\n// further processing of this url.\n\nexport function validateResponse(response, parseNon2xx = false) {\n // Check if we got a valid status code\n if (response.statusMessage !== 'OK') {\n if (!response.statusCode) {\n throw new Error(\n `Unable to fetch content. Original exception was ${response.error}`\n );\n } else if (!parseNon2xx) {\n throw new Error(\n `Resource returned a response status code of ${response.statusCode} and resource was instructed to reject non-2xx level status codes.`\n );\n }\n }\n\n const {\n 'content-type': contentType,\n 'content-length': contentLength,\n } = response.headers;\n\n // Check that the content is not in BAD_CONTENT_TYPES\n if (BAD_CONTENT_TYPES_RE.test(contentType)) {\n throw new Error(\n `Content-type for this resource was ${contentType} and is not allowed.`\n );\n }\n\n // Check that the content length is below maximum\n if (contentLength > MAX_CONTENT_LENGTH) {\n throw new Error(\n `Content for this resource was too large. Maximum content length is ${MAX_CONTENT_LENGTH}.`\n );\n }\n\n return true;\n}\n\n// Grabs the last two pieces of the URL and joins them back together\n// This is to get the 'livejournal.com' from 'erotictrains.livejournal.com'\nexport function baseDomain({ host }) {\n return host.split('.').slice(-2).join('.');\n}\n\n// Set our response attribute to the result of fetching our URL.\n// TODO: This should gracefully handle timeouts and raise the\n// proper exceptions on the many failure cases of HTTP.\n// TODO: Ensure we are not fetching something enormous. Always return\n// unicode content for HTML, with charset conversion.\n\nexport default async function fetchResource(url, parsedUrl) {\n parsedUrl = parsedUrl || URL.parse(encodeURI(url));\n\n const options = {\n url: parsedUrl,\n headers: { ...REQUEST_HEADERS },\n timeout: FETCH_TIMEOUT,\n // Don't set encoding; fixes issues\n // w/gzipped responses\n encoding: null,\n // Accept cookies\n jar: true,\n // Accept and decode gzip\n gzip: true,\n // Follow any redirect\n followAllRedirects: true,\n };\n\n const { response, body } = await get(options);\n\n try {\n validateResponse(response);\n return { body, response };\n } catch (e) {\n return Errors.badUrl;\n }\n}\n","function convertMetaProp($, from, to) {\n $(`meta[${from}]`).each((_, node) => {\n const $node = $(node);\n\n const value = $node.attr(from);\n $node.attr(to, value);\n $node.removeAttr(from);\n });\n\n return $;\n}\n\n// For ease of use in extracting from meta tags,\n// replace the \"content\" attribute on meta tags with the\n// \"value\" attribute.\n//\n// In addition, normalize 'property' attributes to 'name' for ease of\n// querying later. See, e.g., og or twitter meta tags.\n\nexport default function normalizeMetaTags($) {\n $ = convertMetaProp($, 'content', 'value');\n $ = convertMetaProp($, 'property', 'name');\n return $;\n}\n","export const IS_LINK = new RegExp('https?://', 'i');\nexport const IS_IMAGE = new RegExp('.(png|gif|jpe?g)', 'i');\n\nexport const TAGS_TO_REMOVE = [\n 'script',\n 'style',\n 'form',\n].join(',');\n","import {\n IS_LINK,\n IS_IMAGE,\n} from './constants';\n\n// Convert all instances of images with potentially\n// lazy loaded images into normal images.\n// Many sites will have img tags with no source, or an image tag with a src\n// attribute that a is a placeholer. We need to be able to properly fill in\n// the src attribute so the images are no longer lazy loaded.\nexport default function convertLazyLoadedImages($) {\n $('img').each((_, img) => {\n Reflect.ownKeys(img.attribs).forEach((attr) => {\n const value = img.attribs[attr];\n\n if (attr !== 'src' && IS_LINK.test(value) &&\n IS_IMAGE.test(value)) {\n $(img).attr('src', value);\n }\n });\n });\n\n return $;\n}\n","import { TAGS_TO_REMOVE } from './constants';\n\nfunction isComment(index, node) {\n return node.type === 'comment';\n}\n\nfunction cleanComments($) {\n $.root().find('*')\n .contents()\n .filter(isComment)\n .remove();\n\n return $;\n}\n\nexport default function clean($) {\n $(TAGS_TO_REMOVE).remove();\n\n $ = cleanComments($);\n return $;\n}\n","import cheerio from 'cheerio';\n\nimport { fetchResource } from './utils';\nimport {\n normalizeMetaTags,\n convertLazyLoadedImages,\n clean,\n} from './utils/dom';\n\nconst Resource = {\n\n // Create a Resource.\n //\n // :param url: The URL for the document we should retrieve.\n // :param response: If set, use as the response rather than\n // attempting to fetch it ourselves. Expects a\n // string.\n async create(url, preparedResponse, parsedUrl) {\n let result;\n\n if (preparedResponse) {\n const validResponse = {\n statusMessage: 'OK',\n statusCode: 200,\n headers: {\n 'content-type': 'text/html',\n 'content-length': 500,\n },\n };\n\n result = { body: preparedResponse, response: validResponse };\n } else {\n result = await fetchResource(url, parsedUrl);\n }\n\n if (result.error) {\n return result;\n }\n\n return this.generateDoc(result);\n },\n\n generateDoc({ body: content, response }) {\n const { 'content-type': contentType } = response.headers;\n\n // TODO: Implement is_text function from\n // https://github.com/ReadabilityHoldings/readability/blob/8dc89613241d04741ebd42fa9fa7df1b1d746303/readability/utils/text.py#L57\n if (!contentType.includes('html') &&\n !contentType.includes('text')) {\n throw new Error('Content does not appear to be text.');\n }\n\n let $ = cheerio.load(content, { normalizeWhitespace: true });\n\n if ($.root().children().length === 0) {\n throw new Error('No children, likely a bad parse.');\n }\n\n $ = normalizeMetaTags($);\n $ = convertLazyLoadedImages($);\n $ = clean($);\n\n return $;\n },\n};\n\nexport default Resource;\n","export const NYMagExtractor = {\n domain: 'nymag.com',\n content: {\n // Order by most likely. Extractor will stop on first occurrence\n selectors: [\n 'div.article-content',\n 'section.body',\n 'article.article',\n ],\n\n // Selectors to remove from the extracted content\n clean: [\n '.ad',\n '.single-related-story',\n ],\n\n // Object of tranformations to make on matched elements\n // Each key is the selector, each value is the tag to\n // transform to.\n // If a function is given, it should return a string\n // to convert to or nothing (in which case it will not perform\n // the transformation.\n transforms: {\n // Convert h1s to h2s\n h1: 'h2',\n\n // Convert lazy-loaded noscript images to figures\n noscript: ($node) => {\n const $children = $node.children();\n if ($children.length === 1 && $children.get(0).tagName === 'img') {\n return 'figure';\n }\n\n return null;\n },\n },\n },\n\n title: {\n selectors: [\n 'h1.lede-feature-title',\n 'h1.headline-primary',\n 'h1',\n ],\n },\n\n author: {\n selectors: [\n '.by-authors',\n '.lede-feature-author',\n ],\n },\n\n dek: {\n selectors: [\n '.lede-feature-teaser',\n ],\n },\n\n date_published: {\n selectors: [\n ['time.article-timestamp[datetime]', 'datetime'],\n 'time.article-timestamp',\n ],\n },\n};\n","export const BloggerExtractor = {\n domain: 'blogspot.com',\n content: {\n // Blogger is insane and does not load its content\n // initially in the page, but it's all there\n // in noscript\n selectors: [\n '.post-content noscript',\n ],\n\n // Selectors to remove from the extracted content\n clean: [\n ],\n\n // Convert the noscript tag to a div\n transforms: {\n noscript: 'div',\n },\n },\n\n author: {\n selectors: [\n '.post-author-name',\n ],\n },\n\n title: {\n selectors: [\n '.post h2.title',\n ],\n },\n\n date_published: {\n selectors: [\n 'span.publishdate',\n ],\n },\n};\n","export const WikipediaExtractor = {\n domain: 'wikipedia.org',\n content: {\n selectors: [\n '#mw-content-text',\n ],\n\n defaultCleaner: false,\n\n // transform top infobox to an image with caption\n transforms: {\n '.infobox img': ($node) => {\n const $parent = $node.parents('.infobox');\n // Only prepend the first image in .infobox\n if ($parent.children('img').length === 0) {\n $parent.prepend($node);\n }\n },\n '.infobox caption': 'figcaption',\n '.infobox': 'figure',\n },\n\n // Selectors to remove from the extracted content\n clean: [\n '.mw-editsection',\n 'figure tr, figure td, figure tbody',\n '#toc',\n '.navbox',\n ],\n\n },\n\n author: 'Wikipedia Contributors',\n\n title: {\n selectors: [\n 'h2.title',\n ],\n },\n\n date_published: {\n selectors: [\n '#footer-info-lastmod',\n ],\n },\n\n};\n","export const TwitterExtractor = {\n domain: 'twitter.com',\n\n content: {\n transforms: {\n // We're transforming essentially the whole page here.\n // Twitter doesn't have nice selectors, so our initial\n // selector grabs the whole page, then we're re-writing\n // it to fit our needs before we clean it up.\n '.permalink[role=main]': ($node, $) => {\n const tweets = $node.find('.tweet');\n const $tweetContainer = $('');\n $tweetContainer.append(tweets);\n $node.replaceWith($tweetContainer);\n },\n\n // Twitter wraps @ with s, which\n // renders as a strikethrough\n s: 'span',\n },\n\n selectors: [\n '.permalink[role=main]',\n ],\n\n defaultCleaner: false,\n\n clean: [\n '.stream-item-footer',\n 'button',\n '.tweet-details-fixer',\n ],\n },\n\n author: {\n selectors: [\n '.tweet.permalink-tweet .username',\n ],\n },\n\n date_published: {\n selectors: [\n ['.permalink-tweet ._timestamp[data-time-ms]', 'data-time-ms'],\n // '.tweet.permalink-tweet .metadata',\n ],\n },\n\n};\n","export const NYTimesExtractor = {\n title: {\n selectors: [\n '.g-headline',\n 'h1.headline',\n ],\n },\n\n author: {\n selectors: [\n ['meta[name=\"author\"]', 'value'],\n '.g-byline',\n '.byline',\n ],\n },\n\n content: {\n selectors: [\n 'div.g-blocks',\n 'article#story',\n ],\n\n defaultCleaner: false,\n\n transforms: {\n 'img.g-lazy': ($node) => {\n let src = $node.attr('src');\n // const widths = $node.attr('data-widths')\n // .slice(1)\n // .slice(0, -1)\n // .split(',');\n // if (widths.length) {\n // width = widths.slice(-1);\n // } else {\n // width = '900';\n // }\n const width = 640;\n\n src = src.replace('{{size}}', width);\n $node.attr('src', src);\n },\n },\n\n clean: [\n '.ad',\n 'header#story-header',\n '.story-body-1 .lede.video',\n '.visually-hidden',\n '#newsletter-promo',\n '.promo',\n '.comments-button',\n '.hidden',\n ],\n },\n\n date_published: null,\n\n lead_image_url: null,\n\n dek: null,\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\nexport const TheAtlanticExtractor = {\n domain: 'www.theatlantic.com',\n title: {\n selectors: [\n 'h1.hed',\n ],\n },\n\n author: {\n selectors: [\n 'article#article .article-cover-extra .metadata .byline a',\n ],\n },\n\n content: {\n selectors: [\n '.article-body',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: null,\n\n lead_image_url: null,\n\n dek: null,\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const NewYorkerExtractor = {\n domain: 'www.newyorker.com',\n title: {\n selectors: [\n 'h1.title',\n ],\n },\n\n author: {\n selectors: [\n '.contributors',\n ],\n },\n\n content: {\n selectors: [\n 'div#articleBody',\n 'div.articleBody',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: {\n selectors: [\n ['meta[name=\"article:published_time\"]', 'value'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const WiredExtractor = {\n domain: 'www.wired.com',\n title: {\n selectors: [\n 'h1.post-title',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n 'a[rel=\"author\"]',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n 'article.content',\n // enter content selectors\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n '.visually-hidden',\n\n ],\n },\n\n date_published: {\n selectors: [\n ['meta[itemprop=\"datePublished\"]', 'value'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const MSNExtractor = {\n domain: 'www.msn.com',\n title: {\n selectors: [\n 'h1',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n 'span.authorname-txt',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n 'div.richtext',\n // enter content selectors\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n 'span.caption',\n\n ],\n },\n\n date_published: {\n selectors: [\n 'span.time',\n ],\n },\n\n lead_image_url: {\n selectors: [\n\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const YahooExtractor = {\n domain: 'www.yahoo.com',\n title: {\n selectors: [\n 'header.canvas-header',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n 'span.provider-name',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n // enter content selectors\n '.content-canvas',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n '.figure-caption',\n\n ],\n },\n\n date_published: {\n selectors: [\n ['time.date[datetime]', 'datetime'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n // enter dek selectors\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const BuzzfeedExtractor = {\n domain: 'www.buzzfeed.com',\n title: {\n selectors: [\n 'h1[id=\"post-title\"]',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n 'a[data-action=\"user/username\"]', 'byline__author',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n '#buzz_sub_buzz',\n // enter content selectors\n ],\n\n defaultCleaner: false,\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: {\n h2: 'b',\n },\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n '.instapaper_ignore',\n '.suplist_list_hide .buzz_superlist_item .buzz_superlist_number_inline',\n '.share-box',\n ],\n },\n\n date_published: {\n selectors: [\n '.buzz-datetime',\n // enter author selectors\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const WikiaExtractor = {\n domain: 'fandom.wikia.com',\n title: {\n selectors: [\n 'h1.entry-title',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n '.author vcard', '.fn',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n '.grid-content',\n '.entry-content',\n // enter content selectors\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: {\n selectors: [\n ['meta[name=\"article:published_time\"]', 'value'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const LittleThingsExtractor = {\n domain: 'www.littlethings.com',\n title: {\n selectors: [\n 'h1.post-title',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n ['meta[name=\"author\"]', 'value'],\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n // enter content selectors\n '.mainContentIntro',\n '.content-wrapper',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const PoliticoExtractor = {\n domain: 'www.politico.com',\n title: {\n selectors: [\n // enter title selectors\n ['meta[name=\"og:title\"]', 'value'],\n ],\n },\n\n author: {\n selectors: [\n '.story-main-content .byline .vcard',\n ],\n },\n\n content: {\n selectors: [\n // enter content selectors\n '.story-main-content',\n '.content-group', '.story-core',\n '.story-text',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n 'figcaption',\n ],\n },\n\n date_published: {\n selectors: [\n ['.story-main-content .timestamp time[datetime]', 'datetime'],\n\n ],\n },\n\n lead_image_url: {\n selectors: [\n // enter lead_image_url selectors\n ['meta[name=\"og:image\"]', 'value'],\n\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","export const DeadspinExtractor = {\n domain: 'deadspin.com',\n title: {\n selectors: [\n 'h1.headline',\n ],\n },\n\n author: {\n selectors: [\n '.author',\n ],\n },\n\n content: {\n selectors: [\n '.post-content',\n '.entry-content',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: {\n 'iframe.lazyload[data-recommend-id^=\"youtube://\"]': ($node) => {\n const youtubeId = $node.attr('id').split('youtube-')[1];\n $node.attr('src', `https://www.youtube.com/embed/${youtubeId}`);\n },\n },\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n ],\n },\n\n date_published: {\n selectors: [\n ['time.updated[datetime]', 'datetime'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n // enter selectors\n ],\n },\n\n next_page_url: {\n selectors: [\n // enter selectors\n ],\n },\n\n excerpt: {\n selectors: [\n // enter selectors\n ],\n },\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const BroadwayWorldExtractor = {\n domain: 'www.broadwayworld.com',\n title: {\n selectors: [\n 'h1.article-title',\n ],\n },\n\n author: {\n selectors: [\n 'span[itemprop=author]',\n ],\n },\n\n content: {\n selectors: [\n 'div[itemprop=articlebody]',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: {\n },\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: {\n selectors: [\n ['meta[itemprop=datePublished]', 'value'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n ],\n },\n\n next_page_url: {\n selectors: [\n // enter selectors\n ],\n },\n\n excerpt: {\n selectors: [\n // enter selectors\n ],\n },\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const ApartmentTherapyExtractor = {\n domain: 'www.apartmenttherapy.com',\n title: {\n selectors: [\n 'h1.headline',\n ],\n },\n\n author: {\n selectors: [\n '.PostByline__name',\n ],\n },\n\n content: {\n selectors: [\n 'div.post__content',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: {\n 'div[data-render-react-id=\"images/LazyPicture\"]': ($node, $) => {\n const data = JSON.parse($node.attr('data-props'));\n const { src } = data.sources[0];\n const $img = $('![]()
').attr('src', src);\n $node.replaceWith($img);\n },\n },\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: {\n selectors: [\n ['.PostByline__timestamp[datetime]', 'datetime'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=description]', 'value'],\n ],\n },\n\n next_page_url: {\n selectors: [\n // enter selectors\n ],\n },\n\n excerpt: {\n selectors: [\n // enter selectors\n ],\n },\n};\n","import { NYMagExtractor } from './custom/nymag.com';\nimport { BloggerExtractor } from './custom/blogspot.com';\nimport { WikipediaExtractor } from './custom/wikipedia.org';\nimport { TwitterExtractor } from './custom/twitter.com';\nimport { NYTimesExtractor } from './custom/www.nytimes.com';\nimport { TheAtlanticExtractor } from './custom/www.theatlantic.com';\nimport { NewYorkerExtractor } from './custom/www.newyorker.com';\nimport { WiredExtractor } from './custom/www.wired.com';\nimport { MSNExtractor } from './custom/www.msn.com';\nimport { YahooExtractor } from './custom/www.yahoo.com';\nimport { BuzzfeedExtractor } from './custom/www.buzzfeed.com';\nimport { WikiaExtractor } from './custom/fandom.wikia.com';\nimport { LittleThingsExtractor } from './custom/www.littlethings.com';\nimport { PoliticoExtractor } from './custom/www.politico.com';\nimport { DeadspinExtractor } from './custom/deadspin.com';\nimport { BroadwayWorldExtractor } from './custom/www.broadwayworld.com';\nimport { ApartmentTherapyExtractor } from './custom/www.apartmenttherapy.com';\n\nconst Extractors = {\n 'nymag.com': NYMagExtractor,\n 'blogspot.com': BloggerExtractor,\n 'wikipedia.org': WikipediaExtractor,\n 'twitter.com': TwitterExtractor,\n 'www.nytimes.com': NYTimesExtractor,\n 'www.theatlantic.com': TheAtlanticExtractor,\n 'www.newyorker.com': NewYorkerExtractor,\n 'www.wired.com': WiredExtractor,\n 'www.msn.com': MSNExtractor,\n 'www.yahoo.com': YahooExtractor,\n 'www.buzzfeed.com': BuzzfeedExtractor,\n 'fandom.wikia.com': WikiaExtractor,\n 'www.littlethings.com': LittleThingsExtractor,\n 'www.politico.com': PoliticoExtractor,\n 'deadspin.com': DeadspinExtractor,\n 'www.broadwayworld.com': BroadwayWorldExtractor,\n 'www.apartmenttherapy.com': ApartmentTherapyExtractor,\n};\n\nexport default Extractors;\n","// Spacer images to be removed\nexport const SPACER_RE = new RegExp('trans|transparent|spacer|blank', 'i');\n\n// The class we will use to mark elements we want to keep\n// but would normally remove\nexport const KEEP_CLASS = 'mercury-parser-keep';\n\nexport const KEEP_SELECTORS = [\n 'iframe[src^=\"https://www.youtube.com\"]',\n 'iframe[src^=\"http://www.youtube.com\"]',\n 'iframe[src^=\"https://player.vimeo\"]',\n 'iframe[src^=\"http://player.vimeo\"]',\n];\n\n// A list of tags to strip from the output if we encounter them.\nexport const STRIP_OUTPUT_TAGS = [\n 'title',\n 'script',\n 'noscript',\n 'link',\n 'style',\n 'hr',\n 'embed',\n 'iframe',\n 'object',\n];\n\n// cleanAttributes\nexport const REMOVE_ATTRS = ['style', 'align'];\nexport const REMOVE_ATTR_SELECTORS = REMOVE_ATTRS.map(selector => `[${selector}]`);\nexport const REMOVE_ATTR_LIST = REMOVE_ATTRS.join(',');\nexport const WHITELIST_ATTRS = ['src', 'srcset', 'href', 'class', 'id', 'alt'];\nexport const WHITELIST_ATTRS_RE = new RegExp(`^(${WHITELIST_ATTRS.join('|')})$`, 'i');\n\n// removeEmpty\nexport const REMOVE_EMPTY_TAGS = ['p'];\nexport const REMOVE_EMPTY_SELECTORS = REMOVE_EMPTY_TAGS.map(tag => `${tag}:empty`).join(',');\n\n// cleanTags\nexport const CLEAN_CONDITIONALLY_TAGS = ['ul', 'ol', 'table', 'div', 'button', 'form'].join(',');\n\n// cleanHeaders\nconst HEADER_TAGS = ['h2', 'h3', 'h4', 'h5', 'h6'];\nexport const HEADER_TAG_LIST = HEADER_TAGS.join(',');\n\n\n// // CONTENT FETCHING CONSTANTS ////\n\n// A list of strings that can be considered unlikely candidates when\n// extracting content from a resource. These strings are joined together\n// and then tested for existence using re:test, so may contain simple,\n// non-pipe style regular expression queries if necessary.\nexport const UNLIKELY_CANDIDATES_BLACKLIST = [\n 'ad-break',\n 'adbox',\n 'advert',\n 'addthis',\n 'agegate',\n 'aux',\n 'blogger-labels',\n 'combx',\n 'comment',\n 'conversation',\n 'disqus',\n 'entry-unrelated',\n 'extra',\n 'foot',\n // 'form', // This is too generic, has too many false positives\n 'header',\n 'hidden',\n 'loader',\n 'login', // Note: This can hit 'blogindex'.\n 'menu',\n 'meta',\n 'nav',\n 'outbrain',\n 'pager',\n 'pagination',\n 'predicta', // readwriteweb inline ad box\n 'presence_control_external', // lifehacker.com container full of false positives\n 'popup',\n 'printfriendly',\n 'related',\n 'remove',\n 'remark',\n 'rss',\n 'share',\n 'shoutbox',\n 'sidebar',\n 'sociable',\n 'sponsor',\n 'taboola',\n 'tools',\n];\n\n// A list of strings that can be considered LIKELY candidates when\n// extracting content from a resource. Essentially, the inverse of the\n// blacklist above - if something matches both blacklist and whitelist,\n// it is kept. This is useful, for example, if something has a className\n// of \"rss-content entry-content\". It matched 'rss', so it would normally\n// be removed, however, it's also the entry content, so it should be left\n// alone.\n//\n// These strings are joined together and then tested for existence using\n// re:test, so may contain simple, non-pipe style regular expression queries\n// if necessary.\nexport const UNLIKELY_CANDIDATES_WHITELIST = [\n 'and',\n 'article',\n 'body',\n 'blogindex',\n 'column',\n 'content',\n 'entry-content-asset',\n 'format', // misuse of form\n 'hfeed',\n 'hentry',\n 'hatom',\n 'main',\n 'page',\n 'posts',\n 'shadow',\n];\n\n// A list of tags which, if found inside, should cause a to NOT\n// be turned into a paragraph tag. Shallow div tags without these elements\n// should be turned into tags.\nexport const DIV_TO_P_BLOCK_TAGS = [\n 'a',\n 'blockquote',\n 'dl',\n 'div',\n 'img',\n 'p',\n 'pre',\n 'table',\n].join(',');\n\n// A list of tags that should be ignored when trying to find the top candidate\n// for a document.\nexport const NON_TOP_CANDIDATE_TAGS = [\n 'br',\n 'b',\n 'i',\n 'label',\n 'hr',\n 'area',\n 'base',\n 'basefont',\n 'input',\n 'img',\n 'link',\n 'meta',\n];\n\nexport const NON_TOP_CANDIDATE_TAGS_RE =\n new RegExp(`^(${NON_TOP_CANDIDATE_TAGS.join('|')})$`, 'i');\n\n// A list of selectors that specify, very clearly, either hNews or other\n// very content-specific style content, like Blogger templates.\n// More examples here: http://microformats.org/wiki/blog-post-formats\nexport const HNEWS_CONTENT_SELECTORS = [\n ['.hentry', '.entry-content'],\n ['entry', '.entry-content'],\n ['.entry', '.entry_content'],\n ['.post', '.postbody'],\n ['.post', '.post_body'],\n ['.post', '.post-body'],\n];\n\nexport const PHOTO_HINTS = [\n 'figure',\n 'photo',\n 'image',\n 'caption',\n];\nexport const PHOTO_HINTS_RE = new RegExp(PHOTO_HINTS.join('|'), 'i');\n\n\n// A list of strings that denote a positive scoring for this content as being\n// an article container. Checked against className and id.\n//\n// TODO: Perhaps have these scale based on their odds of being quality?\nexport const POSITIVE_SCORE_HINTS = [\n 'article',\n 'articlecontent',\n 'instapaper_body',\n 'blog',\n 'body',\n 'content',\n 'entry-content-asset',\n 'entry',\n 'hentry',\n 'main',\n 'Normal',\n 'page',\n 'pagination',\n 'permalink',\n 'post',\n 'story',\n 'text',\n '[-_]copy', // usatoday\n '\\\\Bcopy',\n];\n\n// The above list, joined into a matching regular expression\nexport const POSITIVE_SCORE_RE = new RegExp(POSITIVE_SCORE_HINTS.join('|'), 'i');\n\n// Readability publisher-specific guidelines\nexport const READABILITY_ASSET = new RegExp('entry-content-asset', 'i');\n\n// A list of strings that denote a negative scoring for this content as being\n// an article container. Checked against className and id.\n//\n// TODO: Perhaps have these scale based on their odds of being quality?\nexport const NEGATIVE_SCORE_HINTS = [\n 'adbox',\n 'advert',\n 'author',\n 'bio',\n 'bookmark',\n 'bottom',\n 'byline',\n 'clear',\n 'com-',\n 'combx',\n 'comment',\n 'comment\\\\B',\n 'contact',\n 'copy',\n 'credit',\n 'crumb',\n 'date',\n 'deck',\n 'excerpt',\n 'featured', // tnr.com has a featured_content which throws us off\n 'foot',\n 'footer',\n 'footnote',\n 'graf',\n 'head',\n 'info',\n 'infotext', // newscientist.com copyright\n 'instapaper_ignore',\n 'jump',\n 'linebreak',\n 'link',\n 'masthead',\n 'media',\n 'meta',\n 'modal',\n 'outbrain', // slate.com junk\n 'promo',\n 'pr_', // autoblog - press release\n 'related',\n 'respond',\n 'roundcontent', // lifehacker restricted content warning\n 'scroll',\n 'secondary',\n 'share',\n 'shopping',\n 'shoutbox',\n 'side',\n 'sidebar',\n 'sponsor',\n 'stamp',\n 'sub',\n 'summary',\n 'tags',\n 'tools',\n 'widget',\n];\n// The above list, joined into a matching regular expression\nexport const NEGATIVE_SCORE_RE = new RegExp(NEGATIVE_SCORE_HINTS.join('|'), 'i');\n\n// XPath to try to determine if a page is wordpress. Not always successful.\nexport const IS_WP_SELECTOR = 'meta[name=generator][value^=WordPress]';\n\n// Match a digit. Pretty clear.\nexport const DIGIT_RE = new RegExp('[0-9]');\n\n// A list of words that, if found in link text or URLs, likely mean that\n// this link is not a next page link.\nexport const EXTRANEOUS_LINK_HINTS = [\n 'print',\n 'archive',\n 'comment',\n 'discuss',\n 'e-mail',\n 'email',\n 'share',\n 'reply',\n 'all',\n 'login',\n 'sign',\n 'single',\n 'adx',\n 'entry-unrelated',\n];\nexport const EXTRANEOUS_LINK_HINTS_RE = new RegExp(EXTRANEOUS_LINK_HINTS.join('|'), 'i');\n\n// Match any phrase that looks like it could be page, or paging, or pagination\nexport const PAGE_RE = new RegExp('pag(e|ing|inat)', 'i');\n\n// Match any link text/classname/id that looks like it could mean the next\n// page. Things like: next, continue, >, >>, » but not >|, »| as those can\n// mean last page.\n// export const NEXT_LINK_TEXT_RE = new RegExp('(next|weiter|continue|>([^\\|]|$)|»([^\\|]|$))', 'i');\nexport const NEXT_LINK_TEXT_RE = /(next|weiter|continue|>([^\\|]|$)|»([^\\|]|$))/i;\n\n// Match any link text/classname/id that looks like it is an end link: things\n// like \"first\", \"last\", \"end\", etc.\nexport const CAP_LINK_TEXT_RE = new RegExp('(first|last|end)', 'i');\n\n// Match any link text/classname/id that looks like it means the previous\n// page.\nexport const PREV_LINK_TEXT_RE = new RegExp('(prev|earl|old|new|<|«)', 'i');\n\n// Match 2 or more consecutive
tags\nexport const BR_TAGS_RE = new RegExp('(]*>[ \\n\\r\\t]*){2,}', 'i');\n\n// Match 1 BR tag.\nexport const BR_TAG_RE = new RegExp(']*>', 'i');\n\n// A list of all of the block level tags known in HTML5 and below. Taken from\n// http://bit.ly/qneNIT\nexport const BLOCK_LEVEL_TAGS = [\n 'article',\n 'aside',\n 'blockquote',\n 'body',\n 'br',\n 'button',\n 'canvas',\n 'caption',\n 'col',\n 'colgroup',\n 'dd',\n 'div',\n 'dl',\n 'dt',\n 'embed',\n 'fieldset',\n 'figcaption',\n 'figure',\n 'footer',\n 'form',\n 'h1',\n 'h2',\n 'h3',\n 'h4',\n 'h5',\n 'h6',\n 'header',\n 'hgroup',\n 'hr',\n 'li',\n 'map',\n 'object',\n 'ol',\n 'output',\n 'p',\n 'pre',\n 'progress',\n 'section',\n 'table',\n 'tbody',\n 'textarea',\n 'tfoot',\n 'th',\n 'thead',\n 'tr',\n 'ul',\n 'video',\n];\nexport const BLOCK_LEVEL_TAGS_RE = new RegExp(`^(${BLOCK_LEVEL_TAGS.join('|')})$`, 'i');\n\n\n// The removal is implemented as a blacklist and whitelist, this test finds\n// blacklisted elements that aren't whitelisted. We do this all in one\n// expression-both because it's only one pass, and because this skips the\n// serialization for whitelisted nodes.\nconst candidatesBlacklist = UNLIKELY_CANDIDATES_BLACKLIST.join('|');\nexport const CANDIDATES_BLACKLIST = new RegExp(candidatesBlacklist, 'i');\n\nconst candidatesWhitelist = UNLIKELY_CANDIDATES_WHITELIST.join('|');\nexport const CANDIDATES_WHITELIST = new RegExp(candidatesWhitelist, 'i');\n\nexport const UNLIKELY_RE = new RegExp(`!(${candidatesWhitelist})|(${candidatesBlacklist})`, 'i');\n\n\nexport const PARAGRAPH_SCORE_TAGS = new RegExp('^(p|li|span|pre)$', 'i');\nexport const CHILD_CONTENT_TAGS = new RegExp('^(td|blockquote|ol|ul|dl)$', 'i');\nexport const BAD_TAGS = new RegExp('^(address|form)$', 'i');\n\nexport const HTML_OR_BODY_RE = new RegExp('^(html|body)$', 'i');\n","import {\n CANDIDATES_WHITELIST,\n CANDIDATES_BLACKLIST,\n} from './constants';\n\nexport default function stripUnlikelyCandidates($) {\n // Loop through the provided document and remove any non-link nodes\n // that are unlikely candidates for article content.\n //\n // Links are ignored because there are very often links to content\n // that are identified as non-body-content, but may be inside\n // article-like content.\n //\n // :param $: a cheerio object to strip nodes from\n // :return $: the cleaned cheerio object\n $('*').not('a').each((index, node) => {\n const $node = $(node);\n const classes = $node.attr('class');\n const id = $node.attr('id');\n if (!id && !classes) return;\n\n const classAndId = `${classes || ''} ${id || ''}`;\n if (CANDIDATES_WHITELIST.test(classAndId)) {\n return;\n } else if (CANDIDATES_BLACKLIST.test(classAndId)) {\n $node.remove();\n }\n });\n\n return $;\n}\n","import { paragraphize } from './index';\n\n// ## NOTES:\n// Another good candidate for refactoring/optimizing.\n// Very imperative code, I don't love it. - AP\n\n\n// Given cheerio object, convert consecutive
tags into\n// tags instead.\n//\n// :param $: A cheerio object\n\nexport default function brsToPs($) {\n let collapsing = false;\n $('br').each((index, element) => {\n const nextElement = $(element).next().get(0);\n\n if (nextElement && nextElement.tagName === 'br') {\n collapsing = true;\n $(element).remove();\n } else if (collapsing) {\n collapsing = false;\n // $(element).replaceWith('')\n paragraphize(element, $, true);\n }\n });\n\n return $;\n}\n","import { BLOCK_LEVEL_TAGS_RE } from './constants';\n\n// Given a node, turn it into a P if it is not already a P, and\n// make sure it conforms to the constraints of a P tag (I.E. does\n// not contain any other block tags.)\n//\n// If the node is a
, it treats the following inline siblings\n// as if they were its children.\n//\n// :param node: The node to paragraphize; this is a raw node\n// :param $: The cheerio object to handle dom manipulation\n// :param br: Whether or not the passed node is a br\n\nexport default function paragraphize(node, $, br = false) {\n const $node = $(node);\n\n if (br) {\n let sibling = node.nextSibling;\n const p = $('');\n\n // while the next node is text or not a block level element\n // append it to a new p node\n while (sibling && !(sibling.tagName && BLOCK_LEVEL_TAGS_RE.test(sibling.tagName))) {\n const nextSibling = sibling.nextSibling;\n $(sibling).appendTo(p);\n sibling = nextSibling;\n }\n\n $node.replaceWith(p);\n $node.remove();\n return $;\n }\n\n return $;\n}\n","import { brsToPs, convertNodeTo } from 'utils/dom';\n\nimport { DIV_TO_P_BLOCK_TAGS } from './constants';\n\nfunction convertDivs($) {\n $('div').each((index, div) => {\n const $div = $(div);\n const convertable = $div.children(DIV_TO_P_BLOCK_TAGS).length === 0;\n\n if (convertable) {\n convertNodeTo($div, $, 'p');\n }\n });\n\n return $;\n}\n\nfunction convertSpans($) {\n $('span').each((index, span) => {\n const $span = $(span);\n const convertable = $span.parents('p, div').length === 0;\n if (convertable) {\n convertNodeTo($span, $, 'p');\n }\n });\n\n return $;\n}\n\n// Loop through the provided doc, and convert any p-like elements to\n// actual paragraph tags.\n//\n// Things fitting this criteria:\n// * Multiple consecutive
tags.\n// * tags without block level elements inside of them\n// * tags who are not children of or tags.\n//\n// :param $: A cheerio object to search\n// :return cheerio object with new p elements\n// (By-reference mutation, though. Returned just for convenience.)\n\nexport default function convertToParagraphs($) {\n $ = brsToPs($);\n $ = convertDivs($);\n $ = convertSpans($);\n\n return $;\n}\n","export default function convertNodeTo($node, $, tag = 'p') {\n const node = $node.get(0);\n if (!node) {\n return $;\n }\n const { attribs } = $node.get(0);\n const attribString = Reflect.ownKeys(attribs)\n .map(key => `${key}=${attribs[key]}`)\n .join(' ');\n\n $node.replaceWith(`<${tag} ${attribString}>${$node.contents()}`);\n return $;\n}\n","import { SPACER_RE } from './constants';\n\nfunction cleanForHeight($img, $) {\n const height = parseInt($img.attr('height'), 10);\n const width = parseInt($img.attr('width'), 10) || 20;\n\n // Remove images that explicitly have very small heights or\n // widths, because they are most likely shims or icons,\n // which aren't very useful for reading.\n if ((height || 20) < 10 || width < 10) {\n $img.remove();\n } else if (height) {\n // Don't ever specify a height on images, so that we can\n // scale with respect to width without screwing up the\n // aspect ratio.\n $img.removeAttr('height');\n }\n\n return $;\n}\n\n// Cleans out images where the source string matches transparent/spacer/etc\n// TODO This seems very aggressive - AP\nfunction removeSpacers($img, $) {\n if (SPACER_RE.test($img.attr('src'))) {\n $img.remove();\n }\n\n return $;\n}\n\nexport default function cleanImages($article, $) {\n $article.find('img').each((index, img) => {\n const $img = $(img);\n\n cleanForHeight($img, $);\n removeSpacers($img, $);\n });\n\n return $;\n}\n","import {\n KEEP_SELECTORS,\n KEEP_CLASS,\n} from './constants';\n\nexport default function markToKeep(article, $, tags = []) {\n if (tags.length === 0) {\n tags = KEEP_SELECTORS;\n }\n\n $(tags.join(','), article).addClass(KEEP_CLASS);\n\n return $;\n}\n","import {\n STRIP_OUTPUT_TAGS,\n KEEP_CLASS,\n} from './constants';\n\nexport default function stripJunkTags(article, $, tags = []) {\n if (tags.length === 0) {\n tags = STRIP_OUTPUT_TAGS;\n }\n\n // Remove matching elements, but ignore\n // any element with a class of mercury-parser-keep\n $(tags.join(','), article).not(`.${KEEP_CLASS}`).remove();\n\n // Remove the mercury-parser-keep class from result\n $(`.${KEEP_CLASS}`, article).removeClass(KEEP_CLASS);\n\n return $;\n}\n","import { convertNodeTo } from 'utils/dom';\n\n// H1 tags are typically the article title, which should be extracted\n// by the title extractor instead. If there's less than 3 of them (<3),\n// strip them. Otherwise, turn 'em into H2s.\nexport default function cleanHOnes(article, $) {\n const $hOnes = $('h1', article);\n\n if ($hOnes.length < 3) {\n $hOnes.each((index, node) => $(node).remove());\n } else {\n $hOnes.each((index, node) => {\n convertNodeTo($(node), $, 'h2');\n });\n }\n\n return $;\n}\n","import { WHITELIST_ATTRS_RE } from './constants';\n\nfunction removeAllButWhitelist($article) {\n // $('*', article).each((index, node) => {\n $article.find('*').each((index, node) => {\n node.attribs = Reflect.ownKeys(node.attribs).reduce((acc, attr) => {\n if (WHITELIST_ATTRS_RE.test(attr)) {\n return { ...acc, [attr]: node.attribs[attr] };\n }\n\n return acc;\n }, {});\n });\n}\n\n// function removeAttrs(article, $) {\n// REMOVE_ATTRS.forEach((attr) => {\n// $(`[${attr}]`, article).removeAttr(attr);\n// });\n// }\n\n// Remove attributes like style or align\nexport default function cleanAttributes($article) {\n removeAllButWhitelist($article);\n\n return $article;\n}\n","export default function removeEmpty($article, $) {\n $article.find('p').each((index, p) => {\n const $p = $(p);\n if ($p.find('iframe, img').length === 0 && $p.text().trim() === '') $p.remove();\n });\n\n return $;\n}\n","// // CONTENT FETCHING CONSTANTS ////\n\n// A list of strings that can be considered unlikely candidates when\n// extracting content from a resource. These strings are joined together\n// and then tested for existence using re:test, so may contain simple,\n// non-pipe style regular expression queries if necessary.\nexport const UNLIKELY_CANDIDATES_BLACKLIST = [\n 'ad-break',\n 'adbox',\n 'advert',\n 'addthis',\n 'agegate',\n 'aux',\n 'blogger-labels',\n 'combx',\n 'comment',\n 'conversation',\n 'disqus',\n 'entry-unrelated',\n 'extra',\n 'foot',\n 'form',\n 'header',\n 'hidden',\n 'loader',\n 'login', // Note: This can hit 'blogindex'.\n 'menu',\n 'meta',\n 'nav',\n 'pager',\n 'pagination',\n 'predicta', // readwriteweb inline ad box\n 'presence_control_external', // lifehacker.com container full of false positives\n 'popup',\n 'printfriendly',\n 'related',\n 'remove',\n 'remark',\n 'rss',\n 'share',\n 'shoutbox',\n 'sidebar',\n 'sociable',\n 'sponsor',\n 'tools',\n];\n\n// A list of strings that can be considered LIKELY candidates when\n// extracting content from a resource. Essentially, the inverse of the\n// blacklist above - if something matches both blacklist and whitelist,\n// it is kept. This is useful, for example, if something has a className\n// of \"rss-content entry-content\". It matched 'rss', so it would normally\n// be removed, however, it's also the entry content, so it should be left\n// alone.\n//\n// These strings are joined together and then tested for existence using\n// re:test, so may contain simple, non-pipe style regular expression queries\n// if necessary.\nexport const UNLIKELY_CANDIDATES_WHITELIST = [\n 'and',\n 'article',\n 'body',\n 'blogindex',\n 'column',\n 'content',\n 'entry-content-asset',\n 'format', // misuse of form\n 'hfeed',\n 'hentry',\n 'hatom',\n 'main',\n 'page',\n 'posts',\n 'shadow',\n];\n\n// A list of tags which, if found inside, should cause a to NOT\n// be turned into a paragraph tag. Shallow div tags without these elements\n// should be turned into tags.\nexport const DIV_TO_P_BLOCK_TAGS = [\n 'a',\n 'blockquote',\n 'dl',\n 'div',\n 'img',\n 'p',\n 'pre',\n 'table',\n].join(',');\n\n// A list of tags that should be ignored when trying to find the top candidate\n// for a document.\nexport const NON_TOP_CANDIDATE_TAGS = [\n 'br',\n 'b',\n 'i',\n 'label',\n 'hr',\n 'area',\n 'base',\n 'basefont',\n 'input',\n 'img',\n 'link',\n 'meta',\n];\n\nexport const NON_TOP_CANDIDATE_TAGS_RE =\n new RegExp(`^(${NON_TOP_CANDIDATE_TAGS.join('|')})$`, 'i');\n\n// A list of selectors that specify, very clearly, either hNews or other\n// very content-specific style content, like Blogger templates.\n// More examples here: http://microformats.org/wiki/blog-post-formats\nexport const HNEWS_CONTENT_SELECTORS = [\n ['.hentry', '.entry-content'],\n ['entry', '.entry-content'],\n ['.entry', '.entry_content'],\n ['.post', '.postbody'],\n ['.post', '.post_body'],\n ['.post', '.post-body'],\n];\n\nexport const PHOTO_HINTS = [\n 'figure',\n 'photo',\n 'image',\n 'caption',\n];\nexport const PHOTO_HINTS_RE = new RegExp(PHOTO_HINTS.join('|'), 'i');\n\n\n// A list of strings that denote a positive scoring for this content as being\n// an article container. Checked against className and id.\n//\n// TODO: Perhaps have these scale based on their odds of being quality?\nexport const POSITIVE_SCORE_HINTS = [\n 'article',\n 'articlecontent',\n 'instapaper_body',\n 'blog',\n 'body',\n 'content',\n 'entry-content-asset',\n 'entry',\n 'hentry',\n 'main',\n 'Normal',\n 'page',\n 'pagination',\n 'permalink',\n 'post',\n 'story',\n 'text',\n '[-_]copy', // usatoday\n '\\\\Bcopy',\n];\n\n// The above list, joined into a matching regular expression\nexport const POSITIVE_SCORE_RE = new RegExp(POSITIVE_SCORE_HINTS.join('|'), 'i');\n\n// Readability publisher-specific guidelines\nexport const READABILITY_ASSET = new RegExp('entry-content-asset', 'i');\n\n// A list of strings that denote a negative scoring for this content as being\n// an article container. Checked against className and id.\n//\n// TODO: Perhaps have these scale based on their odds of being quality?\nexport const NEGATIVE_SCORE_HINTS = [\n 'adbox',\n 'advert',\n 'author',\n 'bio',\n 'bookmark',\n 'bottom',\n 'byline',\n 'clear',\n 'com-',\n 'combx',\n 'comment',\n 'comment\\\\B',\n 'contact',\n 'copy',\n 'credit',\n 'crumb',\n 'date',\n 'deck',\n 'excerpt',\n 'featured', // tnr.com has a featured_content which throws us off\n 'foot',\n 'footer',\n 'footnote',\n 'graf',\n 'head',\n 'info',\n 'infotext', // newscientist.com copyright\n 'instapaper_ignore',\n 'jump',\n 'linebreak',\n 'link',\n 'masthead',\n 'media',\n 'meta',\n 'modal',\n 'outbrain', // slate.com junk\n 'promo',\n 'pr_', // autoblog - press release\n 'related',\n 'respond',\n 'roundcontent', // lifehacker restricted content warning\n 'scroll',\n 'secondary',\n 'share',\n 'shopping',\n 'shoutbox',\n 'side',\n 'sidebar',\n 'sponsor',\n 'stamp',\n 'sub',\n 'summary',\n 'tags',\n 'tools',\n 'widget',\n];\n// The above list, joined into a matching regular expression\nexport const NEGATIVE_SCORE_RE = new RegExp(NEGATIVE_SCORE_HINTS.join('|'), 'i');\n\n// Match a digit. Pretty clear.\nexport const DIGIT_RE = new RegExp('[0-9]');\n\n// Match 2 or more consecutive
tags\nexport const BR_TAGS_RE = new RegExp('(]*>[ \\n\\r\\t]*){2,}', 'i');\n\n// Match 1 BR tag.\nexport const BR_TAG_RE = new RegExp(']*>', 'i');\n\n// A list of all of the block level tags known in HTML5 and below. Taken from\n// http://bit.ly/qneNIT\nexport const BLOCK_LEVEL_TAGS = [\n 'article',\n 'aside',\n 'blockquote',\n 'body',\n 'br',\n 'button',\n 'canvas',\n 'caption',\n 'col',\n 'colgroup',\n 'dd',\n 'div',\n 'dl',\n 'dt',\n 'embed',\n 'fieldset',\n 'figcaption',\n 'figure',\n 'footer',\n 'form',\n 'h1',\n 'h2',\n 'h3',\n 'h4',\n 'h5',\n 'h6',\n 'header',\n 'hgroup',\n 'hr',\n 'li',\n 'map',\n 'object',\n 'ol',\n 'output',\n 'p',\n 'pre',\n 'progress',\n 'section',\n 'table',\n 'tbody',\n 'textarea',\n 'tfoot',\n 'th',\n 'thead',\n 'tr',\n 'ul',\n 'video',\n];\nexport const BLOCK_LEVEL_TAGS_RE = new RegExp(`^(${BLOCK_LEVEL_TAGS.join('|')})$`, 'i');\n\n\n// The removal is implemented as a blacklist and whitelist, this test finds\n// blacklisted elements that aren't whitelisted. We do this all in one\n// expression-both because it's only one pass, and because this skips the\n// serialization for whitelisted nodes.\nconst candidatesBlacklist = UNLIKELY_CANDIDATES_BLACKLIST.join('|');\nexport const CANDIDATES_BLACKLIST = new RegExp(candidatesBlacklist, 'i');\n\nconst candidatesWhitelist = UNLIKELY_CANDIDATES_WHITELIST.join('|');\nexport const CANDIDATES_WHITELIST = new RegExp(candidatesWhitelist, 'i');\n\nexport const UNLIKELY_RE = new RegExp(`!(${candidatesWhitelist})|(${candidatesBlacklist})`, 'i');\n\n\nexport const PARAGRAPH_SCORE_TAGS = new RegExp('^(p|li|span|pre)$', 'i');\nexport const CHILD_CONTENT_TAGS = new RegExp('^(td|blockquote|ol|ul|dl)$', 'i');\nexport const BAD_TAGS = new RegExp('^(address|form)$', 'i');\n\nexport const HTML_OR_BODY_RE = new RegExp('^(html|body)$', 'i');\n","import {\n NEGATIVE_SCORE_RE,\n POSITIVE_SCORE_RE,\n PHOTO_HINTS_RE,\n READABILITY_ASSET,\n} from './constants';\n\n\n// Get the score of a node based on its className and id.\nexport default function getWeight(node) {\n const classes = node.attr('class');\n const id = node.attr('id');\n let score = 0;\n\n if (id) {\n // if id exists, try to score on both positive and negative\n if (POSITIVE_SCORE_RE.test(id)) {\n score += 25;\n }\n if (NEGATIVE_SCORE_RE.test(id)) {\n score -= 25;\n }\n }\n\n if (classes) {\n if (score === 0) {\n // if classes exist and id did not contribute to score\n // try to score on both positive and negative\n if (POSITIVE_SCORE_RE.test(classes)) {\n score += 25;\n }\n if (NEGATIVE_SCORE_RE.test(classes)) {\n score -= 25;\n }\n }\n\n // even if score has been set by id, add score for\n // possible photo matches\n // \"try to keep photos if we can\"\n if (PHOTO_HINTS_RE.test(classes)) {\n score += 10;\n }\n\n // add 25 if class matches entry-content-asset,\n // a class apparently instructed for use in the\n // Readability publisher guidelines\n // https://www.readability.com/developers/guidelines\n if (READABILITY_ASSET.test(classes)) {\n score += 25;\n }\n }\n\n return score;\n}\n\n","// returns the score of a node based on\n// the node's score attribute\n// returns null if no score set\nexport default function getScore($node) {\n return parseFloat($node.attr('score')) || null;\n}\n","// return 1 for every comma in text\nexport default function scoreCommas(text) {\n return (text.match(/,/g) || []).length;\n}\n\n","const idkRe = new RegExp('^(p|pre)$', 'i');\n\nexport default function scoreLength(textLength, tagName = 'p') {\n const chunks = textLength / 50;\n\n if (chunks > 0) {\n let lengthBonus;\n\n // No idea why p or pre are being tamped down here\n // but just following the source for now\n // Not even sure why tagName is included here,\n // since this is only being called from the context\n // of scoreParagraph\n if (idkRe.test(tagName)) {\n lengthBonus = chunks - 2;\n } else {\n lengthBonus = chunks - 1.25;\n }\n\n return Math.min(Math.max(lengthBonus, 0), 3);\n }\n\n return 0;\n}\n\n","import {\n scoreCommas,\n scoreLength,\n} from './index';\n\n// Score a paragraph using various methods. Things like number of\n// commas, etc. Higher is better.\nexport default function scoreParagraph(node) {\n let score = 1;\n const text = node.text().trim();\n const textLength = text.length;\n\n // If this paragraph is less than 25 characters, don't count it.\n if (textLength < 25) {\n return 0;\n }\n\n // Add points for any commas within this paragraph\n score += scoreCommas(text);\n\n // For every 50 characters in this paragraph, add another point. Up\n // to 3 points.\n score += scoreLength(textLength);\n\n // Articles can end with short paragraphs when people are being clever\n // but they can also end with short paragraphs setting up lists of junk\n // that we strip. This negative tweaks junk setup paragraphs just below\n // the cutoff threshold.\n if (text.slice(-1) === ':') {\n score -= 1;\n }\n\n return score;\n}\n\n","\nexport default function setScore($node, $, score) {\n $node.attr('score', score);\n return $node;\n}\n\n","import {\n getOrInitScore,\n setScore,\n} from './index';\n\nexport default function addScore($node, $, amount) {\n try {\n const score = getOrInitScore($node, $) + amount;\n setScore($node, $, score);\n } catch (e) {\n // Ignoring; error occurs in scoreNode\n }\n\n return $node;\n}\n","import { addScore } from './index';\n\n// Adds 1/4 of a child's score to its parent\nexport default function addToParent(node, $, score) {\n const parent = node.parent();\n if (parent) {\n addScore(parent, $, score * 0.25);\n }\n\n return node;\n}\n","import {\n getScore,\n scoreNode,\n getWeight,\n addToParent,\n} from './index';\n\n// gets and returns the score if it exists\n// if not, initializes a score based on\n// the node's tag type\nexport default function getOrInitScore($node, $, weightNodes = true) {\n let score = getScore($node);\n\n if (score) {\n return score;\n }\n\n score = scoreNode($node);\n\n if (weightNodes) {\n score += getWeight($node);\n }\n\n addToParent($node, $, score);\n\n return score;\n}\n\n","import { scoreParagraph } from './index';\nimport {\n PARAGRAPH_SCORE_TAGS,\n CHILD_CONTENT_TAGS,\n BAD_TAGS,\n} from './constants';\n\n// Score an individual node. Has some smarts for paragraphs, otherwise\n// just scores based on tag.\nexport default function scoreNode($node) {\n const { tagName } = $node.get(0);\n\n // TODO: Consider ordering by most likely.\n // E.g., if divs are a more common tag on a page,\n // Could save doing that regex test on every node – AP\n if (PARAGRAPH_SCORE_TAGS.test(tagName)) {\n return scoreParagraph($node);\n } else if (tagName === 'div') {\n return 5;\n } else if (CHILD_CONTENT_TAGS.test(tagName)) {\n return 3;\n } else if (BAD_TAGS.test(tagName)) {\n return -3;\n } else if (tagName === 'th') {\n return -5;\n }\n\n return 0;\n}\n","import { convertNodeTo } from 'utils/dom';\n\nimport { HNEWS_CONTENT_SELECTORS } from './constants';\nimport {\n scoreNode,\n setScore,\n getOrInitScore,\n addScore,\n} from './index';\n\nfunction convertSpans($node, $) {\n if ($node.get(0)) {\n const { tagName } = $node.get(0);\n\n if (tagName === 'span') {\n // convert spans to divs\n convertNodeTo($node, $, 'div');\n }\n }\n}\n\nfunction addScoreTo($node, $, score) {\n if ($node) {\n convertSpans($node, $);\n addScore($node, $, score);\n }\n}\n\nfunction scorePs($, weightNodes) {\n $('p, pre').not('[score]').each((index, node) => {\n // The raw score for this paragraph, before we add any parent/child\n // scores.\n let $node = $(node);\n $node = setScore($node, $, getOrInitScore($node, $, weightNodes));\n\n const $parent = $node.parent();\n const rawScore = scoreNode($node);\n\n addScoreTo($parent, $, rawScore, weightNodes);\n if ($parent) {\n // Add half of the individual content score to the\n // grandparent\n addScoreTo($parent.parent(), $, rawScore / 2, weightNodes);\n }\n });\n\n return $;\n}\n\n// score content. Parents get the full value of their children's\n// content score, grandparents half\nexport default function scoreContent($, weightNodes = true) {\n // First, look for special hNews based selectors and give them a big\n // boost, if they exist\n HNEWS_CONTENT_SELECTORS.forEach(([parentSelector, childSelector]) => {\n $(`${parentSelector} ${childSelector}`).each((index, node) => {\n addScore($(node).parent(parentSelector), $, 80);\n });\n });\n\n // Doubling this again\n // Previous solution caused a bug\n // in which parents weren't retaining\n // scores. This is not ideal, and\n // should be fixed.\n scorePs($, weightNodes);\n scorePs($, weightNodes);\n\n return $;\n}\n","const NORMALIZE_RE = /\\s{2,}/g;\n\nexport default function normalizeSpaces(text) {\n return text.replace(NORMALIZE_RE, ' ').trim();\n}\n","// Given a node type to search for, and a list of regular expressions,\n// look to see if this extraction can be found in the URL. Expects\n// that each expression in r_list will return group(1) as the proper\n// string to be cleaned.\n// Only used for date_published currently.\nexport default function extractFromUrl(url, regexList) {\n const matchRe = regexList.find(re => re.test(url));\n if (matchRe) {\n return matchRe.exec(url)[1];\n }\n\n return null;\n}\n","// An expression that looks to try to find the page digit within a URL, if\n// it exists.\n// Matches:\n// page=1\n// pg=1\n// p=1\n// paging=12\n// pag=7\n// pagination/1\n// paging/88\n// pa/83\n// p/11\n//\n// Does not match:\n// pg=102\n// page:2\nexport const PAGE_IN_HREF_RE = new RegExp('(page|paging|(p(a|g|ag)?(e|enum|ewanted|ing|ination)))?(=|/)([0-9]{1,3})', 'i');\n\nexport const HAS_ALPHA_RE = /[a-z]/i;\n\nexport const IS_ALPHA_RE = /^[a-z]+$/i;\nexport const IS_DIGIT_RE = /^[0-9]+$/i;\n","import { PAGE_IN_HREF_RE } from './constants';\n\nexport default function pageNumFromUrl(url) {\n const matches = url.match(PAGE_IN_HREF_RE);\n if (!matches) return null;\n\n const pageNum = parseInt(matches[6], 10);\n\n // Return pageNum < 100, otherwise\n // return null\n return pageNum < 100 ? pageNum : null;\n}\n","export default function removeAnchor(url) {\n return url.split('#')[0].replace(/\\/$/, '');\n}\n","import URL from 'url';\nimport {\n HAS_ALPHA_RE,\n IS_ALPHA_RE,\n IS_DIGIT_RE,\n PAGE_IN_HREF_RE,\n} from './constants';\n\nfunction isGoodSegment(segment, index, firstSegmentHasLetters) {\n let goodSegment = true;\n\n // If this is purely a number, and it's the first or second\n // url_segment, it's probably a page number. Remove it.\n if (index < 2 && IS_DIGIT_RE.test(segment) && segment.length < 3) {\n goodSegment = true;\n }\n\n // If this is the first url_segment and it's just \"index\",\n // remove it\n if (index === 0 && segment.toLowerCase() === 'index') {\n goodSegment = false;\n }\n\n // If our first or second url_segment is smaller than 3 characters,\n // and the first url_segment had no alphas, remove it.\n if (index < 2 && segment.length < 3 && !firstSegmentHasLetters) {\n goodSegment = false;\n }\n\n return goodSegment;\n}\n\n// Take a URL, and return the article base of said URL. That is, no\n// pagination data exists in it. Useful for comparing to other links\n// that might have pagination data within them.\nexport default function articleBaseUrl(url, parsed) {\n const parsedUrl = parsed || URL.parse(url);\n const { protocol, host, path } = parsedUrl;\n\n let firstSegmentHasLetters = false;\n const cleanedSegments = path.split('/')\n .reverse()\n .reduce((acc, rawSegment, index) => {\n let segment = rawSegment;\n\n // Split off and save anything that looks like a file type.\n if (segment.includes('.')) {\n const [possibleSegment, fileExt] = segment.split('.');\n if (IS_ALPHA_RE.test(fileExt)) {\n segment = possibleSegment;\n }\n }\n\n // If our first or second segment has anything looking like a page\n // number, remove it.\n if (PAGE_IN_HREF_RE.test(segment) && index < 2) {\n segment = segment.replace(PAGE_IN_HREF_RE, '');\n }\n\n // If we're on the first segment, check to see if we have any\n // characters in it. The first segment is actually the last bit of\n // the URL, and this will be helpful to determine if we're on a URL\n // segment that looks like \"/2/\" for example.\n if (index === 0) {\n firstSegmentHasLetters = HAS_ALPHA_RE.test(segment);\n }\n\n // If it's not marked for deletion, push it to cleaned_segments.\n if (isGoodSegment(segment, index, firstSegmentHasLetters)) {\n acc.push(segment);\n }\n\n return acc;\n }, []);\n\n return `${protocol}//${host}${cleanedSegments.reverse().join('/')}`;\n}\n","// Given a string, return True if it appears to have an ending sentence\n// within it, false otherwise.\nconst SENTENCE_END_RE = new RegExp('.( |$)');\nexport default function hasSentenceEnd(text) {\n return SENTENCE_END_RE.test(text);\n}\n\n","export default function excerptContent(content, words = 10) {\n return content.trim()\n .split(/\\s+/)\n .slice(0, words)\n .join(' ');\n}\n","import {\n textLength,\n linkDensity,\n} from 'utils/dom';\nimport { hasSentenceEnd } from 'utils/text';\n\nimport { NON_TOP_CANDIDATE_TAGS_RE } from './constants';\nimport { getScore } from './index';\n\n// Now that we have a top_candidate, look through the siblings of\n// it to see if any of them are decently scored. If they are, they\n// may be split parts of the content (Like two divs, a preamble and\n// a body.) Example:\n// http://articles.latimes.com/2009/oct/14/business/fi-bigtvs14\nexport default function mergeSiblings($candidate, topScore, $) {\n if (!$candidate.parent().length) {\n return $candidate;\n }\n\n const siblingScoreThreshold = Math.max(10, topScore * 0.25);\n const wrappingDiv = $('');\n\n $candidate.parent().children().each((index, sibling) => {\n const $sibling = $(sibling);\n // Ignore tags like BR, HR, etc\n if (NON_TOP_CANDIDATE_TAGS_RE.test(sibling.tagName)) {\n return null;\n }\n\n const siblingScore = getScore($sibling);\n if (siblingScore) {\n if ($sibling === $candidate) {\n wrappingDiv.append($sibling);\n } else {\n let contentBonus = 0;\n const density = linkDensity($sibling);\n\n // If sibling has a very low link density,\n // give it a small bonus\n if (density < 0.05) {\n contentBonus += 20;\n }\n\n // If sibling has a high link density,\n // give it a penalty\n if (density >= 0.5) {\n contentBonus -= 20;\n }\n\n // If sibling node has the same class as\n // candidate, give it a bonus\n if ($sibling.attr('class') === $candidate.attr('class')) {\n contentBonus += topScore * 0.2;\n }\n\n const newScore = siblingScore + contentBonus;\n\n if (newScore >= siblingScoreThreshold) {\n return wrappingDiv.append($sibling);\n } else if (sibling.tagName === 'p') {\n const siblingContent = $sibling.text();\n const siblingContentLength = textLength(siblingContent);\n\n if (siblingContentLength > 80 && density < 0.25) {\n return wrappingDiv.append($sibling);\n } else if (siblingContentLength <= 80 && density === 0 &&\n hasSentenceEnd(siblingContent)) {\n return wrappingDiv.append($sibling);\n }\n }\n }\n }\n\n return null;\n });\n\n return wrappingDiv;\n}\n","import { NON_TOP_CANDIDATE_TAGS_RE } from './constants';\nimport { getScore } from './index';\nimport mergeSiblings from './merge-siblings';\n\n// After we've calculated scores, loop through all of the possible\n// candidate nodes we found and find the one with the highest score.\nexport default function findTopCandidate($) {\n let $candidate;\n let topScore = 0;\n\n $('[score]').each((index, node) => {\n // Ignore tags like BR, HR, etc\n if (NON_TOP_CANDIDATE_TAGS_RE.test(node.tagName)) {\n return;\n }\n\n const $node = $(node);\n const score = getScore($node);\n\n if (score > topScore) {\n topScore = score;\n $candidate = $node;\n }\n });\n\n // If we don't have a candidate, return the body\n // or whatever the first element is\n if (!$candidate) {\n return $('body') || $('*').first();\n }\n\n $candidate = mergeSiblings($candidate, topScore, $);\n\n return $candidate;\n}\n","import {\n getScore,\n setScore,\n getOrInitScore,\n scoreCommas,\n} from 'extractors/generic/content/scoring';\n\nimport { CLEAN_CONDITIONALLY_TAGS } from './constants';\nimport { normalizeSpaces } from '../text';\nimport { linkDensity } from './index';\n\nfunction removeUnlessContent($node, $, weight) {\n // Explicitly save entry-content-asset tags, which are\n // noted as valuable in the Publisher guidelines. For now\n // this works everywhere. We may want to consider making\n // this less of a sure-thing later.\n if ($node.hasClass('entry-content-asset')) {\n return;\n }\n\n const content = normalizeSpaces($node.text());\n\n if (scoreCommas(content) < 10) {\n const pCount = $('p', $node).length;\n const inputCount = $('input', $node).length;\n\n // Looks like a form, too many inputs.\n if (inputCount > (pCount / 3)) {\n $node.remove();\n return;\n }\n\n const contentLength = content.length;\n const imgCount = $('img', $node).length;\n\n // Content is too short, and there are no images, so\n // this is probably junk content.\n if (contentLength < 25 && imgCount === 0) {\n $node.remove();\n return;\n }\n\n const density = linkDensity($node);\n\n // Too high of link density, is probably a menu or\n // something similar.\n // console.log(weight, density, contentLength)\n if (weight < 25 && density > 0.2 && contentLength > 75) {\n $node.remove();\n return;\n }\n\n // Too high of a link density, despite the score being\n // high.\n if (weight >= 25 && density > 0.5) {\n // Don't remove the node if it's a list and the\n // previous sibling starts with a colon though. That\n // means it's probably content.\n const tagName = $node.get(0).tagName;\n const nodeIsList = tagName === 'ol' || tagName === 'ul';\n if (nodeIsList) {\n const previousNode = $node.prev();\n if (previousNode && normalizeSpaces(previousNode.text()).slice(-1) === ':') {\n return;\n }\n }\n\n $node.remove();\n return;\n }\n\n const scriptCount = $('script', $node).length;\n\n // Too many script tags, not enough content.\n if (scriptCount > 0 && contentLength < 150) {\n $node.remove();\n return;\n }\n }\n}\n\n// Given an article, clean it of some superfluous content specified by\n// tags. Things like forms, ads, etc.\n//\n// Tags is an array of tag name's to search through. (like div, form,\n// etc)\n//\n// Return this same doc.\nexport default function cleanTags($article, $) {\n $(CLEAN_CONDITIONALLY_TAGS, $article).each((index, node) => {\n const $node = $(node);\n let weight = getScore($node);\n if (!weight) {\n weight = getOrInitScore($node, $);\n setScore($node, $, weight);\n }\n\n // drop node if its weight is < 0\n if (weight < 0) {\n $node.remove();\n } else {\n // deteremine if node seems like content\n removeUnlessContent($node, $, weight);\n }\n });\n\n return $;\n}\n\n","import { getWeight } from 'extractors/generic/content/scoring';\n\nimport { HEADER_TAG_LIST } from './constants';\nimport { normalizeSpaces } from '../text';\n\nexport default function cleanHeaders($article, $, title = '') {\n $(HEADER_TAG_LIST, $article).each((index, header) => {\n const $header = $(header);\n // Remove any headers that appear before all other p tags in the\n // document. This probably means that it was part of the title, a\n // subtitle or something else extraneous like a datestamp or byline,\n // all of which should be handled by other metadata handling.\n if ($($header, $article).prevAll('p').length === 0) {\n return $header.remove();\n }\n\n // Remove any headers that match the title exactly.\n if (normalizeSpaces($(header).text()) === title) {\n return $header.remove();\n }\n\n // If this header has a negative weight, it's probably junk.\n // Get rid of it.\n if (getWeight($(header)) < 0) {\n return $header.remove();\n }\n\n return $header;\n });\n\n return $;\n}\n","import { convertNodeTo } from 'utils/dom';\n\n// Rewrite the tag name to div if it's a top level node like body or\n// html to avoid later complications with multiple body tags.\nexport default function rewriteTopLevel(article, $) {\n // I'm not using context here because\n // it's problematic when converting the\n // top-level/root node - AP\n $ = convertNodeTo($('html'), $, 'div');\n $ = convertNodeTo($('body'), $, 'div');\n\n return $;\n}\n","import URL from 'url';\n\nfunction absolutize($, rootUrl, attr, $content) {\n $(`[${attr}]`, $content).each((_, node) => {\n const url = node.attribs[attr];\n const absoluteUrl = URL.resolve(rootUrl, url);\n\n node.attribs[attr] = absoluteUrl;\n });\n}\n\nexport default function makeLinksAbsolute($content, $, url) {\n ['href', 'src'].forEach(attr => absolutize($, url, attr, $content));\n\n return $content;\n}\n","\nexport function textLength(text) {\n return text.trim()\n .replace(/\\s+/g, ' ')\n .length;\n}\n\n// Determines what percentage of the text\n// in a node is link text\n// Takes a node, returns a float\nexport function linkDensity($node) {\n const totalTextLength = textLength($node.text());\n\n const linkText = $node.find('a').text();\n const linkLength = textLength(linkText);\n\n if (totalTextLength > 0) {\n return linkLength / totalTextLength;\n } else if (totalTextLength === 0 && linkLength > 0) {\n return 1;\n }\n\n return 0;\n}\n","import { stripTags } from 'utils/dom';\n\n// Given a node type to search for, and a list of meta tag names to\n// search for, find a meta tag associated.\nexport default function extractFromMeta(\n $,\n metaNames,\n cachedNames,\n cleanTags = true\n) {\n const foundNames = metaNames.filter(name => cachedNames.indexOf(name) !== -1);\n\n for (const name of foundNames) {\n const type = 'name';\n const value = 'value';\n\n const nodes = $(`meta[${type}=\"${name}\"]`);\n\n // Get the unique value of every matching node, in case there\n // are two meta tags with the same name and value.\n // Remove empty values.\n const values =\n nodes.map((index, node) => $(node).attr(value))\n .toArray()\n .filter(text => text !== '');\n\n // If we have more than one value for the same name, we have a\n // conflict and can't trust any of them. Skip this name. If we have\n // zero, that means our meta tags had no values. Skip this name\n // also.\n if (values.length === 1) {\n let metaValue;\n // Meta values that contain HTML should be stripped, as they\n // weren't subject to cleaning previously.\n if (cleanTags) {\n metaValue = stripTags(values[0], $);\n } else {\n metaValue = values[0];\n }\n\n return metaValue;\n }\n }\n\n // If nothing is found, return null\n return null;\n}\n","import { withinComment } from 'utils/dom';\n\nfunction isGoodNode($node, maxChildren) {\n // If it has a number of children, it's more likely a container\n // element. Skip it.\n if ($node.children().length > maxChildren) {\n return false;\n }\n // If it looks to be within a comment, skip it.\n if (withinComment($node)) {\n return false;\n }\n\n return true;\n}\n\n// Given a a list of selectors find content that may\n// be extractable from the document. This is for flat\n// meta-information, like author, title, date published, etc.\nexport default function extractFromSelectors(\n $,\n selectors,\n maxChildren = 1,\n textOnly = true\n) {\n for (const selector of selectors) {\n const nodes = $(selector);\n\n // If we didn't get exactly one of this selector, this may be\n // a list of articles or comments. Skip it.\n if (nodes.length === 1) {\n const $node = $(nodes[0]);\n\n if (isGoodNode($node, maxChildren)) {\n let content;\n if (textOnly) {\n content = $node.text();\n } else {\n content = $node.html();\n }\n\n if (content) {\n return content;\n }\n }\n }\n }\n\n return null;\n}\n","// strips all tags from a string of text\nexport default function stripTags(text, $) {\n // Wrapping text in html element prevents errors when text\n // has no html\n const cleanText = $(`${text}`).text();\n return cleanText === '' ? text : cleanText;\n}\n","export default function withinComment($node) {\n const parents = $node.parents().toArray();\n const commentParent = parents.find((parent) => {\n const classAndId = `${parent.attribs.class} ${parent.attribs.id}`;\n return classAndId.includes('comment');\n });\n\n return commentParent !== undefined;\n}\n","// Given a node, determine if it's article-like enough to return\n// param: node (a cheerio node)\n// return: boolean\n\nexport default function nodeIsSufficient($node) {\n return $node.text().trim().length >= 100;\n}\n","import { IS_WP_SELECTOR } from './constants';\n\nexport default function isWordpress($) {\n return $(IS_WP_SELECTOR).length > 0;\n}\n","// CLEAN AUTHOR CONSTANTS\nexport const CLEAN_AUTHOR_RE = /^\\s*(posted |written )?by\\s*:?\\s*(.*)/i;\n // author = re.sub(r'^\\s*(posted |written )?by\\s*:?\\s*(.*)(?i)',\n\n// CLEAN DEK CONSTANTS\nexport const TEXT_LINK_RE = new RegExp('http(s)?://', 'i');\n// An ordered list of meta tag names that denote likely article deks.\n// From most distinct to least distinct.\n//\n// NOTE: There are currently no meta tags that seem to provide the right\n// content consistenty enough. Two options were:\n// - og:description\n// - dc.description\n// However, these tags often have SEO-specific junk in them that's not\n// header-worthy like a dek is. Excerpt material at best.\nexport const DEK_META_TAGS = [\n];\n\n// An ordered list of Selectors to find likely article deks. From\n// most explicit to least explicit.\n//\n// Should be more restrictive than not, as a failed dek can be pretty\n// detrimental to the aesthetics of an article.\nexport const DEK_SELECTORS = [\n '.entry-summary',\n];\n\n// CLEAN DATE PUBLISHED CONSTANTS\nexport const MS_DATE_STRING = /^\\d{13}$/i;\nexport const SEC_DATE_STRING = /^\\d{10}$/i;\nexport const CLEAN_DATE_STRING_RE = /^\\s*published\\s*:?\\s*(.*)/i;\nexport const TIME_MERIDIAN_SPACE_RE = /(.*\\d)(am|pm)(.*)/i;\nexport const TIME_MERIDIAN_DOTS_RE = /\\.m\\./i;\nconst months = [\n 'jan',\n 'feb',\n 'mar',\n 'apr',\n 'may',\n 'jun',\n 'jul',\n 'aug',\n 'sep',\n 'oct',\n 'nov',\n 'dec',\n];\nconst allMonths = months.join('|');\nconst timestamp1 = '[0-9]{1,2}:[0-9]{2,2}( ?[ap].?m.?)?';\nconst timestamp2 = '[0-9]{1,2}[/-][0-9]{1,2}[/-][0-9]{2,4}';\nexport const SPLIT_DATE_STRING =\n new RegExp(`(${timestamp1})|(${timestamp2})|([0-9]{1,4})|(${allMonths})`, 'ig');\n\n// CLEAN TITLE CONSTANTS\n// A regular expression that will match separating characters on a\n// title, that usually denote breadcrumbs or something similar.\nexport const TITLE_SPLITTERS_RE = /(: | - | \\| )/g;\n\nexport const DOMAIN_ENDINGS_RE =\n new RegExp('.com$|.net$|.org$|.co.uk$', 'g');\n","import { CLEAN_AUTHOR_RE } from './constants';\n\n// Take an author string (like 'By David Smith ') and clean it to\n// just the name(s): 'David Smith'.\nexport default function cleanAuthor(author) {\n return author.replace(CLEAN_AUTHOR_RE, '$2').trim();\n}\n","import validUrl from 'valid-url';\n\nexport default function clean(leadImageUrl) {\n leadImageUrl = leadImageUrl.trim();\n if (validUrl.isWebUri(leadImageUrl)) {\n return leadImageUrl;\n }\n\n return null;\n}\n","import { stripTags } from 'utils/dom';\nimport { excerptContent } from 'utils/text';\n\nimport { TEXT_LINK_RE } from './constants';\n\n// Take a dek HTML fragment, and return the cleaned version of it.\n// Return None if the dek wasn't good enough.\nexport default function cleanDek(dek, { $, excerpt }) {\n // Sanity check that we didn't get too short or long of a dek.\n if (dek.length > 1000 || dek.length < 5) return null;\n\n // Check that dek isn't the same as excerpt\n if (excerpt && excerptContent(excerpt, 10) === excerptContent(dek, 10)) return null;\n\n const dekText = stripTags(dek, $);\n\n // Plain text links shouldn't exist in the dek. If we have some, it's\n // not a good dek - bail.\n if (TEXT_LINK_RE.test(dekText)) return null;\n\n return dekText.trim();\n}\n","import moment from 'moment';\n// Is there a compelling reason to use moment here?\n// Mostly only being used for the isValid() method,\n// but could just check for 'Invalid Date' string.\n\nimport {\n MS_DATE_STRING,\n SEC_DATE_STRING,\n CLEAN_DATE_STRING_RE,\n SPLIT_DATE_STRING,\n TIME_MERIDIAN_SPACE_RE,\n TIME_MERIDIAN_DOTS_RE,\n} from './constants';\n\nexport function cleanDateString(dateString) {\n return (dateString.match(SPLIT_DATE_STRING) || [])\n .join(' ')\n .replace(TIME_MERIDIAN_DOTS_RE, 'm')\n .replace(TIME_MERIDIAN_SPACE_RE, '$1 $2 $3')\n .replace(CLEAN_DATE_STRING_RE, '$1')\n .trim();\n}\n\n// Take a date published string, and hopefully return a date out of\n// it. Return none if we fail.\nexport default function cleanDatePublished(dateString) {\n // If string is in milliseconds or seconds, convert to int\n if (MS_DATE_STRING.test(dateString) || SEC_DATE_STRING.test(dateString)) {\n dateString = parseInt(dateString, 10);\n }\n\n let date = moment(new Date(dateString));\n\n if (!date.isValid()) {\n dateString = cleanDateString(dateString);\n date = moment(new Date(dateString));\n }\n\n return date.isValid() ? date.toISOString() : null;\n}\n","import {\n cleanAttributes,\n cleanHeaders,\n cleanHOnes,\n cleanImages,\n cleanTags,\n removeEmpty,\n rewriteTopLevel,\n markToKeep,\n stripJunkTags,\n makeLinksAbsolute,\n} from 'utils/dom';\n\n// Clean our article content, returning a new, cleaned node.\nexport default function extractCleanNode(\n article,\n {\n $,\n cleanConditionally = true,\n title = '',\n url = '',\n defaultCleaner = true,\n }\n) {\n // Rewrite the tag name to div if it's a top level node like body or\n // html to avoid later complications with multiple body tags.\n rewriteTopLevel(article, $);\n\n // Drop small images and spacer images\n // Only do this is defaultCleaner is set to true;\n // this can sometimes be too aggressive.\n if (defaultCleaner) cleanImages(article, $);\n\n // Mark elements to keep that would normally be removed.\n // E.g., stripJunkTags will remove iframes, so we're going to mark\n // YouTube/Vimeo videos as elements we want to keep.\n markToKeep(article, $);\n\n // Drop certain tags like , etc\n // This is -mostly- for cleanliness, not security.\n stripJunkTags(article, $);\n\n // H1 tags are typically the article title, which should be extracted\n // by the title extractor instead. If there's less than 3 of them (<3),\n // strip them. Otherwise, turn 'em into H2s.\n cleanHOnes(article, $);\n\n // Clean headers\n cleanHeaders(article, $, title);\n\n // Make links absolute\n makeLinksAbsolute(article, $, url);\n\n // Remove unnecessary attributes\n cleanAttributes(article);\n\n // We used to clean UL's and OL's here, but it was leading to\n // too many in-article lists being removed. Consider a better\n // way to detect menus particularly and remove them.\n // Also optionally running, since it can be overly aggressive.\n if (defaultCleaner) cleanTags(article, $, cleanConditionally);\n\n // Remove empty paragraph nodes\n removeEmpty(article, $);\n\n return article;\n}\n","import { stripTags } from 'utils/dom';\n\nimport { TITLE_SPLITTERS_RE } from './constants';\nimport { resolveSplitTitle } from './index';\n\nexport default function cleanTitle(title, { url, $ }) {\n // If title has |, :, or - in it, see if\n // we can clean it up.\n if (TITLE_SPLITTERS_RE.test(title)) {\n title = resolveSplitTitle(title, url);\n }\n\n // Final sanity check that we didn't get a crazy title.\n // if (title.length > 150 || title.length < 15) {\n if (title.length > 150) {\n // If we did, return h1 from the document if it exists\n const h1 = $('h1');\n if (h1.length === 1) {\n title = h1.text();\n }\n }\n\n // strip any html tags in the title text\n return stripTags(title, $).trim();\n}\n\n","import URL from 'url';\nimport wuzzy from 'wuzzy';\n\nimport {\n TITLE_SPLITTERS_RE,\n DOMAIN_ENDINGS_RE,\n} from './constants';\n\nfunction extractBreadcrumbTitle(splitTitle, text) {\n // This must be a very breadcrumbed title, like:\n // The Best Gadgets on Earth : Bits : Blogs : NYTimes.com\n // NYTimes - Blogs - Bits - The Best Gadgets on Earth\n if (splitTitle.length >= 6) {\n // Look to see if we can find a breadcrumb splitter that happens\n // more than once. If we can, we'll be able to better pull out\n // the title.\n const termCounts = splitTitle.reduce((acc, titleText) => {\n acc[titleText] = acc[titleText] ? acc[titleText] + 1 : 1;\n return acc;\n }, {});\n\n const [maxTerm, termCount] =\n Reflect.ownKeys(termCounts)\n .reduce((acc, key) => {\n if (acc[1] < termCounts[key]) {\n return [key, termCounts[key]];\n }\n\n return acc;\n }, [0, 0]);\n\n // We found a splitter that was used more than once, so it\n // is probably the breadcrumber. Split our title on that instead.\n // Note: max_term should be <= 4 characters, so that \" >> \"\n // will match, but nothing longer than that.\n if (termCount >= 2 && maxTerm.length <= 4) {\n splitTitle = text.split(maxTerm);\n }\n\n const splitEnds = [splitTitle[0], splitTitle.slice(-1)];\n const longestEnd = splitEnds.reduce((acc, end) => acc.length > end.length ? acc : end, '');\n\n if (longestEnd.length > 10) {\n return longestEnd;\n }\n\n return text;\n }\n\n return null;\n}\n\nfunction cleanDomainFromTitle(splitTitle, url) {\n // Search the ends of the title, looking for bits that fuzzy match\n // the URL too closely. If one is found, discard it and return the\n // rest.\n //\n // Strip out the big TLDs - it just makes the matching a bit more\n // accurate. Not the end of the world if it doesn't strip right.\n const { host } = URL.parse(url);\n const nakedDomain = host.replace(DOMAIN_ENDINGS_RE, '');\n\n const startSlug = splitTitle[0].toLowerCase().replace(' ', '');\n const startSlugRatio = wuzzy.levenshtein(startSlug, nakedDomain);\n\n if (startSlugRatio > 0.4 && startSlug.length > 5) {\n return splitTitle.slice(2).join('');\n }\n\n const endSlug = splitTitle.slice(-1)[0].toLowerCase().replace(' ', '');\n const endSlugRatio = wuzzy.levenshtein(endSlug, nakedDomain);\n\n if (endSlugRatio > 0.4 && endSlug.length >= 5) {\n return splitTitle.slice(0, -2).join('');\n }\n\n return null;\n}\n\n// Given a title with separators in it (colons, dashes, etc),\n// resolve whether any of the segments should be removed.\nexport default function resolveSplitTitle(title, url = '') {\n // Splits while preserving splitters, like:\n // ['The New New York', ' - ', 'The Washington Post']\n const splitTitle = title.split(TITLE_SPLITTERS_RE);\n if (splitTitle.length === 1) {\n return title;\n }\n\n let newTitle = extractBreadcrumbTitle(splitTitle, title);\n if (newTitle) return newTitle;\n\n newTitle = cleanDomainFromTitle(splitTitle, url);\n if (newTitle) return newTitle;\n\n // Fuzzy ratio didn't find anything, so this title is probably legit.\n // Just return it all.\n return title;\n}\n","import cleanAuthor from './author';\nimport cleanImage from './lead-image-url';\nimport cleanDek from './dek';\nimport cleanDatePublished from './date-published';\nimport cleanContent from './content';\nimport cleanTitle from './title';\n\nconst Cleaners = {\n author: cleanAuthor,\n lead_image_url: cleanImage,\n dek: cleanDek,\n date_published: cleanDatePublished,\n content: cleanContent,\n title: cleanTitle,\n};\n\n\nexport default Cleaners;\n\nexport { cleanAuthor };\nexport { cleanImage };\nexport { cleanDek };\nexport { cleanDatePublished };\nexport { cleanContent };\nexport { cleanTitle };\nexport { default as resolveSplitTitle } from './resolve-split-title';\n","import {\n stripUnlikelyCandidates,\n convertToParagraphs,\n} from 'utils/dom';\n\nimport {\n scoreContent,\n findTopCandidate,\n} from './scoring';\n\n// Using a variety of scoring techniques, extract the content most\n// likely to be article text.\n//\n// If strip_unlikely_candidates is True, remove any elements that\n// match certain criteria first. (Like, does this element have a\n// classname of \"comment\")\n//\n// If weight_nodes is True, use classNames and IDs to determine the\n// worthiness of nodes.\n//\n// Returns a cheerio object $\nexport default function extractBestNode($, opts) {\n // clone the node so we can get back to our\n // initial parsed state if needed\n // TODO Do I need this? – AP\n // let $root = $.root().clone()\n\n\n if (opts.stripUnlikelyCandidates) {\n $ = stripUnlikelyCandidates($);\n }\n\n $ = convertToParagraphs($);\n $ = scoreContent($, opts.weightNodes);\n const $topCandidate = findTopCandidate($);\n\n return $topCandidate;\n}\n","import cheerio from 'cheerio';\n\nimport { nodeIsSufficient } from 'utils/dom';\nimport { cleanContent } from 'cleaners';\nimport { normalizeSpaces } from 'utils/text';\n\nimport extractBestNode from './extract-best-node';\n\nconst GenericContentExtractor = {\n defaultOpts: {\n stripUnlikelyCandidates: true,\n weightNodes: true,\n cleanConditionally: true,\n },\n\n // Extract the content for this resource - initially, pass in our\n // most restrictive opts which will return the highest quality\n // content. On each failure, retry with slightly more lax opts.\n //\n // :param return_type: string. If \"node\", should return the content\n // as a cheerio node rather than as an HTML string.\n //\n // Opts:\n // stripUnlikelyCandidates: Remove any elements that match\n // non-article-like criteria first.(Like, does this element\n // have a classname of \"comment\")\n //\n // weightNodes: Modify an elements score based on whether it has\n // certain classNames or IDs. Examples: Subtract if a node has\n // a className of 'comment', Add if a node has an ID of\n // 'entry-content'.\n //\n // cleanConditionally: Clean the node to return of some\n // superfluous content. Things like forms, ads, etc.\n extract({ $, html, title, url }, opts) {\n opts = { ...this.defaultOpts, ...opts };\n\n $ = $ || cheerio.load(html);\n\n // Cascade through our extraction-specific opts in an ordered fashion,\n // turning them off as we try to extract content.\n let node = this.getContentNode($, title, url, opts);\n\n if (nodeIsSufficient(node)) {\n return this.cleanAndReturnNode(node, $);\n }\n\n // We didn't succeed on first pass, one by one disable our\n // extraction opts and try again.\n for (const key of Reflect.ownKeys(opts).filter(k => opts[k] === true)) {\n opts[key] = false;\n $ = cheerio.load(html);\n\n node = this.getContentNode($, title, url, opts);\n\n if (nodeIsSufficient(node)) {\n break;\n }\n }\n\n return this.cleanAndReturnNode(node, $);\n },\n\n // Get node given current options\n getContentNode($, title, url, opts) {\n return cleanContent(\n extractBestNode($, opts),\n {\n $,\n cleanConditionally: opts.cleanConditionally,\n title,\n url,\n });\n },\n\n // Once we got here, either we're at our last-resort node, or\n // we broke early. Make sure we at least have -something- before we\n // move forward.\n cleanAndReturnNode(node, $) {\n if (!node) {\n return null;\n }\n\n return normalizeSpaces($.html(node));\n\n // if return_type == \"html\":\n // return normalize_spaces(node_to_html(node))\n // else:\n // return node\n },\n\n};\n\nexport default GenericContentExtractor;\n","// TODO: It would be great if we could merge the meta and selector lists into\n// a list of objects, because we could then rank them better. For example,\n// .hentry .entry-title is far better suited than <meta title>.\n\n// An ordered list of meta tag names that denote likely article titles. All\n// attributes should be lowercase for faster case-insensitive matching. From\n// most distinct to least distinct.\nexport const STRONG_TITLE_META_TAGS = [\n 'tweetmeme-title',\n 'dc.title',\n 'rbtitle',\n 'headline',\n 'title',\n];\n\n// og:title is weak because it typically contains context that we don't like,\n// for example the source site's name. Gotta get that brand into facebook!\nexport const WEAK_TITLE_META_TAGS = [\n 'og:title',\n];\n\n// An ordered list of XPath Selectors to find likely article titles. From\n// most explicit to least explicit.\n//\n// Note - this does not use classes like CSS. This checks to see if the string\n// exists in the className, which is not as accurate as .className (which\n// splits on spaces/endlines), but for our purposes it's close enough. The\n// speed tradeoff is worth the accuracy hit.\nexport const STRONG_TITLE_SELECTORS = [\n '.hentry .entry-title',\n 'h1#articleHeader',\n 'h1.articleHeader',\n 'h1.article',\n '.instapaper_title',\n '#meebo-title',\n];\n\nexport const WEAK_TITLE_SELECTORS = [\n 'article h1',\n '#entry-title',\n '.entry-title',\n '#entryTitle',\n '#entrytitle',\n '.entryTitle',\n '.entrytitle',\n '#articleTitle',\n '.articleTitle',\n 'post post-title',\n 'h1.title',\n 'h2.article',\n 'h1',\n 'html head title',\n 'title',\n];\n","import { cleanTitle } from 'cleaners';\nimport {\n extractFromMeta,\n extractFromSelectors,\n} from 'utils/dom';\n\nimport {\n STRONG_TITLE_META_TAGS,\n WEAK_TITLE_META_TAGS,\n STRONG_TITLE_SELECTORS,\n WEAK_TITLE_SELECTORS,\n} from './constants';\n\nconst GenericTitleExtractor = {\n extract({ $, url, metaCache }) {\n // First, check to see if we have a matching meta tag that we can make\n // use of that is strongly associated with the headline.\n let title;\n\n title = extractFromMeta($, STRONG_TITLE_META_TAGS, metaCache);\n if (title) return cleanTitle(title, { url, $ });\n\n // Second, look through our content selectors for the most likely\n // article title that is strongly associated with the headline.\n title = extractFromSelectors($, STRONG_TITLE_SELECTORS);\n if (title) return cleanTitle(title, { url, $ });\n\n // Third, check for weaker meta tags that may match.\n title = extractFromMeta($, WEAK_TITLE_META_TAGS, metaCache);\n if (title) return cleanTitle(title, { url, $ });\n\n // Last, look for weaker selector tags that may match.\n title = extractFromSelectors($, WEAK_TITLE_SELECTORS);\n if (title) return cleanTitle(title, { url, $ });\n\n // If no matches, return an empty string\n return '';\n },\n};\n\nexport default GenericTitleExtractor;\n","// An ordered list of meta tag names that denote likely article authors. All\n// attributes should be lowercase for faster case-insensitive matching. From\n// most distinct to least distinct.\n//\n// Note: \"author\" is too often the -developer- of the page, so it is not\n// added here.\nexport const AUTHOR_META_TAGS = [\n 'byl',\n 'clmst',\n 'dc.author',\n 'dcsext.author',\n 'dc.creator',\n 'rbauthors',\n 'authors',\n];\n\nexport const AUTHOR_MAX_LENGTH = 300;\n\n// An ordered list of XPath Selectors to find likely article authors. From\n// most explicit to least explicit.\n//\n// Note - this does not use classes like CSS. This checks to see if the string\n// exists in the className, which is not as accurate as .className (which\n// splits on spaces/endlines), but for our purposes it's close enough. The\n// speed tradeoff is worth the accuracy hit.\nexport const AUTHOR_SELECTORS = [\n '.entry .entry-author',\n '.author.vcard .fn',\n '.author .vcard .fn',\n '.byline.vcard .fn',\n '.byline .vcard .fn',\n '.byline .by .author',\n '.byline .by',\n '.byline .author',\n '.post-author.vcard',\n '.post-author .vcard',\n 'a[rel=author]',\n '#by_author',\n '.by_author',\n '#entryAuthor',\n '.entryAuthor',\n '.byline a[href*=author]',\n '#author .authorname',\n '.author .authorname',\n '#author',\n '.author',\n '.articleauthor',\n '.ArticleAuthor',\n '.byline',\n];\n\n// An ordered list of Selectors to find likely article authors, with\n// regular expression for content.\nconst bylineRe = /^[\\n\\s]*By/i;\nexport const BYLINE_SELECTORS_RE = [\n ['#byline', bylineRe],\n ['.byline', bylineRe],\n];\n","import { cleanAuthor } from 'cleaners';\nimport {\n extractFromMeta,\n extractFromSelectors,\n} from 'utils/dom';\n\nimport {\n AUTHOR_META_TAGS,\n AUTHOR_MAX_LENGTH,\n AUTHOR_SELECTORS,\n BYLINE_SELECTORS_RE,\n} from './constants';\n\nconst GenericAuthorExtractor = {\n extract({ $, metaCache }) {\n let author;\n\n // First, check to see if we have a matching\n // meta tag that we can make use of.\n author = extractFromMeta($, AUTHOR_META_TAGS, metaCache);\n if (author && author.length < AUTHOR_MAX_LENGTH) {\n return cleanAuthor(author);\n }\n\n // Second, look through our selectors looking for potential authors.\n author = extractFromSelectors($, AUTHOR_SELECTORS, 2);\n if (author && author.length < AUTHOR_MAX_LENGTH) {\n return cleanAuthor(author);\n }\n\n // Last, use our looser regular-expression based selectors for\n // potential authors.\n for (const [selector, regex] of BYLINE_SELECTORS_RE) {\n const node = $(selector);\n if (node.length === 1) {\n const text = node.text();\n if (regex.test(text)) {\n return cleanAuthor(text);\n }\n }\n }\n\n return null;\n },\n};\n\nexport default GenericAuthorExtractor;\n\n","// An ordered list of meta tag names that denote\n// likely date published dates. All attributes\n// should be lowercase for faster case-insensitive matching.\n// From most distinct to least distinct.\nexport const DATE_PUBLISHED_META_TAGS = [\n 'article:published_time',\n 'displaydate',\n 'dc.date',\n 'dc.date.issued',\n 'rbpubdate',\n 'publish_date',\n 'pub_date',\n 'pagedate',\n 'pubdate',\n 'revision_date',\n 'doc_date',\n 'date_created',\n 'content_create_date',\n 'lastmodified',\n 'created',\n 'date',\n];\n\n// An ordered list of XPath Selectors to find\n// likely date published dates. From most explicit\n// to least explicit.\nexport const DATE_PUBLISHED_SELECTORS = [\n '.hentry .dtstamp.published',\n '.hentry .published',\n '.hentry .dtstamp.updated',\n '.hentry .updated',\n '.single .published',\n '.meta .published',\n '.meta .postDate',\n '.entry-date',\n '.byline .date',\n '.postmetadata .date',\n '.article_datetime',\n '.date-header',\n '.story-date',\n '.dateStamp',\n '#story .datetime',\n '.dateline',\n '.pubdate',\n];\n\n// An ordered list of compiled regular expressions to find likely date\n// published dates from the URL. These should always have the first\n// reference be a date string that is parseable by dateutil.parser.parse\nconst abbrevMonthsStr = '(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)';\nexport const DATE_PUBLISHED_URL_RES = [\n // /2012/01/27/ but not /2012/01/293\n new RegExp('/(20\\\\d{2}/\\\\d{2}/\\\\d{2})/', 'i'),\n // 20120127 or 20120127T but not 2012012733 or 8201201733\n // /[^0-9](20\\d{2}[01]\\d[0-3]\\d)([^0-9]|$)/i,\n // 2012-01-27\n new RegExp('(20\\\\d{2}-[01]\\\\d-[0-3]\\\\d)', 'i'),\n // /2012/jan/27/\n new RegExp(`/(20\\\\d{2}/${abbrevMonthsStr}/[0-3]\\\\d)/`, 'i'),\n];\n\n","import { cleanDatePublished } from 'cleaners';\nimport {\n extractFromMeta,\n extractFromSelectors,\n} from 'utils/dom';\nimport { extractFromUrl } from 'utils/text';\n\nimport {\n DATE_PUBLISHED_META_TAGS,\n DATE_PUBLISHED_SELECTORS,\n DATE_PUBLISHED_URL_RES,\n} from './constants';\n\nconst GenericDatePublishedExtractor = {\n extract({ $, url, metaCache }) {\n let datePublished;\n // First, check to see if we have a matching meta tag\n // that we can make use of.\n // Don't try cleaning tags from this string\n datePublished = extractFromMeta($, DATE_PUBLISHED_META_TAGS, metaCache, false);\n if (datePublished) return cleanDatePublished(datePublished);\n\n // Second, look through our selectors looking for potential\n // date_published's.\n datePublished = extractFromSelectors($, DATE_PUBLISHED_SELECTORS);\n if (datePublished) return cleanDatePublished(datePublished);\n\n // Lastly, look to see if a dately string exists in the URL\n datePublished = extractFromUrl(url, DATE_PUBLISHED_URL_RES);\n if (datePublished) return cleanDatePublished(datePublished);\n\n return null;\n },\n};\n\nexport default GenericDatePublishedExtractor;\n","// import {\n// DEK_META_TAGS,\n// DEK_SELECTORS,\n// DEK_URL_RES,\n// } from './constants';\n\n// import { cleanDek } from 'cleaners';\n\n// import {\n// extractFromMeta,\n// extractFromSelectors,\n// } from 'utils/dom';\n\n// Currently there is only one selector for\n// deks. We should simply return null here\n// until we have a more robust generic option.\n// Below is the original source for this, for reference.\nconst GenericDekExtractor = {\n // extract({ $, content, metaCache }) {\n extract() {\n return null;\n },\n};\n\nexport default GenericDekExtractor;\n\n// def extract_dek(self):\n// # First, check to see if we have a matching meta tag that we can make\n// # use of.\n// dek = self.extract_from_meta('dek', constants.DEK_META_TAGS)\n// if not dek:\n// # Second, look through our CSS/XPath selectors. This may return\n// # an HTML fragment.\n// dek = self.extract_from_selectors('dek',\n// constants.DEK_SELECTORS,\n// text_only=False)\n//\n// if dek:\n// # Make sure our dek isn't in the first few thousand characters\n// # of the content, otherwise it's just the start of the article\n// # and not a true dek.\n// content = self.extract_content()\n// content_chunk = normalize_spaces(strip_tags(content[:2000]))\n// dek_chunk = normalize_spaces(dek[:100]) # Already has no tags.\n//\n// # 80% or greater similarity means the dek was very similar to some\n// # of the starting content, so we skip it.\n// if fuzz.partial_ratio(content_chunk, dek_chunk) < 80:\n// return dek\n//\n// return None\n","// An ordered list of meta tag names that denote likely article leading images.\n// All attributes should be lowercase for faster case-insensitive matching.\n// From most distinct to least distinct.\nexport const LEAD_IMAGE_URL_META_TAGS = [\n 'og:image',\n 'twitter:image',\n 'image_src',\n];\n\nexport const LEAD_IMAGE_URL_SELECTORS = [\n 'link[rel=image_src]',\n];\n\nexport const POSITIVE_LEAD_IMAGE_URL_HINTS = [\n 'upload',\n 'wp-content',\n 'large',\n 'photo',\n 'wp-image',\n];\nexport const POSITIVE_LEAD_IMAGE_URL_HINTS_RE = new RegExp(POSITIVE_LEAD_IMAGE_URL_HINTS.join('|'), 'i');\n\nexport const NEGATIVE_LEAD_IMAGE_URL_HINTS = [\n 'spacer',\n 'sprite',\n 'blank',\n 'throbber',\n 'gradient',\n 'tile',\n 'bg',\n 'background',\n 'icon',\n 'social',\n 'header',\n 'hdr',\n 'advert',\n 'spinner',\n 'loader',\n 'loading',\n 'default',\n 'rating',\n 'share',\n 'facebook',\n 'twitter',\n 'theme',\n 'promo',\n 'ads',\n 'wp-includes',\n];\nexport const NEGATIVE_LEAD_IMAGE_URL_HINTS_RE = new RegExp(NEGATIVE_LEAD_IMAGE_URL_HINTS.join('|'), 'i');\n\nexport const GIF_RE = /\\.gif(\\?.*)?$/i;\nexport const JPG_RE = /\\.jpe?g(\\?.*)?$/i;\n","import {\n POSITIVE_LEAD_IMAGE_URL_HINTS_RE,\n NEGATIVE_LEAD_IMAGE_URL_HINTS_RE,\n GIF_RE,\n JPG_RE,\n} from './constants';\n\nimport { PHOTO_HINTS_RE } from '../content/scoring/constants';\n\nfunction getSig($node) {\n return `${$node.attr('class') || ''} ${$node.attr('id') || ''}`;\n}\n\n// Scores image urls based on a variety of heuristics.\nexport function scoreImageUrl(url) {\n url = url.trim();\n let score = 0;\n\n if (POSITIVE_LEAD_IMAGE_URL_HINTS_RE.test(url)) {\n score += 20;\n }\n\n if (NEGATIVE_LEAD_IMAGE_URL_HINTS_RE.test(url)) {\n score -= 20;\n }\n\n // TODO: We might want to consider removing this as\n // gifs are much more common/popular than they once were\n if (GIF_RE.test(url)) {\n score -= 10;\n }\n\n if (JPG_RE.test(url)) {\n score += 10;\n }\n\n // PNGs are neutral.\n\n return score;\n}\n\n// Alt attribute usually means non-presentational image.\nexport function scoreAttr($img) {\n if ($img.attr('alt')) {\n return 5;\n }\n\n return 0;\n}\n\n// Look through our parent and grandparent for figure-like\n// container elements, give a bonus if we find them\nexport function scoreByParents($img) {\n let score = 0;\n const $figParent = $img.parents('figure').first();\n\n if ($figParent.length === 1) {\n score += 25;\n }\n\n const $parent = $img.parent();\n let $gParent;\n if ($parent.length === 1) {\n $gParent = $parent.parent();\n }\n\n [$parent, $gParent].forEach(($node) => {\n if (PHOTO_HINTS_RE.test(getSig($node))) {\n score += 15;\n }\n });\n\n return score;\n}\n\n// Look at our immediate sibling and see if it looks like it's a\n// caption. Bonus if so.\nexport function scoreBySibling($img) {\n let score = 0;\n const $sibling = $img.next();\n const sibling = $sibling.get(0);\n\n if (sibling && sibling.tagName === 'figcaption') {\n score += 25;\n }\n\n if (PHOTO_HINTS_RE.test(getSig($sibling))) {\n score += 15;\n }\n\n return score;\n}\n\nexport function scoreByDimensions($img) {\n let score = 0;\n\n const width = parseFloat($img.attr('width'));\n const height = parseFloat($img.attr('height'));\n const src = $img.attr('src');\n\n // Penalty for skinny images\n if (width && width <= 50) {\n score -= 50;\n }\n\n // Penalty for short images\n if (height && height <= 50) {\n score -= 50;\n }\n\n if (width && height && !src.includes('sprite')) {\n const area = width * height;\n if (area < 5000) { // Smaller than 50 x 100\n score -= 100;\n } else {\n score += Math.round(area / 1000);\n }\n }\n\n return score;\n}\n\nexport function scoreByPosition($imgs, index) {\n return ($imgs.length / 2) - index;\n}\n","import { extractFromMeta } from 'utils/dom';\nimport { cleanImage } from 'cleaners';\n\nimport {\n LEAD_IMAGE_URL_META_TAGS,\n LEAD_IMAGE_URL_SELECTORS,\n} from './constants';\n\nimport {\n scoreImageUrl,\n scoreAttr,\n scoreByParents,\n scoreBySibling,\n scoreByDimensions,\n scoreByPosition,\n} from './score-image';\n\n// Given a resource, try to find the lead image URL from within\n// it. Like content and next page extraction, uses a scoring system\n// to determine what the most likely image may be. Short circuits\n// on really probable things like og:image meta tags.\n//\n// Potential signals to still take advantage of:\n// * domain\n// * weird aspect ratio\nconst GenericLeadImageUrlExtractor = {\n extract({ $, content, metaCache }) {\n let cleanUrl;\n\n // Check to see if we have a matching meta tag that we can make use of.\n // Moving this higher because common practice is now to use large\n // images on things like Open Graph or Twitter cards.\n // images usually have for things like Open Graph.\n const imageUrl =\n extractFromMeta(\n $,\n LEAD_IMAGE_URL_META_TAGS,\n metaCache,\n false\n );\n\n if (imageUrl) {\n cleanUrl = cleanImage(imageUrl);\n\n if (cleanUrl) return cleanUrl;\n }\n\n // Next, try to find the \"best\" image via the content.\n // We'd rather not have to fetch each image and check dimensions,\n // so try to do some analysis and determine them instead.\n const imgs = $('img', content).toArray();\n const imgScores = {};\n\n imgs.forEach((img, index) => {\n const $img = $(img);\n const src = $img.attr('src');\n\n if (!src) return;\n\n let score = scoreImageUrl(src);\n score += scoreAttr($img);\n score += scoreByParents($img);\n score += scoreBySibling($img);\n score += scoreByDimensions($img);\n score += scoreByPosition(imgs, index);\n\n imgScores[src] = score;\n });\n\n const [topUrl, topScore] =\n Reflect.ownKeys(imgScores).reduce((acc, key) =>\n imgScores[key] > acc[1] ? [key, imgScores[key]] : acc\n , [null, 0]);\n\n if (topScore > 0) {\n cleanUrl = cleanImage(topUrl);\n\n if (cleanUrl) return cleanUrl;\n }\n\n // If nothing else worked, check to see if there are any really\n // probable nodes in the doc, like <link rel=\"image_src\" />.\n for (const selector of LEAD_IMAGE_URL_SELECTORS) {\n const $node = $(selector).first();\n const src = $node.attr('src');\n if (src) {\n cleanUrl = cleanImage(src);\n if (cleanUrl) return cleanUrl;\n }\n\n const href = $node.attr('href');\n if (href) {\n cleanUrl = cleanImage(href);\n if (cleanUrl) return cleanUrl;\n }\n\n const value = $node.attr('value');\n if (value) {\n cleanUrl = cleanImage(value);\n if (cleanUrl) return cleanUrl;\n }\n }\n\n return null;\n },\n};\n\nexport default GenericLeadImageUrlExtractor;\n\n// def extract(self):\n// \"\"\"\n// # First, try to find the \"best\" image via the content.\n// # We'd rather not have to fetch each image and check dimensions,\n// # so try to do some analysis and determine them instead.\n// content = self.extractor.extract_content(return_type=\"node\")\n// imgs = content.xpath('.//img')\n// img_scores = defaultdict(int)\n// logger.debug('Scoring %d images from content', len(imgs))\n// for (i, img) in enumerate(imgs):\n// img_score = 0\n//\n// if not 'src' in img.attrib:\n// logger.debug('No src attribute found')\n// continue\n//\n// try:\n// parsed_img = urlparse(img.attrib['src'])\n// img_path = parsed_img.path.lower()\n// except ValueError:\n// logger.debug('ValueError getting img path.')\n// continue\n// logger.debug('Image path is %s', img_path)\n//\n// if constants.POSITIVE_LEAD_IMAGE_URL_HINTS_RE.match(img_path):\n// logger.debug('Positive URL hints match. Adding 20.')\n// img_score += 20\n//\n// if constants.NEGATIVE_LEAD_IMAGE_URL_HINTS_RE.match(img_path):\n// logger.debug('Negative URL hints match. Subtracting 20.')\n// img_score -= 20\n//\n// # Gifs are more often structure than photos\n// if img_path.endswith('gif'):\n// logger.debug('gif found. Subtracting 10.')\n// img_score -= 10\n//\n// # JPGs are more often photographs\n// if img_path.endswith('jpg'):\n// logger.debug('jpg found. Adding 10.')\n// img_score += 10\n//\n// # PNGs are neutral.\n//\n// # Alt attribute usually means non-presentational image.\n// if 'alt' in img.attrib and len(img.attrib['alt']) > 5:\n// logger.debug('alt attribute found. Adding 5.')\n// img_score += 5\n//\n// # Look through our parent and grandparent for figure-like\n// # container elements, give a bonus if we find them\n// parents = [img.getparent()]\n// if parents[0] is not None and parents[0].getparent() is not None:\n// parents.append(parents[0].getparent())\n// for p in parents:\n// if p.tag == 'figure':\n// logger.debug('Parent with <figure> tag found. Adding 25.')\n// img_score += 25\n//\n// p_sig = ' '.join([p.get('id', ''), p.get('class', '')])\n// if constants.PHOTO_HINTS_RE.search(p_sig):\n// logger.debug('Photo hints regex match. Adding 15.')\n// img_score += 15\n//\n// # Look at our immediate sibling and see if it looks like it's a\n// # caption. Bonus if so.\n// sibling = img.getnext()\n// if sibling is not None:\n// if sibling.tag == 'figcaption':\n// img_score += 25\n//\n// sib_sig = ' '.join([sibling.get('id', ''),\n// sibling.get('class', '')]).lower()\n// if 'caption' in sib_sig:\n// img_score += 15\n//\n// # Pull out width/height if they were set.\n// img_width = None\n// img_height = None\n// if 'width' in img.attrib:\n// try:\n// img_width = float(img.get('width'))\n// except ValueError:\n// pass\n// if 'height' in img.attrib:\n// try:\n// img_height = float(img.get('height'))\n// except ValueError:\n// pass\n//\n// # Penalty for skinny images\n// if img_width and img_width <= 50:\n// logger.debug('Skinny image found. Subtracting 50.')\n// img_score -= 50\n//\n// # Penalty for short images\n// if img_height and img_height <= 50:\n// # Wide, short images are more common than narrow, tall ones\n// logger.debug('Short image found. Subtracting 25.')\n// img_score -= 25\n//\n// if img_width and img_height and not 'sprite' in img_path:\n// area = img_width * img_height\n//\n// if area < 5000: # Smaller than 50x100\n// logger.debug('Image with small area found. Subtracting 100.')\n// img_score -= 100\n// else:\n// img_score += round(area/1000.0)\n//\n// # If the image is higher on the page than other images,\n// # it gets a bonus. Penalty if lower.\n// logger.debug('Adding page placement bonus of %d.', len(imgs)/2 - i)\n// img_score += len(imgs)/2 - i\n//\n// # Use the raw src here because we munged img_path for case\n// # insensitivity\n// logger.debug('Final score is %d.', img_score)\n// img_scores[img.attrib['src']] += img_score\n//\n// top_score = 0\n// top_url = None\n// for (url, score) in img_scores.items():\n// if score > top_score:\n// top_url = url\n// top_score = score\n//\n// if top_score > 0:\n// logger.debug('Using top score image from content. Score was %d', top_score)\n// return top_url\n//\n//\n// # If nothing else worked, check to see if there are any really\n// # probable nodes in the doc, like <link rel=\"image_src\" />.\n// logger.debug('Trying to find lead image in probable nodes')\n// for selector in constants.LEAD_IMAGE_URL_SELECTORS:\n// nodes = self.resource.extract_by_selector(selector)\n// for node in nodes:\n// clean_value = None\n// if node.attrib.get('src'):\n// clean_value = self.clean(node.attrib['src'])\n//\n// if not clean_value and node.attrib.get('href'):\n// clean_value = self.clean(node.attrib['href'])\n//\n// if not clean_value and node.attrib.get('value'):\n// clean_value = self.clean(node.attrib['value'])\n//\n// if clean_value:\n// logger.debug('Found lead image in probable nodes.')\n// logger.debug('Node was: %s', node)\n// return clean_value\n//\n// return None\n","import difflib from 'difflib';\n\nexport default function scoreSimilarity(score, articleUrl, href) {\n // Do this last and only if we have a real candidate, because it's\n // potentially expensive computationally. Compare the link to this\n // URL using difflib to get the % similarity of these URLs. On a\n // sliding scale, subtract points from this link based on\n // similarity.\n if (score > 0) {\n const similarity = new difflib.SequenceMatcher(null, articleUrl, href).ratio();\n // Subtract .1 from diff_percent when calculating modifier,\n // which means that if it's less than 10% different, we give a\n // bonus instead. Ex:\n // 3% different = +17.5 points\n // 10% different = 0 points\n // 20% different = -25 points\n const diffPercent = 1.0 - similarity;\n const diffModifier = -(250 * (diffPercent - 0.2));\n return score + diffModifier;\n }\n\n return 0;\n}\n","import { IS_DIGIT_RE } from 'utils/text/constants';\n\nexport default function scoreLinkText(linkText, pageNum) {\n // If the link text can be parsed as a number, give it a minor\n // bonus, with a slight bias towards lower numbered pages. This is\n // so that pages that might not have 'next' in their text can still\n // get scored, and sorted properly by score.\n let score = 0;\n\n if (IS_DIGIT_RE.test(linkText.trim())) {\n const linkTextAsNum = parseInt(linkText, 10);\n // If it's the first page, we already got it on the first call.\n // Give it a negative score. Otherwise, up to page 10, give a\n // small bonus.\n if (linkTextAsNum < 2) {\n score = -30;\n } else {\n score = Math.max(0, 10 - linkTextAsNum);\n }\n\n // If it appears that the current page number is greater than\n // this links page number, it's a very bad sign. Give it a big\n // penalty.\n if (pageNum && pageNum >= linkTextAsNum) {\n score -= 50;\n }\n }\n\n return score;\n}\n","export default function scorePageInLink(pageNum, isWp) {\n // page in the link = bonus. Intentionally ignore wordpress because\n // their ?p=123 link style gets caught by this even though it means\n // separate documents entirely.\n if (pageNum && !isWp) {\n return 50;\n }\n\n return 0;\n}\n","export const DIGIT_RE = /\\d/;\n\n// A list of words that, if found in link text or URLs, likely mean that\n// this link is not a next page link.\nexport const EXTRANEOUS_LINK_HINTS = [\n 'print',\n 'archive',\n 'comment',\n 'discuss',\n 'e-mail',\n 'email',\n 'share',\n 'reply',\n 'all',\n 'login',\n 'sign',\n 'single',\n 'adx',\n 'entry-unrelated',\n];\nexport const EXTRANEOUS_LINK_HINTS_RE = new RegExp(EXTRANEOUS_LINK_HINTS.join('|'), 'i');\n\n// Match any link text/classname/id that looks like it could mean the next\n// page. Things like: next, continue, >, >>, » but not >|, »| as those can\n// mean last page.\nexport const NEXT_LINK_TEXT_RE = new RegExp('(next|weiter|continue|>([^|]|$)|»([^|]|$))', 'i');\n\n// Match any link text/classname/id that looks like it is an end link: things\n// like \"first\", \"last\", \"end\", etc.\nexport const CAP_LINK_TEXT_RE = new RegExp('(first|last|end)', 'i');\n\n// Match any link text/classname/id that looks like it means the previous\n// page.\nexport const PREV_LINK_TEXT_RE = new RegExp('(prev|earl|old|new|<|«)', 'i');\n\n// Match any phrase that looks like it could be page, or paging, or pagination\nexport const PAGE_RE = new RegExp('pag(e|ing|inat)', 'i');\n\n","import { EXTRANEOUS_LINK_HINTS_RE } from '../constants';\n\nexport default function scoreExtraneousLinks(href) {\n // If the URL itself contains extraneous values, give a penalty.\n if (EXTRANEOUS_LINK_HINTS_RE.test(href)) {\n return -25;\n }\n\n return 0;\n}\n","import { range } from 'utils';\nimport {\n NEGATIVE_SCORE_RE,\n POSITIVE_SCORE_RE,\n PAGE_RE,\n} from 'utils/dom/constants';\nimport { EXTRANEOUS_LINK_HINTS_RE } from '../constants';\n\nfunction makeSig($link) {\n return `${$link.attr('class') || ''} ${$link.attr('id') || ''}`;\n}\n\nexport default function scoreByParents($link) {\n // If a parent node contains paging-like classname or id, give a\n // bonus. Additionally, if a parent_node contains bad content\n // (like 'sponsor'), give a penalty.\n let $parent = $link.parent();\n let positiveMatch = false;\n let negativeMatch = false;\n let score = 0;\n\n Array.from(range(0, 4)).forEach(() => {\n if ($parent.length === 0) {\n return;\n }\n\n const parentData = makeSig($parent, ' ');\n\n // If we have 'page' or 'paging' in our data, that's a good\n // sign. Add a bonus.\n if (!positiveMatch && PAGE_RE.test(parentData)) {\n positiveMatch = true;\n score += 25;\n }\n\n // If we have 'comment' or something in our data, and\n // we don't have something like 'content' as well, that's\n // a bad sign. Give a penalty.\n if (!negativeMatch && NEGATIVE_SCORE_RE.test(parentData)\n && EXTRANEOUS_LINK_HINTS_RE.test(parentData)) {\n if (!POSITIVE_SCORE_RE.test(parentData)) {\n negativeMatch = true;\n score -= 25;\n }\n }\n\n $parent = $parent.parent();\n });\n\n return score;\n}\n\n","import { PREV_LINK_TEXT_RE } from '../constants';\n\nexport default function scorePrevLink(linkData) {\n // If the link has something like \"previous\", its definitely\n // an old link, skip it.\n if (PREV_LINK_TEXT_RE.test(linkData)) {\n return -200;\n }\n\n return 0;\n}\n","import URL from 'url';\n\nimport {\n DIGIT_RE,\n EXTRANEOUS_LINK_HINTS_RE,\n} from '../constants';\n\nexport default function shouldScore(\n href,\n articleUrl,\n baseUrl,\n parsedUrl,\n linkText,\n previousUrls\n) {\n // skip if we've already fetched this url\n if (previousUrls.find(url => href === url) !== undefined) {\n return false;\n }\n\n // If we've already parsed this URL, or the URL matches the base\n // URL, or is empty, skip it.\n if (!href || href === articleUrl || href === baseUrl) {\n return false;\n }\n\n const { hostname } = parsedUrl;\n const { hostname: linkHost } = URL.parse(href);\n\n // Domain mismatch.\n if (linkHost !== hostname) {\n return false;\n }\n\n // If href doesn't contain a digit after removing the base URL,\n // it's certainly not the next page.\n const fragment = href.replace(baseUrl, '');\n if (!DIGIT_RE.test(fragment)) {\n return false;\n }\n\n // This link has extraneous content (like \"comment\") in its link\n // text, so we skip it.\n if (EXTRANEOUS_LINK_HINTS_RE.test(linkText)) {\n return false;\n }\n\n // Next page link text is never long, skip if it is too long.\n if (linkText.length > 25) {\n return false;\n }\n\n return true;\n}\n\n","export default function scoreBaseUrl(href, baseRegex) {\n // If the baseUrl isn't part of this URL, penalize this\n // link. It could still be the link, but the odds are lower.\n // Example:\n // http://www.actionscript.org/resources/articles/745/1/JavaScript-and-VBScript-Injection-in-ActionScript-3/Page1.html\n if (!baseRegex.test(href)) {\n return -25;\n }\n\n return 0;\n}\n","import { NEXT_LINK_TEXT_RE } from '../constants';\n\nexport default function scoreNextLinkText(linkData) {\n // Things like \"next\", \">>\", etc.\n if (NEXT_LINK_TEXT_RE.test(linkData)) {\n return 50;\n }\n\n return 0;\n}\n","import {\n NEXT_LINK_TEXT_RE,\n CAP_LINK_TEXT_RE,\n} from '../constants';\n\nexport default function scoreCapLinks(linkData) {\n // Cap links are links like \"last\", etc.\n if (CAP_LINK_TEXT_RE.test(linkData)) {\n // If we found a link like \"last\", but we've already seen that\n // this link is also \"next\", it's fine. If it's not been\n // previously marked as \"next\", then it's probably bad.\n // Penalize.\n if (NEXT_LINK_TEXT_RE.test(linkData)) {\n return -65;\n }\n }\n\n return 0;\n}\n","import URL from 'url';\n\nimport { isWordpress } from 'utils/dom';\nimport {\n removeAnchor,\n pageNumFromUrl,\n} from 'utils/text';\n\nimport {\n scoreSimilarity,\n scoreLinkText,\n scorePageInLink,\n scoreExtraneousLinks,\n scoreByParents,\n scorePrevLink,\n shouldScore,\n scoreBaseUrl,\n scoreCapLinks,\n scoreNextLinkText,\n} from './utils';\n\nexport function makeBaseRegex(baseUrl) {\n return new RegExp(`^${baseUrl}`, 'i');\n}\n\nfunction makeSig($link, linkText) {\n return `${linkText || $link.text()} ${$link.attr('class') || ''} ${$link.attr('id') || ''}`;\n}\n\nexport default function scoreLinks({\n links,\n articleUrl,\n baseUrl,\n parsedUrl,\n $,\n previousUrls = [],\n}) {\n parsedUrl = parsedUrl || URL.parse(articleUrl);\n const baseRegex = makeBaseRegex(baseUrl);\n const isWp = isWordpress($);\n\n // Loop through all links, looking for hints that they may be next-page\n // links. Things like having \"page\" in their textContent, className or\n // id, or being a child of a node with a page-y className or id.\n //\n // After we do that, assign each page a score, and pick the one that\n // looks most like the next page link, as long as its score is strong\n // enough to have decent confidence.\n const scoredPages = links.reduce((possiblePages, link) => {\n // Remove any anchor data since we don't do a good job\n // standardizing URLs (it's hard), we're going to do\n // some checking with and without a trailing slash\n const href = removeAnchor(link.attribs.href);\n const $link = $(link);\n const linkText = $link.text();\n\n if (!shouldScore(href, articleUrl, baseUrl, parsedUrl, linkText, previousUrls)) {\n return possiblePages;\n }\n\n // ## PASSED THE FIRST-PASS TESTS. Start scoring. ##\n if (!possiblePages[href]) {\n possiblePages[href] = {\n score: 0,\n linkText,\n href,\n };\n } else {\n possiblePages[href].linkText = `${possiblePages[href].linkText}|${linkText}`;\n }\n\n const possiblePage = possiblePages[href];\n const linkData = makeSig($link, linkText);\n const pageNum = pageNumFromUrl(href);\n\n let score = scoreBaseUrl(href, baseRegex);\n score += scoreNextLinkText(linkData);\n score += scoreCapLinks(linkData);\n score += scorePrevLink(linkData);\n score += scoreByParents($link);\n score += scoreExtraneousLinks(href);\n score += scorePageInLink(pageNum, isWp);\n score += scoreLinkText(linkText, pageNum);\n score += scoreSimilarity(score, articleUrl, href);\n\n possiblePage.score = score;\n\n return possiblePages;\n }, {});\n\n return Reflect.ownKeys(scoredPages).length === 0 ? null : scoredPages;\n}\n","import URL from 'url';\n\nimport {\n articleBaseUrl,\n removeAnchor,\n} from 'utils/text';\nimport scoreLinks from './scoring/score-links';\n\n// Looks for and returns next page url\n// for multi-page articles\nconst GenericNextPageUrlExtractor = {\n extract({ $, url, parsedUrl, previousUrls = [] }) {\n parsedUrl = parsedUrl || URL.parse(url);\n\n const articleUrl = removeAnchor(url);\n const baseUrl = articleBaseUrl(url, parsedUrl);\n\n const links = $('a[href]').toArray();\n\n const scoredLinks = scoreLinks({\n links,\n articleUrl,\n baseUrl,\n parsedUrl,\n $,\n previousUrls,\n });\n\n // If no links were scored, return null\n if (!scoredLinks) return null;\n\n // now that we've scored all possible pages,\n // find the biggest one.\n const topPage = Reflect.ownKeys(scoredLinks).reduce((acc, link) => {\n const scoredLink = scoredLinks[link];\n return scoredLink.score > acc.score ? scoredLink : acc;\n }, { score: -100 });\n\n // If the score is less than 50, we're not confident enough to use it,\n // so we fail.\n if (topPage.score >= 50) {\n return topPage.href;\n }\n\n return null;\n },\n};\n\n\nexport default GenericNextPageUrlExtractor;\n","export const CANONICAL_META_SELECTORS = [\n 'og:url',\n];\n","import URL from 'url';\nimport { extractFromMeta } from 'utils/dom';\n\nimport { CANONICAL_META_SELECTORS } from './constants';\n\nfunction parseDomain(url) {\n const parsedUrl = URL.parse(url);\n const { hostname } = parsedUrl;\n return hostname;\n}\n\nfunction result(url) {\n return {\n url,\n domain: parseDomain(url),\n };\n}\n\nconst GenericUrlExtractor = {\n extract({ $, url, metaCache }) {\n const $canonical = $('link[rel=canonical]');\n if ($canonical.length !== 0) {\n const href = $canonical.attr('href');\n if (href) {\n return result(href);\n }\n }\n\n const metaUrl = extractFromMeta($, CANONICAL_META_SELECTORS, metaCache);\n if (metaUrl) {\n return result(metaUrl);\n }\n\n return result(url);\n },\n\n};\n\nexport default GenericUrlExtractor;\n","export const EXCERPT_META_SELECTORS = [\n 'og:description',\n 'twitter:description',\n];\n","import ellipsize from 'ellipsize';\n\nimport {\n extractFromMeta,\n stripTags,\n} from 'utils/dom';\n\nimport { EXCERPT_META_SELECTORS } from './constants';\n\nexport function clean(content, $, maxLength = 200) {\n content = content.replace(/[\\s\\n]+/g, ' ').trim();\n return ellipsize(content, maxLength, { ellipse: '…' });\n}\n\nconst GenericExcerptExtractor = {\n extract({ $, content, metaCache }) {\n const excerpt = extractFromMeta($, EXCERPT_META_SELECTORS, metaCache);\n if (excerpt) {\n return clean(stripTags(excerpt, $));\n }\n // Fall back to excerpting from the extracted content\n const maxLength = 200;\n const shortContent = content.slice(0, maxLength * 5);\n return clean($(shortContent).text(), $, maxLength);\n },\n};\n\nexport default GenericExcerptExtractor;\n","import cheerio from 'cheerio';\n\nimport { normalizeSpaces } from 'utils/text';\n\nconst GenericWordCountExtractor = {\n extract({ content }) {\n const $ = cheerio.load(content);\n\n const text = normalizeSpaces($('div').first().text());\n return text.split(/\\s/).length;\n },\n};\n\nexport default GenericWordCountExtractor;\n","import cheerio from 'cheerio';\nimport stringDirection from 'string-direction';\n\nimport GenericContentExtractor from './content/extractor';\nimport GenericTitleExtractor from './title/extractor';\nimport GenericAuthorExtractor from './author/extractor';\nimport GenericDatePublishedExtractor from './date-published/extractor';\nimport GenericDekExtractor from './dek/extractor';\nimport GenericLeadImageUrlExtractor from './lead-image-url/extractor';\nimport GenericNextPageUrlExtractor from './next-page-url/extractor';\nimport GenericUrlExtractor from './url/extractor';\nimport GenericExcerptExtractor from './excerpt/extractor';\nimport GenericWordCountExtractor from './word-count/extractor';\n\nconst GenericExtractor = {\n // This extractor is the default for all domains\n domain: '*',\n title: GenericTitleExtractor.extract,\n date_published: GenericDatePublishedExtractor.extract,\n author: GenericAuthorExtractor.extract,\n content: GenericContentExtractor.extract.bind(GenericContentExtractor),\n lead_image_url: GenericLeadImageUrlExtractor.extract,\n dek: GenericDekExtractor.extract,\n next_page_url: GenericNextPageUrlExtractor.extract,\n url_and_domain: GenericUrlExtractor.extract,\n excerpt: GenericExcerptExtractor.extract,\n word_count: GenericWordCountExtractor.extract,\n direction: ({ title }) => stringDirection.getDirection(title),\n\n extract(options) {\n const { html } = options;\n\n if (html) {\n const $ = cheerio.load(html);\n options.$ = $;\n }\n\n const title = this.title(options);\n const date_published = this.date_published(options);\n const author = this.author(options);\n const content = this.content({ ...options, title });\n const lead_image_url = this.lead_image_url({ ...options, content });\n const dek = this.dek({ ...options, content });\n const next_page_url = this.next_page_url(options);\n const excerpt = this.excerpt({ ...options, content });\n const word_count = this.word_count({ ...options, content });\n const direction = this.direction({ title });\n const { url, domain } = this.url_and_domain(options);\n\n return {\n title,\n author,\n date_published: date_published || null,\n dek,\n lead_image_url,\n content,\n next_page_url,\n url,\n domain,\n excerpt,\n word_count,\n direction,\n };\n },\n};\n\nexport default GenericExtractor;\n","import URL from 'url';\n\nimport Extractors from './all';\nimport GenericExtractor from './generic';\n\nexport default function getExtractor(url, parsedUrl) {\n parsedUrl = parsedUrl || URL.parse(url);\n const { hostname } = parsedUrl;\n const baseDomain = hostname.split('.').slice(-2).join('.');\n\n return Extractors[hostname] || Extractors[baseDomain] || GenericExtractor;\n}\n","import Cleaners from 'cleaners';\nimport { convertNodeTo } from 'utils/dom';\nimport GenericExtractor from './generic';\n\n// Remove elements by an array of selectors\nexport function cleanBySelectors($content, $, { clean }) {\n if (!clean) return $content;\n\n $(clean.join(','), $content).remove();\n\n return $content;\n}\n\n// Transform matching elements\nexport function transformElements($content, $, { transforms }) {\n if (!transforms) return $content;\n\n Reflect.ownKeys(transforms).forEach((key) => {\n const $matches = $(key, $content);\n const value = transforms[key];\n\n // If value is a string, convert directly\n if (typeof value === 'string') {\n $matches.each((index, node) => {\n convertNodeTo($(node), $, transforms[key]);\n });\n } else if (typeof value === 'function') {\n // If value is function, apply function to node\n $matches.each((index, node) => {\n const result = value($(node), $);\n // If function returns a string, convert node to that value\n if (typeof result === 'string') {\n convertNodeTo($(node), $, result);\n }\n });\n }\n });\n\n return $content;\n}\n\nfunction findMatchingSelector($, selectors) {\n return selectors.find((selector) => {\n if (Array.isArray(selector)) {\n const [s, attr] = selector;\n return $(s).length === 1 && $(s).attr(attr) && $(s).attr(attr).trim() !== '';\n }\n\n return $(selector).length === 1 && $(selector).text().trim() !== '';\n });\n}\n\nexport function select(opts) {\n const { $, type, extractionOpts, extractHtml = false } = opts;\n // Skip if there's not extraction for this type\n if (!extractionOpts) return null;\n\n // If a string is hardcoded for a type (e.g., Wikipedia\n // contributors), return the string\n if (typeof extractionOpts === 'string') return extractionOpts;\n\n const { selectors, defaultCleaner = true } = extractionOpts;\n\n const matchingSelector = findMatchingSelector($, selectors);\n\n if (!matchingSelector) return null;\n\n // Declaring result; will contain either\n // text or html, which will be cleaned\n // by the appropriate cleaner type\n\n // If the selector type requests html as its return type\n // transform and clean the element with provided selectors\n if (extractHtml) {\n let $content = $(matchingSelector);\n\n // Wrap in div so transformation can take place on root element\n $content.wrap($('<div></div>'));\n $content = $content.parent();\n\n $content = transformElements($content, $, extractionOpts);\n $content = cleanBySelectors($content, $, extractionOpts);\n\n $content = Cleaners[type]($content, { ...opts, defaultCleaner });\n\n return $.html($content);\n }\n\n let result;\n\n // if selector is an array (e.g., ['img', 'src']),\n // extract the attr\n if (Array.isArray(matchingSelector)) {\n const [selector, attr] = matchingSelector;\n result = $(selector).attr(attr).trim();\n } else {\n result = $(matchingSelector).text().trim();\n }\n\n // Allow custom extractor to skip default cleaner\n // for this type; defaults to true\n if (defaultCleaner) {\n return Cleaners[type](result, opts);\n }\n\n return result;\n}\n\nfunction extractResult(opts) {\n const { type, extractor, fallback = true } = opts;\n\n const result = select({ ...opts, extractionOpts: extractor[type] });\n\n // If custom parser succeeds, return the result\n if (result) {\n return result;\n }\n\n // If nothing matches the selector, and fallback is enabled,\n // run the Generic extraction\n if (fallback) return GenericExtractor[type](opts);\n\n return null;\n}\n\nconst RootExtractor = {\n extract(extractor = GenericExtractor, opts) {\n const { contentOnly, extractedTitle } = opts;\n // This is the generic extractor. Run its extract method\n if (extractor.domain === '*') return extractor.extract(opts);\n\n opts = {\n ...opts,\n extractor,\n };\n\n if (contentOnly) {\n const content = extractResult({\n ...opts, type: 'content', extractHtml: true, title: extractedTitle,\n });\n return {\n content,\n };\n }\n const title = extractResult({ ...opts, type: 'title' });\n const date_published = extractResult({ ...opts, type: 'date_published' });\n const author = extractResult({ ...opts, type: 'author' });\n const next_page_url = extractResult({ ...opts, type: 'next_page_url' });\n const content = extractResult({\n ...opts, type: 'content', extractHtml: true, title,\n });\n const lead_image_url = extractResult({ ...opts, type: 'lead_image_url', content });\n const excerpt = extractResult({ ...opts, type: 'excerpt', content });\n const dek = extractResult({ ...opts, type: 'dek', content, excerpt });\n const word_count = extractResult({ ...opts, type: 'word_count', content });\n const direction = extractResult({ ...opts, type: 'direction', title });\n const { url, domain } =\n extractResult({ ...opts, type: 'url_and_domain' }) || { url: null, domain: null };\n\n return {\n title,\n content,\n author,\n date_published,\n lead_image_url,\n dek,\n next_page_url,\n url,\n domain,\n excerpt,\n word_count,\n direction,\n };\n },\n};\n\nexport default RootExtractor;\n","import { removeAnchor } from 'utils/text';\nimport RootExtractor from 'extractors/root-extractor';\nimport GenericExtractor from 'extractors/generic';\nimport Resource from 'resource';\n\nexport default async function collectAllPages(\n {\n next_page_url,\n html,\n $,\n metaCache,\n result,\n Extractor,\n title,\n url,\n }\n) {\n // At this point, we've fetched just the first page\n let pages = 1;\n const previousUrls = [removeAnchor(url)];\n\n // If we've gone over 26 pages, something has\n // likely gone wrong.\n while (next_page_url && pages < 26) {\n pages += 1;\n $ = await Resource.create(next_page_url);\n html = $.html();\n\n const extractorOpts = {\n url: next_page_url,\n html,\n $,\n metaCache,\n contentOnly: true,\n extractedTitle: title,\n previousUrls,\n };\n\n const nextPageResult = RootExtractor.extract(Extractor, extractorOpts);\n\n previousUrls.push(next_page_url);\n result = {\n ...result,\n content: `\n ${result.content}\n <hr>\n <h4>Page ${pages}</h4>\n ${nextPageResult.content}\n `,\n };\n\n next_page_url = nextPageResult.next_page_url;\n }\n\n const word_count = GenericExtractor.word_count({ content: `<div>${result.content}</div>` });\n return {\n ...result,\n total_pages: pages,\n pages_rendered: pages,\n word_count,\n };\n}\n","import URL from 'url';\n\nimport Resource from 'resource';\nimport {\n validateUrl,\n Errors,\n} from 'utils';\nimport getExtractor from 'extractors/get-extractor';\nimport RootExtractor from 'extractors/root-extractor';\nimport collectAllPages from 'extractors/collect-all-pages';\n\nconst Mercury = {\n async parse(url, html, opts = {}) {\n const {\n fetchAllPages = true,\n fallback = true,\n } = opts;\n\n const parsedUrl = URL.parse(url);\n\n if (!validateUrl(parsedUrl)) {\n return Errors.badUrl;\n }\n\n const Extractor = getExtractor(url, parsedUrl);\n // console.log(`Using extractor for ${Extractor.domain}`);\n\n const $ = await Resource.create(url, html, parsedUrl);\n\n // If we found an error creating the resource, return that error\n if ($.error) {\n return $;\n }\n\n html = $.html();\n\n // Cached value of every meta name in our document.\n // Used when extracting title/author/date_published/dek\n const metaCache = $('meta').map((_, node) => $(node).attr('name')).toArray();\n\n let result = RootExtractor.extract(Extractor, { url, html, $, metaCache, parsedUrl, fallback });\n const { title, next_page_url } = result;\n\n // Fetch more pages if next_page_url found\n if (fetchAllPages && next_page_url) {\n result = await collectAllPages(\n {\n Extractor,\n next_page_url,\n html,\n $,\n metaCache,\n result,\n title,\n url,\n }\n );\n } else {\n result = {\n ...result,\n total_pages: 1,\n rendered_pages: 1,\n };\n }\n\n return result;\n },\n\n // A convenience method for getting a resource\n // to work with, e.g., for custom extractor generator\n async fetchResource(url) {\n return await Resource.create(url);\n },\n\n};\n\nexport default Mercury;\n"],"names":["range","start","end","validateUrl","hostname","Errors","REQUEST_HEADERS","FETCH_TIMEOUT","BAD_CONTENT_TYPES","BAD_CONTENT_TYPES_RE","RegExp","join","MAX_CONTENT_LENGTH","get","options","resolve","reject","err","response","body","validateResponse","parseNon2xx","statusMessage","statusCode","Error","error","headers","contentType","contentLength","test","url","parsedUrl","URL","parse","encodeURI","badUrl","fetchResource","convertMetaProp","$","from","to","each","_","node","$node","value","attr","removeAttr","normalizeMetaTags","IS_LINK","IS_IMAGE","TAGS_TO_REMOVE","convertLazyLoadedImages","img","attribs","forEach","isComment","index","type","cleanComments","root","find","contents","filter","remove","clean","Resource","preparedResponse","validResponse","result","generateDoc","content","includes","cheerio","load","normalizeWhitespace","children","length","NYMagExtractor","$children","tagName","BloggerExtractor","WikipediaExtractor","$parent","parents","prepend","TwitterExtractor","tweets","$tweetContainer","append","replaceWith","NYTimesExtractor","src","width","replace","TheAtlanticExtractor","NewYorkerExtractor","WiredExtractor","MSNExtractor","YahooExtractor","BuzzfeedExtractor","WikiaExtractor","LittleThingsExtractor","PoliticoExtractor","DeadspinExtractor","youtubeId","split","BroadwayWorldExtractor","ApartmentTherapyExtractor","data","JSON","sources","$img","Extractors","SPACER_RE","KEEP_CLASS","KEEP_SELECTORS","STRIP_OUTPUT_TAGS","REMOVE_ATTRS","REMOVE_ATTR_SELECTORS","map","selector","REMOVE_ATTR_LIST","WHITELIST_ATTRS","WHITELIST_ATTRS_RE","REMOVE_EMPTY_TAGS","REMOVE_EMPTY_SELECTORS","tag","CLEAN_CONDITIONALLY_TAGS","HEADER_TAGS","HEADER_TAG_LIST","UNLIKELY_CANDIDATES_BLACKLIST","UNLIKELY_CANDIDATES_WHITELIST","DIV_TO_P_BLOCK_TAGS","NON_TOP_CANDIDATE_TAGS","NON_TOP_CANDIDATE_TAGS_RE","PHOTO_HINTS","PHOTO_HINTS_RE","POSITIVE_SCORE_HINTS","POSITIVE_SCORE_RE","NEGATIVE_SCORE_HINTS","NEGATIVE_SCORE_RE","IS_WP_SELECTOR","EXTRANEOUS_LINK_HINTS","EXTRANEOUS_LINK_HINTS_RE","PAGE_RE","BLOCK_LEVEL_TAGS","BLOCK_LEVEL_TAGS_RE","candidatesBlacklist","CANDIDATES_BLACKLIST","candidatesWhitelist","CANDIDATES_WHITELIST","stripUnlikelyCandidates","not","classes","id","classAndId","brsToPs","collapsing","element","nextElement","next","paragraphize","br","sibling","nextSibling","p","appendTo","convertDivs","div","$div","convertable","convertSpans","span","$span","convertToParagraphs","convertNodeTo","attribString","key","cleanForHeight","height","parseInt","removeSpacers","cleanImages","$article","markToKeep","article","tags","addClass","stripJunkTags","removeClass","cleanHOnes","$hOnes","removeAllButWhitelist","reduce","acc","cleanAttributes","removeEmpty","$p","text","trim","HNEWS_CONTENT_SELECTORS","READABILITY_ASSET","PARAGRAPH_SCORE_TAGS","CHILD_CONTENT_TAGS","BAD_TAGS","getWeight","score","getScore","parseFloat","scoreCommas","match","idkRe","scoreLength","textLength","chunks","lengthBonus","Math","min","max","scoreParagraph","slice","setScore","addScore","amount","getOrInitScore","e","addToParent","parent","weightNodes","scoreNode","addScoreTo","scorePs","rawScore","scoreContent","parentSelector","childSelector","NORMALIZE_RE","normalizeSpaces","extractFromUrl","regexList","matchRe","re","exec","PAGE_IN_HREF_RE","HAS_ALPHA_RE","IS_ALPHA_RE","IS_DIGIT_RE","pageNumFromUrl","matches","pageNum","removeAnchor","isGoodSegment","segment","firstSegmentHasLetters","goodSegment","toLowerCase","articleBaseUrl","parsed","protocol","host","path","cleanedSegments","reverse","rawSegment","possibleSegment","fileExt","push","SENTENCE_END_RE","hasSentenceEnd","excerptContent","words","mergeSiblings","$candidate","topScore","siblingScoreThreshold","wrappingDiv","$sibling","siblingScore","contentBonus","density","linkDensity","newScore","siblingContent","siblingContentLength","findTopCandidate","first","removeUnlessContent","weight","hasClass","pCount","inputCount","imgCount","nodeIsList","previousNode","prev","scriptCount","cleanTags","cleanHeaders","title","header","$header","prevAll","rewriteTopLevel","absolutize","rootUrl","$content","absoluteUrl","makeLinksAbsolute","totalTextLength","linkText","linkLength","extractFromMeta","metaNames","cachedNames","foundNames","indexOf","name","nodes","values","toArray","metaValue","stripTags","isGoodNode","maxChildren","withinComment","extractFromSelectors","selectors","textOnly","html","cleanText","commentParent","class","undefined","nodeIsSufficient","isWordpress","CLEAN_AUTHOR_RE","TEXT_LINK_RE","MS_DATE_STRING","SEC_DATE_STRING","CLEAN_DATE_STRING_RE","TIME_MERIDIAN_SPACE_RE","TIME_MERIDIAN_DOTS_RE","months","allMonths","timestamp1","timestamp2","SPLIT_DATE_STRING","TITLE_SPLITTERS_RE","DOMAIN_ENDINGS_RE","cleanAuthor","author","leadImageUrl","validUrl","isWebUri","cleanDek","dek","excerpt","dekText","cleanDateString","dateString","cleanDatePublished","date","moment","Date","isValid","toISOString","extractCleanNode","cleanConditionally","defaultCleaner","cleanTitle","resolveSplitTitle","h1","extractBreadcrumbTitle","splitTitle","termCounts","titleText","maxTerm","termCount","splitEnds","longestEnd","cleanDomainFromTitle","nakedDomain","startSlug","startSlugRatio","wuzzy","levenshtein","endSlug","endSlugRatio","newTitle","Cleaners","cleanImage","cleanContent","extractBestNode","opts","$topCandidate","GenericContentExtractor","defaultOpts","getContentNode","cleanAndReturnNode","k","STRONG_TITLE_META_TAGS","WEAK_TITLE_META_TAGS","STRONG_TITLE_SELECTORS","WEAK_TITLE_SELECTORS","GenericTitleExtractor","metaCache","AUTHOR_META_TAGS","AUTHOR_MAX_LENGTH","AUTHOR_SELECTORS","bylineRe","BYLINE_SELECTORS_RE","GenericAuthorExtractor","regex","DATE_PUBLISHED_META_TAGS","DATE_PUBLISHED_SELECTORS","abbrevMonthsStr","DATE_PUBLISHED_URL_RES","GenericDatePublishedExtractor","datePublished","GenericDekExtractor","LEAD_IMAGE_URL_META_TAGS","LEAD_IMAGE_URL_SELECTORS","POSITIVE_LEAD_IMAGE_URL_HINTS","POSITIVE_LEAD_IMAGE_URL_HINTS_RE","NEGATIVE_LEAD_IMAGE_URL_HINTS","NEGATIVE_LEAD_IMAGE_URL_HINTS_RE","GIF_RE","JPG_RE","getSig","scoreImageUrl","scoreAttr","scoreByParents","$figParent","$gParent","scoreBySibling","scoreByDimensions","area","round","scoreByPosition","$imgs","GenericLeadImageUrlExtractor","cleanUrl","imageUrl","imgs","imgScores","topUrl","href","scoreSimilarity","articleUrl","similarity","difflib","SequenceMatcher","ratio","diffPercent","diffModifier","scoreLinkText","linkTextAsNum","scorePageInLink","isWp","DIGIT_RE","NEXT_LINK_TEXT_RE","CAP_LINK_TEXT_RE","PREV_LINK_TEXT_RE","scoreExtraneousLinks","makeSig","$link","positiveMatch","negativeMatch","parentData","scorePrevLink","linkData","shouldScore","baseUrl","previousUrls","linkHost","fragment","scoreBaseUrl","baseRegex","scoreNextLinkText","scoreCapLinks","makeBaseRegex","scoreLinks","links","scoredPages","possiblePages","link","possiblePage","GenericNextPageUrlExtractor","scoredLinks","topPage","scoredLink","CANONICAL_META_SELECTORS","parseDomain","GenericUrlExtractor","$canonical","metaUrl","EXCERPT_META_SELECTORS","maxLength","ellipsize","ellipse","GenericExcerptExtractor","shortContent","GenericWordCountExtractor","GenericExtractor","extract","bind","stringDirection","getDirection","date_published","lead_image_url","next_page_url","word_count","direction","url_and_domain","domain","getExtractor","baseDomain","cleanBySelectors","transformElements","transforms","$matches","findMatchingSelector","Array","isArray","s","select","extractionOpts","extractHtml","matchingSelector","wrap","extractResult","extractor","fallback","RootExtractor","contentOnly","extractedTitle","Extractor","pages","create","extractorOpts","nextPageResult","collectAllPages","Mercury","fetchAllPages"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;eAAyBA;;AAAzB,AAAe,SAAUA,KAAV;MAAgBC,KAAhB,uEAAwB,CAAxB;MAA2BC,GAA3B,uEAAiC,CAAjC;;;;;gBACND,SAASC,GADH;;;;;;iBAELD,SAAS,CAFJ;;;;;;;;;;;;;;ACAf;AACA,AAAe,SAASE,WAAT,OAAmC;MAAZC,QAAY,QAAZA,QAAY;;;SAEzC,CAAC,CAACA,QAAT;;;ACHF,IAAMC,SAAS;UACL;WACC,IADD;cAEI;;CAHd,CAOA;;ACPO,IAAMC,kBAAkB;gBACf;CADT;;;AAKP,AAAO,IAAMC,gBAAgB,KAAtB;;;AAGP,IAAMC,oBAAoB,CACxB,YADwB,EAExB,WAFwB,EAGxB,YAHwB,EAIxB,WAJwB,CAA1B;;AAOA,AAAO,IAAMC,uBAAuB,IAAIC,MAAJ,QAAgBF,kBAAkBG,IAAlB,CAAuB,GAAvB,CAAhB,SAAiD,GAAjD,CAA7B;;;;AAKP,AAAO,IAAMC,qBAAqB,OAA3B,CAEP,AAIA,AAKA;;ACpBA,SAASC,GAAT,CAAaC,OAAb,EAAsB;SACb,aAAY,UAACC,OAAD,EAAUC,MAAV,EAAqB;YAC9BF,OAAR,EAAiB,UAACG,GAAD,EAAMC,QAAN,EAAgBC,IAAhB,EAAyB;UACpCF,GAAJ,EAAS;eACAA,GAAP;OADF,MAEO;gBACG,EAAEE,UAAF,EAAQD,kBAAR,EAAR;;KAJJ;GADK,CAAP;;;;;;;;AAgBF,AAAO,SAASE,gBAAT,CAA0BF,QAA1B,EAAyD;MAArBG,WAAqB,uEAAP,KAAO;;;MAE1DH,SAASI,aAAT,KAA2B,IAA/B,EAAqC;QAC/B,CAACJ,SAASK,UAAd,EAA0B;YAClB,IAAIC,KAAJ,sDAC+CN,SAASO,KADxD,CAAN;KADF,MAIO,IAAI,CAACJ,WAAL,EAAkB;YACjB,IAAIG,KAAJ,kDAC2CN,SAASK,UADpD,wEAAN;;;;0BASAL,SAASQ,OAjBiD;MAe5CC,WAf4C,qBAe5D,cAf4D;MAgB1CC,aAhB0C,qBAgB5D,gBAhB4D;;;;MAoB1DnB,qBAAqBoB,IAArB,CAA0BF,WAA1B,CAAJ,EAA4C;UACpC,IAAIH,KAAJ,yCACkCG,WADlC,0BAAN;;;;MAMEC,gBAAgBhB,kBAApB,EAAwC;UAChC,IAAIY,KAAJ,yEACkEZ,kBADlE,OAAN;;;SAKK,IAAP;;;AAGF,AAMA;;;;;;AAMA;yDAAe,iBAA6BkB,GAA7B,EAAkCC,SAAlC;;;;;;;wBACDA,aAAaC,IAAIC,KAAJ,CAAUC,UAAUJ,GAAV,CAAV,CAAzB;;mBADa,GAGG;mBACTC,SADS;oCAEAzB,eAAd,CAFc;uBAGLC,aAHK;;;wBAMJ,IANI;;mBAQT,IARS;;oBAUR,IAVQ;;kCAYM;aAfT;;mBAkBoBM,IAAIC,OAAJ,CAlBpB;;;;oBAAA,SAkBLI,QAlBK;gBAAA,SAkBKC,IAlBL;;;6BAqBMD,QAAjB;6CACO,EAAEC,UAAF,EAAQD,kBAAR,EAtBI;;;;;6CAwBJb,OAAO8B,MAxBH;;;;;;;;GAAf;;WAA8BC,aAA9B;;;;SAA8BA,aAA9B;;;AC5EA,SAASC,eAAT,CAAyBC,CAAzB,EAA4BC,IAA5B,EAAkCC,EAAlC,EAAsC;cAC1BD,IAAV,QAAmBE,IAAnB,CAAwB,UAACC,CAAD,EAAIC,IAAJ,EAAa;QAC7BC,QAAQN,EAAEK,IAAF,CAAd;;QAEME,QAAQD,MAAME,IAAN,CAAWP,IAAX,CAAd;UACMO,IAAN,CAAWN,EAAX,EAAeK,KAAf;UACME,UAAN,CAAiBR,IAAjB;GALF;;SAQOD,CAAP;;;;;;;;;;AAUF,AAAe,SAASU,iBAAT,CAA2BV,CAA3B,EAA8B;MACvCD,gBAAgBC,CAAhB,EAAmB,SAAnB,EAA8B,OAA9B,CAAJ;MACID,gBAAgBC,CAAhB,EAAmB,UAAnB,EAA+B,MAA/B,CAAJ;SACOA,CAAP;;;ACtBK,IAAMW,UAAU,IAAIvC,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAhB;AACP,AAAO,IAAMwC,WAAW,IAAIxC,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB;;AAEP,AAAO,IAAMyC,iBAAiB,CAC5B,QAD4B,EAE5B,OAF4B,EAG5B,MAH4B,EAI5BxC,IAJ4B,CAIvB,GAJuB,CAAvB;;ACEP;;;;;AAKA,AAAe,SAASyC,uBAAT,CAAiCd,CAAjC,EAAoC;IAC/C,KAAF,EAASG,IAAT,CAAc,UAACC,CAAD,EAAIW,GAAJ,EAAY;qBACRA,IAAIC,OAApB,EAA6BC,OAA7B,CAAqC,UAACT,IAAD,EAAU;UACvCD,QAAQQ,IAAIC,OAAJ,CAAYR,IAAZ,CAAd;;UAEIA,SAAS,KAAT,IAAkBG,QAAQpB,IAAR,CAAagB,KAAb,CAAlB,IACAK,SAASrB,IAAT,CAAcgB,KAAd,CADJ,EAC0B;UACtBQ,GAAF,EAAOP,IAAP,CAAY,KAAZ,EAAmBD,KAAnB;;KALJ;GADF;;SAWOP,CAAP;;;ACpBF,SAASkB,SAAT,CAAmBC,KAAnB,EAA0Bd,IAA1B,EAAgC;SACvBA,KAAKe,IAAL,KAAc,SAArB;;;AAGF,SAASC,aAAT,CAAuBrB,CAAvB,EAA0B;IACtBsB,IAAF,GAASC,IAAT,CAAc,GAAd,EACSC,QADT,GAESC,MAFT,CAEgBP,SAFhB,EAGSQ,MAHT;;SAKO1B,CAAP;;;AAGF,AAAe,SAAS2B,KAAT,CAAe3B,CAAf,EAAkB;IAC7Ba,cAAF,EAAkBa,MAAlB;;MAEIL,cAAcrB,CAAd,CAAJ;SACOA,CAAP;;;ACVF,IAAM4B,WAAW;;;;;;;;QAAA,kBAQFpC,GARE,EAQGqC,gBARH,EAQqBpC,SARrB,EAQgC;;;;;;;;;oBAAA;;mBAGzCoC,gBAHyC;;;;;2BAAA,GAIrB;+BACL,IADK;4BAER,GAFQ;yBAGX;kCACS,WADT;oCAEW;;eATqB;;;uBAalC,EAAEhD,MAAMgD,gBAAR,EAA0BjD,UAAUkD,aAApC,EAAT;;;;;;qBAEehC,cAAcN,GAAd,EAAmBC,SAAnB,CAf4B;;;oBAAA;;;mBAkBzCsC,OAAO5C,KAlBkC;;;;;+CAmBpC4C,MAnBoC;;;+CAsBtC,MAAKC,WAAL,CAAiBD,MAAjB,CAtBsC;;;;;;;;;GARhC;aAAA,6BAiC0B;QAArBE,OAAqB,QAA3BpD,IAA2B;QAAZD,QAAY,QAAZA,QAAY;QACfS,WADe,GACCT,SAASQ,OADV,CAC/B,cAD+B;;;;;QAKnC,CAACC,YAAY6C,QAAZ,CAAqB,MAArB,CAAD,IACA,CAAC7C,YAAY6C,QAAZ,CAAqB,MAArB,CADL,EACmC;YAC3B,IAAIhD,KAAJ,CAAU,qCAAV,CAAN;;;QAGEc,IAAImC,QAAQC,IAAR,CAAaH,OAAb,EAAsB,EAAEI,qBAAqB,IAAvB,EAAtB,CAAR;;QAEIrC,EAAEsB,IAAF,GAASgB,QAAT,GAAoBC,MAApB,KAA+B,CAAnC,EAAsC;YAC9B,IAAIrD,KAAJ,CAAU,kCAAV,CAAN;;;QAGEwB,kBAAkBV,CAAlB,CAAJ;QACIc,wBAAwBd,CAAxB,CAAJ;QACI2B,MAAM3B,CAAN,CAAJ;;WAEOA,CAAP;;CArDJ,CAyDA;;AClEO,IAAMwC,iBAAiB;UACpB,WADoB;WAEnB;;eAEI,CACT,qBADS,EAET,cAFS,EAGT,iBAHS,CAFJ;;;WASA,CACL,KADK,EAEL,uBAFK,CATA;;;;;;;;gBAoBK;;UAEN,IAFM;;;gBAKA,kBAAClC,KAAD,EAAW;YACbmC,YAAYnC,MAAMgC,QAAN,EAAlB;YACIG,UAAUF,MAAV,KAAqB,CAArB,IAA0BE,UAAUlE,GAAV,CAAc,CAAd,EAAiBmE,OAAjB,KAA6B,KAA3D,EAAkE;iBACzD,QAAP;;;eAGK,IAAP;;;GAjCsB;;SAsCrB;eACM,CACT,uBADS,EAET,qBAFS,EAGT,IAHS;GAvCe;;UA8CpB;eACK,CACT,aADS,EAET,sBAFS;GA/Ce;;OAqDvB;eACQ,CACT,sBADS;GAtDe;;kBA2DZ;eACH,CACT,CAAC,kCAAD,EAAqC,UAArC,CADS,EAET,wBAFS;;CA5DR;;ACAA,IAAMC,mBAAmB;UACtB,cADsB;WAErB;;;;eAII,CACT,wBADS,CAJJ;;;WASA,EATA;;;gBAaK;gBACA;;GAhBgB;;UAoBtB;eACK,CACT,mBADS;GArBiB;;SA0BvB;eACM,CACT,gBADS;GA3BiB;;kBAgCd;eACH,CACT,kBADS;;CAjCR;;ACAA,IAAMC,qBAAqB;UACxB,eADwB;WAEvB;eACI,CACT,kBADS,CADJ;;oBAKS,KALT;;;gBAQK;sBACM,oBAACtC,KAAD,EAAW;YACnBuC,UAAUvC,MAAMwC,OAAN,CAAc,UAAd,CAAhB;;YAEID,QAAQP,QAAR,CAAiB,KAAjB,EAAwBC,MAAxB,KAAmC,CAAvC,EAA0C;kBAChCQ,OAAR,CAAgBzC,KAAhB;;OALM;0BAQU,YARV;kBASE;KAjBP;;;WAqBA,CACL,iBADK,EAEL,oCAFK,EAGL,MAHK,EAIL,SAJK;;GAvBuB;;UAgCxB,wBAhCwB;;SAkCzB;eACM,CACT,UADS;GAnCmB;;kBAwChB;eACH,CACT,sBADS;;;CAzCR;;ACAA,IAAM0C,mBAAmB;UACtB,aADsB;;WAGrB;gBACK;;;;;+BAKe,2BAAC1C,KAAD,EAAQN,CAAR,EAAc;YAC/BiD,SAAS3C,MAAMiB,IAAN,CAAW,QAAX,CAAf;YACM2B,kBAAkBlD,EAAE,iCAAF,CAAxB;wBACgBmD,MAAhB,CAAuBF,MAAvB;cACMG,WAAN,CAAkBF,eAAlB;OATQ;;;;SAcP;KAfE;;eAkBI,CACT,uBADS,CAlBJ;;oBAsBS,KAtBT;;WAwBA,CACL,qBADK,EAEL,QAFK,EAGL,sBAHK;GA3BqB;;UAkCtB;eACK,CACT,kCADS;GAnCiB;;kBAwCd;eACH,CACT,CAAC,4CAAD,EAA+C,cAA/C,CADS;;;CAzCR;;ACAA,IAAMG,mBAAmB;SACvB;eACM,CACT,aADS,EAET,aAFS;GAFiB;;UAQtB;eACK,CACT,CAAC,qBAAD,EAAwB,OAAxB,CADS,EAET,WAFS,EAGT,SAHS;GATiB;;WAgBrB;eACI,CACT,cADS,EAET,eAFS,CADJ;;oBAMS,KANT;;gBAQK;oBACI,kBAAC/C,KAAD,EAAW;YACnBgD,MAAMhD,MAAME,IAAN,CAAW,KAAX,CAAV;;;;;;;;;;YAUM+C,QAAQ,GAAd;;cAEMD,IAAIE,OAAJ,CAAY,UAAZ,EAAwBD,KAAxB,CAAN;cACM/C,IAAN,CAAW,KAAX,EAAkB8C,GAAlB;;KAvBG;;WA2BA,CACL,KADK,EAEL,qBAFK,EAGL,2BAHK,EAIL,kBAJK,EAKL,mBALK,EAML,QANK,EAOL,kBAPK,EAQL,SARK;GA3CqB;;kBAuDd,IAvDc;;kBAyDd,IAzDc;;OA2DzB,IA3DyB;;iBA6Df,IA7De;;WA+DrB;CA/DJ;;ACAP;;AAEA,AAAO,IAAMG,uBAAuB;UAC1B,qBAD0B;SAE3B;eACM,CACT,QADS;GAHqB;;UAQ1B;eACK,CACT,0DADS;GATqB;;WAczB;eACI,CACT,eADS,CADJ;;;;gBAOK,EAPL;;;;;WAaA;GA3ByB;;kBAgClB,IAhCkB;;kBAkClB,IAlCkB;;OAoC7B,IApC6B;;iBAsCnB,IAtCmB;;WAwCzB;CAxCJ;;ACFP;;;AAGA,AAAO,IAAMC,qBAAqB;UACxB,mBADwB;SAEzB;eACM,CACT,UADS;GAHmB;;UAQxB;eACK,CACT,eADS;GATmB;;WAcvB;eACI,CACT,iBADS,EAET,iBAFS,CADJ;;;;gBAQK,EARL;;;;;WAcA;GA5BuB;;kBAiChB;eACH,CACT,CAAC,qCAAD,EAAwC,OAAxC,CADS;GAlCmB;;kBAuChB;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GAxCmB;;OA6C3B;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GA9CmB;;iBAmDjB,IAnDiB;;WAqDvB;CArDJ;;ACHP;;;AAGA,AAAO,IAAMC,iBAAiB;UACpB,eADoB;SAErB;eACM,CACT,eADS;GAHe;;UASpB;eACK,CACT,iBADS;GAVe;;WAgBnB;eACI,CACT,iBADS,CADJ;;;;gBAQK,EARL;;;;;WAcA,CACL,kBADK;GA9BmB;;kBAoCZ;eACH,CACT,CAAC,gCAAD,EAAmC,OAAnC,CADS;GArCe;;kBA0CZ;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA3Ce;;OAgDvB;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GAjDe;;iBAsDb,IAtDa;;WAwDnB;CAxDJ;;ACHP;;;AAGA,AAAO,IAAMC,eAAe;UAClB,aADkB;SAEnB;eACM,CACT,IADS;GAHa;;UASlB;eACK,CACT,qBADS;GAVa;;WAgBjB;eACI,CACT,cADS,CADJ;;;;gBAQK,EARL;;;;;WAcA,CACL,cADK;GA9BiB;;kBAoCV;eACH,CACT,WADS;GArCa;;kBA0CV;eACH;GA3Ca;;OAgDrB;eACQ,CACT,CAAC,0BAAD,EAA6B,OAA7B,CADS;GAjDa;;iBAsDX,IAtDW;;WAwDjB;CAxDJ;;ACHP;;;AAGA,AAAO,IAAMC,iBAAiB;UACpB,eADoB;SAErB;eACM,CACT,sBADS;GAHe;;UASpB;eACK,CACT,oBADS;GAVe;;WAgBnB;eACI;;qBAAA,CADJ;;;;gBAQK,EARL;;;;;WAcA,CACL,iBADK;GA9BmB;;kBAoCZ;eACH,CACT,CAAC,qBAAD,EAAwB,UAAxB,CADS;GArCe;;kBA0CZ;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA3Ce;;OAgDvB;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GAjDe;;iBAuDb,IAvDa;;WAyDnB;CAzDJ;;ACHP;;;AAGA,AAAO,IAAMC,oBAAoB;UACvB,kBADuB;SAExB;eACM,CACT,qBADS;GAHkB;;UASvB;eACK,CACT,gCADS,EACyB,gBADzB;GAVkB;;WAgBtB;eACI,CACT,gBADS,CADJ;;oBAMS,KANT;;;;gBAUK;UACN;KAXC;;;;;WAiBA,CACL,oBADK,EAEL,uEAFK,EAGL,YAHK;GAjCsB;;kBAwCf;eACH,CACT,gBADS;GAzCkB;;kBA+Cf;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GAhDkB;;OAqD1B;eACQ,CACT,CAAC,0BAAD,EAA6B,OAA7B,CADS;GAtDkB;;iBA2DhB,IA3DgB;;WA6DtB;CA7DJ;;ACHP;;;AAGA,AAAO,IAAMC,iBAAiB;UACpB,kBADoB;SAErB;eACM,CACT,gBADS;GAHe;;UASpB;eACK,CACT,eADS,EACQ,KADR;GAVe;;WAgBnB;eACI,CACT,eADS,EAET,gBAFS,CADJ;;;;gBASK,EATL;;;;;WAeA;GA/BmB;;kBAoCZ;eACH,CACT,CAAC,qCAAD,EAAwC,OAAxC,CADS;GArCe;;kBA0CZ;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA3Ce;;OAgDvB;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GAjDe;;iBAsDb,IAtDa;;WAwDnB;CAxDJ;;ACHP;;;AAGA,AAAO,IAAMC,wBAAwB;UAC3B,sBAD2B;SAE5B;eACM,CACT,eADS;GAHsB;;UAS3B;eACK,CACT,CAAC,qBAAD,EAAwB,OAAxB,CADS;GAVsB;;WAgB1B;eACI;;uBAAA,EAGT,kBAHS,CADJ;;;;gBASK,EATL;;;;;WAeA;GA/B0B;;kBAoCnB;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GArCsB;;iBA0CpB,IA1CoB;;WA4C1B;CA5CJ;;ACHP;;;AAGA,AAAO,IAAMC,oBAAoB;UACvB,kBADuB;SAExB;eACM;;KAER,uBAAD,EAA0B,OAA1B,CAFS;GAHkB;;UASvB;eACK,CACT,oCADS;GAVkB;;WAetB;eACI;;yBAAA,EAGT,gBAHS,EAGS,aAHT,EAIT,aAJS,CADJ;;;;gBAUK,EAVL;;;;;WAgBA,CACL,YADK;GA/BsB;;kBAoCf;eACH,CACT,CAAC,+CAAD,EAAkD,UAAlD,CADS;GArCkB;;kBA2Cf;eACH;;KAER,uBAAD,EAA0B,OAA1B,CAFS;GA5CkB;;OAmD1B;eACQ,CACT,CAAC,0BAAD,EAA6B,OAA7B,CADS;GApDkB;;iBAyDhB,IAzDgB;;WA2DtB;CA3DJ;;ACHA,IAAMC,oBAAoB;UACvB,cADuB;SAExB;eACM,CACT,aADS;GAHkB;;UAQvB;eACK,CACT,SADS;GATkB;;WActB;eACI,CACT,eADS,EAET,gBAFS,CADJ;;;;gBAQK;0DAC0C,8CAAC5D,KAAD,EAAW;YACvD6D,YAAY7D,MAAME,IAAN,CAAW,IAAX,EAAiB4D,KAAjB,CAAuB,UAAvB,EAAmC,CAAnC,CAAlB;cACM5D,IAAN,CAAW,KAAX,qCAAmD2D,SAAnD;;KAXG;;;;;WAkBA;GAhCsB;;kBAoCf;eACH,CACT,CAAC,wBAAD,EAA2B,UAA3B,CADS;GArCkB;;kBA0Cf;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA3CkB;;OAgD1B;eACQ;;;GAjDkB;;iBAsDhB;eACF;;;GAvDkB;;WA4DtB;eACI;;;;CA7DR;;ACAP;;;AAGA,AAAO,IAAME,yBAAyB;UAC5B,uBAD4B;SAE7B;eACM,CACT,kBADS;GAHuB;;UAQ5B;eACK,CACT,uBADS;GATuB;;WAc3B;eACI,CACT,2BADS,CADJ;;;;gBAOK,EAPL;;;;;WAaA;GA3B2B;;kBAgCpB;eACH,CACT,CAAC,8BAAD,EAAiC,OAAjC,CADS;GAjCuB;;kBAsCpB;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GAvCuB;;OA4C/B;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GA7CuB;;iBAkDrB;eACF;;;GAnDuB;;WAwD3B;eACI;;;;CAzDR;;ACHP;;;AAGA,AAAO,IAAMC,4BAA4B;UAC/B,0BAD+B;SAEhC;eACM,CACT,aADS;GAH0B;;UAQ/B;eACK,CACT,mBADS;GAT0B;;WAc9B;eACI,CACT,mBADS,CADJ;;;;gBAOK;wDACwC,+CAAChE,KAAD,EAAQN,CAAR,EAAc;YACxDuE,OAAOC,KAAK7E,KAAL,CAAWW,MAAME,IAAN,CAAW,YAAX,CAAX,CAAb;YACQ8C,GAFsD,GAE9CiB,KAAKE,OAAL,CAAa,CAAb,CAF8C,CAEtDnB,GAFsD;;YAGxDoB,OAAO1E,EAAE,SAAF,EAAaQ,IAAb,CAAkB,KAAlB,EAAyB8C,GAAzB,CAAb;cACMF,WAAN,CAAkBsB,IAAlB;;KAZG;;;;;WAmBA;GAjC8B;;kBAsCvB;eACH,CACT,CAAC,kCAAD,EAAqC,UAArC,CADS;GAvC0B;;kBA4CvB;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA7C0B;;OAkDlC;eACQ,CACT,CAAC,wBAAD,EAA2B,OAA3B,CADS;GAnD0B;;iBAwDxB;eACF;;;GAzD0B;;WA8D9B;eACI;;;;CA/DR;;ACeP,IAAMC,aAAa;eACJnC,cADI;kBAEDG,gBAFC;mBAGAC,kBAHA;iBAIFI,gBAJE;qBAKEK,gBALF;yBAMMI,oBANN;uBAOIC,kBAPJ;mBAQAC,cARA;iBASFC,YATE;mBAUAC,cAVA;sBAWGC,iBAXH;sBAYGC,cAZH;0BAaOC,qBAbP;sBAcGC,iBAdH;kBAeDC,iBAfC;2BAgBQG,sBAhBR;8BAiBWC;CAjB9B,CAoBA;;ACtCA;AACA,AAAO,IAAMM,YAAY,IAAIxG,MAAJ,CAAW,gCAAX,EAA6C,GAA7C,CAAlB;;;;AAIP,AAAO,IAAMyG,aAAa,qBAAnB;;AAEP,AAAO,IAAMC,iBAAiB,CAC5B,wCAD4B,EAE5B,uCAF4B,EAG5B,qCAH4B,EAI5B,oCAJ4B,CAAvB;;;AAQP,AAAO,IAAMC,oBAAoB,CAC/B,OAD+B,EAE/B,QAF+B,EAG/B,UAH+B,EAI/B,MAJ+B,EAK/B,OAL+B,EAM/B,IAN+B,EAO/B,OAP+B,EAQ/B,QAR+B,EAS/B,QAT+B,CAA1B;;;AAaP,AAAO,IAAMC,eAAe,CAAC,OAAD,EAAU,OAAV,CAArB;AACP,AAAO,IAAMC,wBAAwBD,aAAaE,GAAb,CAAiB;eAAgBC,QAAhB;CAAjB,CAA9B;AACP,AAAO,IAAMC,mBAAmBJ,aAAa3G,IAAb,CAAkB,GAAlB,CAAzB;AACP,AAAO,IAAMgH,kBAAkB,CAAC,KAAD,EAAQ,QAAR,EAAkB,MAAlB,EAA0B,OAA1B,EAAmC,IAAnC,EAAyC,KAAzC,CAAxB;AACP,AAAO,IAAMC,qBAAqB,IAAIlH,MAAJ,QAAgBiH,gBAAgBhH,IAAhB,CAAqB,GAArB,CAAhB,SAA+C,GAA/C,CAA3B;;;AAGP,AAAO,IAAMkH,oBAAoB,CAAC,GAAD,CAA1B;AACP,AAAO,IAAMC,yBAAyBD,kBAAkBL,GAAlB,CAAsB;SAAUO,GAAV;CAAtB,EAA6CpH,IAA7C,CAAkD,GAAlD,CAA/B;;;AAGP,AAAO,IAAMqH,2BAA2B,CAAC,IAAD,EAAO,IAAP,EAAa,OAAb,EAAsB,KAAtB,EAA6B,QAA7B,EAAuC,MAAvC,EAA+CrH,IAA/C,CAAoD,GAApD,CAAjC;;;AAGP,IAAMsH,cAAc,CAAC,IAAD,EAAO,IAAP,EAAa,IAAb,EAAmB,IAAnB,EAAyB,IAAzB,CAApB;AACA,AAAO,IAAMC,kBAAkBD,YAAYtH,IAAZ,CAAiB,GAAjB,CAAxB;;;;;;;;AASP,AAAO,IAAMwH,gCAAgC,CAC3C,UAD2C,EAE3C,OAF2C,EAG3C,QAH2C,EAI3C,SAJ2C,EAK3C,SAL2C,EAM3C,KAN2C,EAO3C,gBAP2C,EAQ3C,OAR2C,EAS3C,SAT2C,EAU3C,cAV2C,EAW3C,QAX2C,EAY3C,iBAZ2C,EAa3C,OAb2C,EAc3C,MAd2C;;AAgB3C,QAhB2C,EAiB3C,QAjB2C,EAkB3C,QAlB2C,EAmB3C,OAnB2C;AAoB3C,MApB2C,EAqB3C,MArB2C,EAsB3C,KAtB2C,EAuB3C,UAvB2C,EAwB3C,OAxB2C,EAyB3C,YAzB2C,EA0B3C,UA1B2C;AA2B3C,2BA3B2C;AA4B3C,OA5B2C,EA6B3C,eA7B2C,EA8B3C,SA9B2C,EA+B3C,QA/B2C,EAgC3C,QAhC2C,EAiC3C,KAjC2C,EAkC3C,OAlC2C,EAmC3C,UAnC2C,EAoC3C,SApC2C,EAqC3C,UArC2C,EAsC3C,SAtC2C,EAuC3C,SAvC2C,EAwC3C,OAxC2C,CAAtC;;;;;;;;;;;;;AAsDP,AAAO,IAAMC,gCAAgC,CAC3C,KAD2C,EAE3C,SAF2C,EAG3C,MAH2C,EAI3C,WAJ2C,EAK3C,QAL2C,EAM3C,SAN2C,EAO3C,qBAP2C,EAQ3C,QAR2C;AAS3C,OAT2C,EAU3C,QAV2C,EAW3C,OAX2C,EAY3C,MAZ2C,EAa3C,MAb2C,EAc3C,OAd2C,EAe3C,QAf2C,CAAtC;;;;;AAqBP,AAAO,IAAMC,sBAAsB,CACjC,GADiC,EAEjC,YAFiC,EAGjC,IAHiC,EAIjC,KAJiC,EAKjC,KALiC,EAMjC,GANiC,EAOjC,KAPiC,EAQjC,OARiC,EASjC1H,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM2H,yBAAyB,CACpC,IADoC,EAEpC,GAFoC,EAGpC,GAHoC,EAIpC,OAJoC,EAKpC,IALoC,EAMpC,MANoC,EAOpC,MAPoC,EAQpC,UARoC,EASpC,OAToC,EAUpC,KAVoC,EAWpC,MAXoC,EAYpC,MAZoC,CAA/B;;AAeP,AAAO,IAAMC,4BACX,IAAI7H,MAAJ,QAAgB4H,uBAAuB3H,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;AAGP,AAYA,AAAO,IAAM6H,cAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,iBAAiB,IAAI/H,MAAJ,CAAW8H,YAAY7H,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAM+H,uBAAuB,CAClC,SADkC,EAElC,gBAFkC,EAGlC,iBAHkC,EAIlC,MAJkC,EAKlC,MALkC,EAMlC,SANkC,EAOlC,qBAPkC,EAQlC,OARkC,EASlC,QATkC,EAUlC,MAVkC,EAWlC,QAXkC,EAYlC,MAZkC,EAalC,YAbkC,EAclC,WAdkC,EAelC,MAfkC,EAgBlC,OAhBkC,EAiBlC,MAjBkC,EAkBlC,UAlBkC;AAmBlC,SAnBkC,CAA7B;;;AAuBP,AAAO,IAAMC,oBAAoB,IAAIjI,MAAJ,CAAWgI,qBAAqB/H,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA;;;;AAIA,AAAO,IAAMiI,uBAAuB,CAClC,OADkC,EAElC,QAFkC,EAGlC,QAHkC,EAIlC,KAJkC,EAKlC,UALkC,EAMlC,QANkC,EAOlC,QAPkC,EAQlC,OARkC,EASlC,MATkC,EAUlC,OAVkC,EAWlC,SAXkC,EAYlC,YAZkC,EAalC,SAbkC,EAclC,MAdkC,EAelC,QAfkC,EAgBlC,OAhBkC,EAiBlC,MAjBkC,EAkBlC,MAlBkC,EAmBlC,SAnBkC,EAoBlC,UApBkC;AAqBlC,MArBkC,EAsBlC,QAtBkC,EAuBlC,UAvBkC,EAwBlC,MAxBkC,EAyBlC,MAzBkC,EA0BlC,MA1BkC,EA2BlC,UA3BkC;AA4BlC,mBA5BkC,EA6BlC,MA7BkC,EA8BlC,WA9BkC,EA+BlC,MA/BkC,EAgClC,UAhCkC,EAiClC,OAjCkC,EAkClC,MAlCkC,EAmClC,OAnCkC,EAoClC,UApCkC;AAqClC,OArCkC,EAsClC,KAtCkC;AAuClC,SAvCkC,EAwClC,SAxCkC,EAyClC,cAzCkC;AA0ClC,QA1CkC,EA2ClC,WA3CkC,EA4ClC,OA5CkC,EA6ClC,UA7CkC,EA8ClC,UA9CkC,EA+ClC,MA/CkC,EAgDlC,SAhDkC,EAiDlC,SAjDkC,EAkDlC,OAlDkC,EAmDlC,KAnDkC,EAoDlC,SApDkC,EAqDlC,MArDkC,EAsDlC,OAtDkC,EAuDlC,QAvDkC,CAA7B;;AA0DP,AAAO,IAAMC,oBAAoB,IAAInI,MAAJ,CAAWkI,qBAAqBjI,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMmI,iBAAiB,wCAAvB;;AAEP,AAGA;;AAEA,AAAO,IAAMC,wBAAwB,CACnC,OADmC,EAEnC,SAFmC,EAGnC,SAHmC,EAInC,SAJmC,EAKnC,QALmC,EAMnC,OANmC,EAOnC,OAPmC,EAQnC,OARmC,EASnC,KATmC,EAUnC,OAVmC,EAWnC,MAXmC,EAYnC,QAZmC,EAanC,KAbmC,EAcnC,iBAdmC,CAA9B;AAgBP,AAAO,IAAMC,2BAA2B,IAAItI,MAAJ,CAAWqI,sBAAsBpI,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;AAGP,AAAO,IAAMsI,UAAU,IAAIvI,MAAJ,CAAW,iBAAX,EAA8B,GAA9B,CAAhB;;AAEP,AAMA,AAIA,AAIA,AAGA,AAGA;;AAEA,AAAO,IAAMwI,mBAAmB,CAC9B,SAD8B,EAE9B,OAF8B,EAG9B,YAH8B,EAI9B,MAJ8B,EAK9B,IAL8B,EAM9B,QAN8B,EAO9B,QAP8B,EAQ9B,SAR8B,EAS9B,KAT8B,EAU9B,UAV8B,EAW9B,IAX8B,EAY9B,KAZ8B,EAa9B,IAb8B,EAc9B,IAd8B,EAe9B,OAf8B,EAgB9B,UAhB8B,EAiB9B,YAjB8B,EAkB9B,QAlB8B,EAmB9B,QAnB8B,EAoB9B,MApB8B,EAqB9B,IArB8B,EAsB9B,IAtB8B,EAuB9B,IAvB8B,EAwB9B,IAxB8B,EAyB9B,IAzB8B,EA0B9B,IA1B8B,EA2B9B,QA3B8B,EA4B9B,QA5B8B,EA6B9B,IA7B8B,EA8B9B,IA9B8B,EA+B9B,KA/B8B,EAgC9B,QAhC8B,EAiC9B,IAjC8B,EAkC9B,QAlC8B,EAmC9B,GAnC8B,EAoC9B,KApC8B,EAqC9B,UArC8B,EAsC9B,SAtC8B,EAuC9B,OAvC8B,EAwC9B,OAxC8B,EAyC9B,UAzC8B,EA0C9B,OA1C8B,EA2C9B,IA3C8B,EA4C9B,OA5C8B,EA6C9B,IA7C8B,EA8C9B,IA9C8B,EA+C9B,OA/C8B,CAAzB;AAiDP,AAAO,IAAMC,sBAAsB,IAAIzI,MAAJ,QAAgBwI,iBAAiBvI,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAMyI,sBAAsBjB,8BAA8BxH,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM0I,uBAAuB,IAAI3I,MAAJ,CAAW0I,mBAAX,EAAgC,GAAhC,CAA7B;;AAEP,IAAME,sBAAsBlB,8BAA8BzH,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM4I,uBAAuB,IAAI7I,MAAJ,CAAW4I,mBAAX,EAAgC,GAAhC,CAA7B,CAEP,AAGA,AACA,AACA,AAEA;;ACtYe,SAASE,uBAAT,CAAiClH,CAAjC,EAAoC;;;;;;;;;;IAU/C,GAAF,EAAOmH,GAAP,CAAW,GAAX,EAAgBhH,IAAhB,CAAqB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;QAC9BC,QAAQN,EAAEK,IAAF,CAAd;QACM+G,UAAU9G,MAAME,IAAN,CAAW,OAAX,CAAhB;QACM6G,KAAK/G,MAAME,IAAN,CAAW,IAAX,CAAX;QACI,CAAC6G,EAAD,IAAO,CAACD,OAAZ,EAAqB;;QAEfE,cAAgBF,WAAW,EAA3B,WAAiCC,MAAM,EAAvC,CAAN;QACIJ,qBAAqB1H,IAArB,CAA0B+H,UAA1B,CAAJ,EAA2C;;KAA3C,MAEO,IAAIP,qBAAqBxH,IAArB,CAA0B+H,UAA1B,CAAJ,EAA2C;YAC1C5F,MAAN;;GAVJ;;SAcO1B,CAAP;;;AC3BF;;;;;;;;;;AAUA,AAAe,SAASuH,OAAT,CAAiBvH,CAAjB,EAAoB;MAC7BwH,aAAa,KAAjB;IACE,IAAF,EAAQrH,IAAR,CAAa,UAACgB,KAAD,EAAQsG,OAAR,EAAoB;QACzBC,cAAc1H,EAAEyH,OAAF,EAAWE,IAAX,GAAkBpJ,GAAlB,CAAsB,CAAtB,CAApB;;QAEImJ,eAAeA,YAAYhF,OAAZ,KAAwB,IAA3C,EAAiD;mBAClC,IAAb;QACE+E,OAAF,EAAW/F,MAAX;KAFF,MAGO,IAAI8F,UAAJ,EAAgB;mBACR,KAAb;;mBAEaC,OAAb,EAAsBzH,CAAtB,EAAyB,IAAzB;;GATJ;;SAaOA,CAAP;;;ACzBF;;;;;;;;;;;AAWA,AAAe,SAAS4H,YAAT,CAAsBvH,IAAtB,EAA4BL,CAA5B,EAA2C;MAAZ6H,EAAY,uEAAP,KAAO;;MAClDvH,QAAQN,EAAEK,IAAF,CAAd;;MAEIwH,EAAJ,EAAQ;QACFC,UAAUzH,KAAK0H,WAAnB;QACMC,IAAIhI,EAAE,SAAF,CAAV;;;;WAIO8H,WAAW,EAAEA,QAAQpF,OAAR,IAAmBmE,oBAAoBtH,IAApB,CAAyBuI,QAAQpF,OAAjC,CAArB,CAAlB,EAAmF;UAC3EqF,cAAcD,QAAQC,WAA5B;QACED,OAAF,EAAWG,QAAX,CAAoBD,CAApB;gBACUD,WAAV;;;UAGI3E,WAAN,CAAkB4E,CAAlB;UACMtG,MAAN;WACO1B,CAAP;;;SAGKA,CAAP;;;AC7BF,SAASkI,WAAT,CAAqBlI,CAArB,EAAwB;IACpB,KAAF,EAASG,IAAT,CAAc,UAACgB,KAAD,EAAQgH,GAAR,EAAgB;QACtBC,OAAOpI,EAAEmI,GAAF,CAAb;QACME,cAAcD,KAAK9F,QAAL,CAAcyD,mBAAd,EAAmCxD,MAAnC,KAA8C,CAAlE;;QAEI8F,WAAJ,EAAiB;oBACDD,IAAd,EAAoBpI,CAApB,EAAuB,GAAvB;;GALJ;;SASOA,CAAP;;;AAGF,SAASsI,YAAT,CAAsBtI,CAAtB,EAAyB;IACrB,MAAF,EAAUG,IAAV,CAAe,UAACgB,KAAD,EAAQoH,IAAR,EAAiB;QACxBC,QAAQxI,EAAEuI,IAAF,CAAd;QACMF,cAAcG,MAAM1F,OAAN,CAAc,QAAd,EAAwBP,MAAxB,KAAmC,CAAvD;QACI8F,WAAJ,EAAiB;oBACDG,KAAd,EAAqBxI,CAArB,EAAwB,GAAxB;;GAJJ;;SAQOA,CAAP;;;;;;;;;;;;;;;AAeF,AAAe,SAASyI,mBAAT,CAA6BzI,CAA7B,EAAgC;MACzCuH,QAAQvH,CAAR,CAAJ;MACIkI,YAAYlI,CAAZ,CAAJ;MACIsI,aAAatI,CAAb,CAAJ;;SAEOA,CAAP;;;AC9Ca,SAAS0I,aAAT,CAAuBpI,KAAvB,EAA8BN,CAA9B,EAA4C;MAAXyF,GAAW,uEAAL,GAAK;;MACnDpF,OAAOC,MAAM/B,GAAN,CAAU,CAAV,CAAb;MACI,CAAC8B,IAAL,EAAW;WACFL,CAAP;;;mBAEkBM,MAAM/B,GAAN,CAAU,CAAV,CALqC;;MAKjDyC,OALiD,cAKjDA,OALiD;;MAMnD2H,eAAe,iBAAgB3H,OAAhB,EACQkE,GADR,CACY;WAAU0D,GAAV,SAAiB5H,QAAQ4H,GAAR,CAAjB;GADZ,EAEQvK,IAFR,CAEa,GAFb,CAArB;;QAIM+E,WAAN,OAAsBqC,GAAtB,SAA6BkD,YAA7B,SAA6CrI,MAAMkB,QAAN,EAA7C,UAAkEiE,GAAlE;SACOzF,CAAP;;;ACTF,SAAS6I,cAAT,CAAwBnE,IAAxB,EAA8B1E,CAA9B,EAAiC;MACzB8I,SAASC,SAASrE,KAAKlE,IAAL,CAAU,QAAV,CAAT,EAA8B,EAA9B,CAAf;MACM+C,QAAQwF,SAASrE,KAAKlE,IAAL,CAAU,OAAV,CAAT,EAA6B,EAA7B,KAAoC,EAAlD;;;;;MAKI,CAACsI,UAAU,EAAX,IAAiB,EAAjB,IAAuBvF,QAAQ,EAAnC,EAAuC;SAChC7B,MAAL;GADF,MAEO,IAAIoH,MAAJ,EAAY;;;;SAIZrI,UAAL,CAAgB,QAAhB;;;SAGKT,CAAP;;;;;AAKF,SAASgJ,aAAT,CAAuBtE,IAAvB,EAA6B1E,CAA7B,EAAgC;MAC1B4E,UAAUrF,IAAV,CAAemF,KAAKlE,IAAL,CAAU,KAAV,CAAf,CAAJ,EAAsC;SAC/BkB,MAAL;;;SAGK1B,CAAP;;;AAGF,AAAe,SAASiJ,WAAT,CAAqBC,QAArB,EAA+BlJ,CAA/B,EAAkC;WACtCuB,IAAT,CAAc,KAAd,EAAqBpB,IAArB,CAA0B,UAACgB,KAAD,EAAQJ,GAAR,EAAgB;QAClC2D,OAAO1E,EAAEe,GAAF,CAAb;;mBAEe2D,IAAf,EAAqB1E,CAArB;kBACc0E,IAAd,EAAoB1E,CAApB;GAJF;;SAOOA,CAAP;;;AClCa,SAASmJ,UAAT,CAAoBC,OAApB,EAA6BpJ,CAA7B,EAA2C;MAAXqJ,IAAW,uEAAJ,EAAI;;MACpDA,KAAK9G,MAAL,KAAgB,CAApB,EAAuB;WACduC,cAAP;;;IAGAuE,KAAKhL,IAAL,CAAU,GAAV,CAAF,EAAkB+K,OAAlB,EAA2BE,QAA3B,CAAoCzE,UAApC;;SAEO7E,CAAP;;;ACPa,SAASuJ,aAAT,CAAuBH,OAAvB,EAAgCpJ,CAAhC,EAA8C;MAAXqJ,IAAW,uEAAJ,EAAI;;MACvDA,KAAK9G,MAAL,KAAgB,CAApB,EAAuB;WACdwC,iBAAP;;;;;IAKAsE,KAAKhL,IAAL,CAAU,GAAV,CAAF,EAAkB+K,OAAlB,EAA2BjC,GAA3B,OAAmCtC,UAAnC,EAAiDnD,MAAjD;;;UAGMmD,UAAN,EAAoBuE,OAApB,EAA6BI,WAA7B,CAAyC3E,UAAzC;;SAEO7E,CAAP;;;ACfF;;;;AAGA,AAAe,SAASyJ,UAAT,CAAoBL,OAApB,EAA6BpJ,CAA7B,EAAgC;MACvC0J,SAAS1J,EAAE,IAAF,EAAQoJ,OAAR,CAAf;;MAEIM,OAAOnH,MAAP,GAAgB,CAApB,EAAuB;WACdpC,IAAP,CAAY,UAACgB,KAAD,EAAQd,IAAR;aAAiBL,EAAEK,IAAF,EAAQqB,MAAR,EAAjB;KAAZ;GADF,MAEO;WACEvB,IAAP,CAAY,UAACgB,KAAD,EAAQd,IAAR,EAAiB;oBACbL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0B,IAA1B;KADF;;;SAKKA,CAAP;;;ACdF,SAAS2J,qBAAT,CAA+BT,QAA/B,EAAyC;;WAE9B3H,IAAT,CAAc,GAAd,EAAmBpB,IAAnB,CAAwB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;SAClCW,OAAL,GAAe,iBAAgBX,KAAKW,OAArB,EAA8B4I,MAA9B,CAAqC,UAACC,GAAD,EAAMrJ,IAAN,EAAe;UAC7D8E,mBAAmB/F,IAAnB,CAAwBiB,IAAxB,CAAJ,EAAmC;4BACrBqJ,GAAZ,sBAAkBrJ,IAAlB,EAAyBH,KAAKW,OAAL,CAAaR,IAAb,CAAzB;;;aAGKqJ,GAAP;KALa,EAMZ,EANY,CAAf;GADF;;;;;;;;;;AAkBF,AAAe,SAASC,eAAT,CAAyBZ,QAAzB,EAAmC;wBAC1BA,QAAtB;;SAEOA,QAAP;;;ACzBa,SAASa,WAAT,CAAqBb,QAArB,EAA+BlJ,CAA/B,EAAkC;WACtCuB,IAAT,CAAc,GAAd,EAAmBpB,IAAnB,CAAwB,UAACgB,KAAD,EAAQ6G,CAAR,EAAc;QAC9BgC,KAAKhK,EAAEgI,CAAF,CAAX;QACIgC,GAAGzI,IAAH,CAAQ,aAAR,EAAuBgB,MAAvB,KAAkC,CAAlC,IAAuCyH,GAAGC,IAAH,GAAUC,IAAV,OAAqB,EAAhE,EAAoEF,GAAGtI,MAAH;GAFtE;;SAKO1B,CAAP;;;ACNF;;;;;;AAMA,AAAO,IAAM6F,kCAAgC,CAC3C,UAD2C,EAE3C,OAF2C,EAG3C,QAH2C,EAI3C,SAJ2C,EAK3C,SAL2C,EAM3C,KAN2C,EAO3C,gBAP2C,EAQ3C,OAR2C,EAS3C,SAT2C,EAU3C,cAV2C,EAW3C,QAX2C,EAY3C,iBAZ2C,EAa3C,OAb2C,EAc3C,MAd2C,EAe3C,MAf2C,EAgB3C,QAhB2C,EAiB3C,QAjB2C,EAkB3C,QAlB2C,EAmB3C,OAnB2C;AAoB3C,MApB2C,EAqB3C,MArB2C,EAsB3C,KAtB2C,EAuB3C,OAvB2C,EAwB3C,YAxB2C,EAyB3C,UAzB2C;AA0B3C,2BA1B2C;AA2B3C,OA3B2C,EA4B3C,eA5B2C,EA6B3C,SA7B2C,EA8B3C,QA9B2C,EA+B3C,QA/B2C,EAgC3C,KAhC2C,EAiC3C,OAjC2C,EAkC3C,UAlC2C,EAmC3C,SAnC2C,EAoC3C,UApC2C,EAqC3C,SArC2C,EAsC3C,OAtC2C,CAAtC;;;;;;;;;;;;;AAoDP,AAAO,IAAMC,kCAAgC,CAC3C,KAD2C,EAE3C,SAF2C,EAG3C,MAH2C,EAI3C,WAJ2C,EAK3C,QAL2C,EAM3C,SAN2C,EAO3C,qBAP2C,EAQ3C,QAR2C;AAS3C,OAT2C,EAU3C,QAV2C,EAW3C,OAX2C,EAY3C,MAZ2C,EAa3C,MAb2C,EAc3C,OAd2C,EAe3C,QAf2C,CAAtC;;;;;AAqBP,AAAO,IAAMC,wBAAsB,CACjC,GADiC,EAEjC,YAFiC,EAGjC,IAHiC,EAIjC,KAJiC,EAKjC,KALiC,EAMjC,GANiC,EAOjC,KAPiC,EAQjC,OARiC,EASjC1H,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM2H,2BAAyB,CACpC,IADoC,EAEpC,GAFoC,EAGpC,GAHoC,EAIpC,OAJoC,EAKpC,IALoC,EAMpC,MANoC,EAOpC,MAPoC,EAQpC,UARoC,EASpC,OAToC,EAUpC,KAVoC,EAWpC,MAXoC,EAYpC,MAZoC,CAA/B;;AAeP,AAAO,IAAMC,8BACX,IAAI7H,MAAJ,QAAgB4H,yBAAuB3H,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;;;;AAMP,AAAO,IAAM8L,4BAA0B,CACrC,CAAC,SAAD,EAAY,gBAAZ,CADqC,EAErC,CAAC,OAAD,EAAU,gBAAV,CAFqC,EAGrC,CAAC,QAAD,EAAW,gBAAX,CAHqC,EAIrC,CAAC,OAAD,EAAU,WAAV,CAJqC,EAKrC,CAAC,OAAD,EAAU,YAAV,CALqC,EAMrC,CAAC,OAAD,EAAU,YAAV,CANqC,CAAhC;;AASP,AAAO,IAAMjE,gBAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,mBAAiB,IAAI/H,MAAJ,CAAW8H,cAAY7H,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAM+H,yBAAuB,CAClC,SADkC,EAElC,gBAFkC,EAGlC,iBAHkC,EAIlC,MAJkC,EAKlC,MALkC,EAMlC,SANkC,EAOlC,qBAPkC,EAQlC,OARkC,EASlC,QATkC,EAUlC,MAVkC,EAWlC,QAXkC,EAYlC,MAZkC,EAalC,YAbkC,EAclC,WAdkC,EAelC,MAfkC,EAgBlC,OAhBkC,EAiBlC,MAjBkC,EAkBlC,UAlBkC;AAmBlC,SAnBkC,CAA7B;;;AAuBP,AAAO,IAAMC,sBAAoB,IAAIjI,MAAJ,CAAWgI,uBAAqB/H,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAM+L,sBAAoB,IAAIhM,MAAJ,CAAW,qBAAX,EAAkC,GAAlC,CAA1B;;;;;;AAMP,AAAO,IAAMkI,yBAAuB,CAClC,OADkC,EAElC,QAFkC,EAGlC,QAHkC,EAIlC,KAJkC,EAKlC,UALkC,EAMlC,QANkC,EAOlC,QAPkC,EAQlC,OARkC,EASlC,MATkC,EAUlC,OAVkC,EAWlC,SAXkC,EAYlC,YAZkC,EAalC,SAbkC,EAclC,MAdkC,EAelC,QAfkC,EAgBlC,OAhBkC,EAiBlC,MAjBkC,EAkBlC,MAlBkC,EAmBlC,SAnBkC,EAoBlC,UApBkC;AAqBlC,MArBkC,EAsBlC,QAtBkC,EAuBlC,UAvBkC,EAwBlC,MAxBkC,EAyBlC,MAzBkC,EA0BlC,MA1BkC,EA2BlC,UA3BkC;AA4BlC,mBA5BkC,EA6BlC,MA7BkC,EA8BlC,WA9BkC,EA+BlC,MA/BkC,EAgClC,UAhCkC,EAiClC,OAjCkC,EAkClC,MAlCkC,EAmClC,OAnCkC,EAoClC,UApCkC;AAqClC,OArCkC,EAsClC,KAtCkC;AAuClC,SAvCkC,EAwClC,SAxCkC,EAyClC,cAzCkC;AA0ClC,QA1CkC,EA2ClC,WA3CkC,EA4ClC,OA5CkC,EA6ClC,UA7CkC,EA8ClC,UA9CkC,EA+ClC,MA/CkC,EAgDlC,SAhDkC,EAiDlC,SAjDkC,EAkDlC,OAlDkC,EAmDlC,KAnDkC,EAoDlC,SApDkC,EAqDlC,MArDkC,EAsDlC,OAtDkC,EAuDlC,QAvDkC,CAA7B;;AA0DP,AAAO,IAAMC,sBAAoB,IAAInI,MAAJ,CAAWkI,uBAAqBjI,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA,AAGA,AAGA;;AAEA,AAAO,IAAMuI,qBAAmB,CAC9B,SAD8B,EAE9B,OAF8B,EAG9B,YAH8B,EAI9B,MAJ8B,EAK9B,IAL8B,EAM9B,QAN8B,EAO9B,QAP8B,EAQ9B,SAR8B,EAS9B,KAT8B,EAU9B,UAV8B,EAW9B,IAX8B,EAY9B,KAZ8B,EAa9B,IAb8B,EAc9B,IAd8B,EAe9B,OAf8B,EAgB9B,UAhB8B,EAiB9B,YAjB8B,EAkB9B,QAlB8B,EAmB9B,QAnB8B,EAoB9B,MApB8B,EAqB9B,IArB8B,EAsB9B,IAtB8B,EAuB9B,IAvB8B,EAwB9B,IAxB8B,EAyB9B,IAzB8B,EA0B9B,IA1B8B,EA2B9B,QA3B8B,EA4B9B,QA5B8B,EA6B9B,IA7B8B,EA8B9B,IA9B8B,EA+B9B,KA/B8B,EAgC9B,QAhC8B,EAiC9B,IAjC8B,EAkC9B,QAlC8B,EAmC9B,GAnC8B,EAoC9B,KApC8B,EAqC9B,UArC8B,EAsC9B,SAtC8B,EAuC9B,OAvC8B,EAwC9B,OAxC8B,EAyC9B,UAzC8B,EA0C9B,OA1C8B,EA2C9B,IA3C8B,EA4C9B,OA5C8B,EA6C9B,IA7C8B,EA8C9B,IA9C8B,EA+C9B,OA/C8B,CAAzB;AAiDP,AAAO,IAAMC,wBAAsB,IAAIzI,MAAJ,QAAgBwI,mBAAiBvI,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAMyI,wBAAsBjB,gCAA8BxH,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,IAAM2I,wBAAsBlB,gCAA8BzH,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,AAGA,AAAO,IAAMgM,yBAAuB,IAAIjM,MAAJ,CAAW,mBAAX,EAAgC,GAAhC,CAA7B;AACP,AAAO,IAAMkM,uBAAqB,IAAIlM,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAA3B;AACP,AAAO,IAAMmM,aAAW,IAAInM,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB,CAEP;;AC3SA;AACA,AAAe,SAASoM,SAAT,CAAmBnK,IAAnB,EAAyB;MAChC+G,UAAU/G,KAAKG,IAAL,CAAU,OAAV,CAAhB;MACM6G,KAAKhH,KAAKG,IAAL,CAAU,IAAV,CAAX;MACIiK,QAAQ,CAAZ;;MAEIpD,EAAJ,EAAQ;;QAEFhB,oBAAkB9G,IAAlB,CAAuB8H,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;QAEEd,oBAAkBhH,IAAlB,CAAuB8H,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;;;MAIAD,OAAJ,EAAa;QACPqD,UAAU,CAAd,EAAiB;;;UAGXpE,oBAAkB9G,IAAlB,CAAuB6H,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;UAEEb,oBAAkBhH,IAAlB,CAAuB6H,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;;;;;;QAOAjB,iBAAe5G,IAAf,CAAoB6H,OAApB,CAAJ,EAAkC;eACvB,EAAT;;;;;;;QAOEgD,oBAAkB7K,IAAlB,CAAuB6H,OAAvB,CAAJ,EAAqC;eAC1B,EAAT;;;;SAIGqD,KAAP;;;ACpDF;;;AAGA,AAAe,SAASC,QAAT,CAAkBpK,KAAlB,EAAyB;SAC/BqK,WAAWrK,MAAME,IAAN,CAAW,OAAX,CAAX,KAAmC,IAA1C;;;ACJF;AACA,AAAe,SAASoK,WAAT,CAAqBX,IAArB,EAA2B;SACjC,CAACA,KAAKY,KAAL,CAAW,IAAX,KAAoB,EAArB,EAAyBtI,MAAhC;;;ACFF,IAAMuI,QAAQ,IAAI1M,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAd;;AAEA,AAAe,SAAS2M,WAAT,CAAqBC,UAArB,EAAgD;MAAftI,OAAe,uEAAL,GAAK;;MACvDuI,SAASD,aAAa,EAA5B;;MAEIC,SAAS,CAAb,EAAgB;QACVC,oBAAJ;;;;;;;QAOIJ,MAAMvL,IAAN,CAAWmD,OAAX,CAAJ,EAAyB;oBACTuI,SAAS,CAAvB;KADF,MAEO;oBACSA,SAAS,IAAvB;;;WAGKE,KAAKC,GAAL,CAASD,KAAKE,GAAL,CAASH,WAAT,EAAsB,CAAtB,CAAT,EAAmC,CAAnC,CAAP;;;SAGK,CAAP;;;ACjBF;;AAEA,AAAe,SAASI,cAAT,CAAwBjL,IAAxB,EAA8B;MACvCoK,QAAQ,CAAZ;MACMR,OAAO5J,KAAK4J,IAAL,GAAYC,IAAZ,EAAb;MACMc,aAAaf,KAAK1H,MAAxB;;;MAGIyI,aAAa,EAAjB,EAAqB;WACZ,CAAP;;;;WAIOJ,YAAYX,IAAZ,CAAT;;;;WAISc,YAAYC,UAAZ,CAAT;;;;;;MAMIf,KAAKsB,KAAL,CAAW,CAAC,CAAZ,MAAmB,GAAvB,EAA4B;aACjB,CAAT;;;SAGKd,KAAP;;;AC/Ba,SAASe,QAAT,CAAkBlL,KAAlB,EAAyBN,CAAzB,EAA4ByK,KAA5B,EAAmC;QAC1CjK,IAAN,CAAW,OAAX,EAAoBiK,KAApB;SACOnK,KAAP;;;ACEa,SAASmL,QAAT,CAAkBnL,KAAlB,EAAyBN,CAAzB,EAA4B0L,MAA5B,EAAoC;MAC7C;QACIjB,QAAQkB,eAAerL,KAAf,EAAsBN,CAAtB,IAA2B0L,MAAzC;aACSpL,KAAT,EAAgBN,CAAhB,EAAmByK,KAAnB;GAFF,CAGE,OAAOmB,CAAP,EAAU;;;;SAILtL,KAAP;;;ACXF;AACA,AAAe,SAASuL,WAAT,CAAqBxL,IAArB,EAA2BL,CAA3B,EAA8ByK,KAA9B,EAAqC;MAC5CqB,SAASzL,KAAKyL,MAAL,EAAf;MACIA,MAAJ,EAAY;aACDA,MAAT,EAAiB9L,CAAjB,EAAoByK,QAAQ,IAA5B;;;SAGKpK,IAAP;;;ACFF;;;AAGA,AAAe,SAASsL,cAAT,CAAwBrL,KAAxB,EAA+BN,CAA/B,EAAsD;MAApB+L,WAAoB,uEAAN,IAAM;;MAC/DtB,QAAQC,SAASpK,KAAT,CAAZ;;MAEImK,KAAJ,EAAW;WACFA,KAAP;;;UAGMuB,UAAU1L,KAAV,CAAR;;MAEIyL,WAAJ,EAAiB;aACNvB,UAAUlK,KAAV,CAAT;;;cAGUA,KAAZ,EAAmBN,CAAnB,EAAsByK,KAAtB;;SAEOA,KAAP;;;AClBF;;AAEA,AAAe,SAASuB,SAAT,CAAmB1L,KAAnB,EAA0B;mBACnBA,MAAM/B,GAAN,CAAU,CAAV,CADmB;;MAC/BmE,OAD+B,cAC/BA,OAD+B;;;;;;MAMnC2H,uBAAqB9K,IAArB,CAA0BmD,OAA1B,CAAJ,EAAwC;WAC/B4I,eAAehL,KAAf,CAAP;GADF,MAEO,IAAIoC,YAAY,KAAhB,EAAuB;WACrB,CAAP;GADK,MAEA,IAAI4H,qBAAmB/K,IAAnB,CAAwBmD,OAAxB,CAAJ,EAAsC;WACpC,CAAP;GADK,MAEA,IAAI6H,WAAShL,IAAT,CAAcmD,OAAd,CAAJ,EAA4B;WAC1B,CAAC,CAAR;GADK,MAEA,IAAIA,YAAY,IAAhB,EAAsB;WACpB,CAAC,CAAR;;;SAGK,CAAP;;;ACjBF,SAAS4F,cAAT,CAAsBhI,KAAtB,EAA6BN,CAA7B,EAAgC;MAC1BM,MAAM/B,GAAN,CAAU,CAAV,CAAJ,EAAkB;qBACI+B,MAAM/B,GAAN,CAAU,CAAV,CADJ;;QACRmE,OADQ,cACRA,OADQ;;;QAGZA,YAAY,MAAhB,EAAwB;;oBAERpC,KAAd,EAAqBN,CAArB,EAAwB,KAAxB;;;;;AAKN,SAASiM,UAAT,CAAoB3L,KAApB,EAA2BN,CAA3B,EAA8ByK,KAA9B,EAAqC;MAC/BnK,KAAJ,EAAW;mBACIA,KAAb,EAAoBN,CAApB;aACSM,KAAT,EAAgBN,CAAhB,EAAmByK,KAAnB;;;;AAIJ,SAASyB,OAAT,CAAiBlM,CAAjB,EAAoB+L,WAApB,EAAiC;IAC7B,QAAF,EAAY5E,GAAZ,CAAgB,SAAhB,EAA2BhH,IAA3B,CAAgC,UAACgB,KAAD,EAAQd,IAAR,EAAiB;;;QAG3CC,QAAQN,EAAEK,IAAF,CAAZ;YACQmL,SAASlL,KAAT,EAAgBN,CAAhB,EAAmB2L,eAAerL,KAAf,EAAsBN,CAAtB,EAAyB+L,WAAzB,CAAnB,CAAR;;QAEMlJ,UAAUvC,MAAMwL,MAAN,EAAhB;QACMK,WAAWH,UAAU1L,KAAV,CAAjB;;eAEWuC,OAAX,EAAoB7C,CAApB,EAAuBmM,QAAvB,EAAiCJ,WAAjC;QACIlJ,OAAJ,EAAa;;;iBAGAA,QAAQiJ,MAAR,EAAX,EAA6B9L,CAA7B,EAAgCmM,WAAW,CAA3C,EAA8CJ,WAA9C;;GAbJ;;SAiBO/L,CAAP;;;;;AAKF,AAAe,SAASoM,YAAT,CAAsBpM,CAAtB,EAA6C;MAApB+L,WAAoB,uEAAN,IAAM;;;;4BAGlC9K,OAAxB,CAAgC,gBAAqC;;;QAAnCoL,cAAmC;QAAnBC,aAAmB;;MAC9DD,cAAL,SAAuBC,aAAvB,EAAwCnM,IAAxC,CAA6C,UAACgB,KAAD,EAAQd,IAAR,EAAiB;eACnDL,EAAEK,IAAF,EAAQyL,MAAR,CAAeO,cAAf,CAAT,EAAyCrM,CAAzC,EAA4C,EAA5C;KADF;GADF;;;;;;;UAWQA,CAAR,EAAW+L,WAAX;UACQ/L,CAAR,EAAW+L,WAAX;;SAEO/L,CAAP;;;ACpEF,IAAMuM,eAAe,SAArB;;AAEA,AAAe,SAASC,eAAT,CAAyBvC,IAAzB,EAA+B;SACrCA,KAAKzG,OAAL,CAAa+I,YAAb,EAA2B,GAA3B,EAAgCrC,IAAhC,EAAP;;;ACHF;;;;;AAKA,AAAe,SAASuC,cAAT,CAAwBjN,GAAxB,EAA6BkN,SAA7B,EAAwC;MAC/CC,UAAUD,UAAUnL,IAAV,CAAe;WAAMqL,GAAGrN,IAAH,CAAQC,GAAR,CAAN;GAAf,CAAhB;MACImN,OAAJ,EAAa;WACJA,QAAQE,IAAR,CAAarN,GAAb,EAAkB,CAAlB,CAAP;;;SAGK,IAAP;;;ACXF;;;;;;;;;;;;;;;;AAgBA,AAAO,IAAMsN,kBAAkB,IAAI1O,MAAJ,CAAW,0EAAX,EAAuF,GAAvF,CAAxB;;AAEP,AAAO,IAAM2O,eAAe,QAArB;;AAEP,AAAO,IAAMC,cAAc,WAApB;AACP,AAAO,IAAMC,cAAc,WAApB;;ACnBQ,SAASC,cAAT,CAAwB1N,GAAxB,EAA6B;MACpC2N,UAAU3N,IAAIqL,KAAJ,CAAUiC,eAAV,CAAhB;MACI,CAACK,OAAL,EAAc,OAAO,IAAP;;MAERC,UAAUrE,SAASoE,QAAQ,CAAR,CAAT,EAAqB,EAArB,CAAhB;;;;SAIOC,UAAU,GAAV,GAAgBA,OAAhB,GAA0B,IAAjC;;;ACVa,SAASC,YAAT,CAAsB7N,GAAtB,EAA2B;SACjCA,IAAI4E,KAAJ,CAAU,GAAV,EAAe,CAAf,EAAkBZ,OAAlB,CAA0B,KAA1B,EAAiC,EAAjC,CAAP;;;ACOF,SAAS8J,aAAT,CAAuBC,OAAvB,EAAgCpM,KAAhC,EAAuCqM,sBAAvC,EAA+D;MACzDC,cAAc,IAAlB;;;;MAIItM,QAAQ,CAAR,IAAa8L,YAAY1N,IAAZ,CAAiBgO,OAAjB,CAAb,IAA0CA,QAAQhL,MAAR,GAAiB,CAA/D,EAAkE;kBAClD,IAAd;;;;;MAKEpB,UAAU,CAAV,IAAeoM,QAAQG,WAAR,OAA0B,OAA7C,EAAsD;kBACtC,KAAd;;;;;MAKEvM,QAAQ,CAAR,IAAaoM,QAAQhL,MAAR,GAAiB,CAA9B,IAAmC,CAACiL,sBAAxC,EAAgE;kBAChD,KAAd;;;SAGKC,WAAP;;;;;;AAMF,AAAe,SAASE,cAAT,CAAwBnO,GAAxB,EAA6BoO,MAA7B,EAAqC;MAC5CnO,YAAYmO,UAAUlO,IAAIC,KAAJ,CAAUH,GAAV,CAA5B;MACQqO,QAF0C,GAEjBpO,SAFiB,CAE1CoO,QAF0C;MAEhCC,IAFgC,GAEjBrO,SAFiB,CAEhCqO,IAFgC;MAE1BC,IAF0B,GAEjBtO,SAFiB,CAE1BsO,IAF0B;;;MAI9CP,yBAAyB,KAA7B;MACMQ,kBAAkBD,KAAK3J,KAAL,CAAW,GAAX,EACvB6J,OADuB,GAEvBrE,MAFuB,CAEhB,UAACC,GAAD,EAAMqE,UAAN,EAAkB/M,KAAlB,EAA4B;QAC9BoM,UAAUW,UAAd;;;QAGIX,QAAQrL,QAAR,CAAiB,GAAjB,CAAJ,EAA2B;2BACUqL,QAAQnJ,KAAR,CAAc,GAAd,CADV;;;;UAClB+J,eADkB;UACDC,OADC;;UAErBpB,YAAYzN,IAAZ,CAAiB6O,OAAjB,CAAJ,EAA+B;kBACnBD,eAAV;;;;;;QAMArB,gBAAgBvN,IAAhB,CAAqBgO,OAArB,KAAiCpM,QAAQ,CAA7C,EAAgD;gBACpCoM,QAAQ/J,OAAR,CAAgBsJ,eAAhB,EAAiC,EAAjC,CAAV;;;;;;;QAOE3L,UAAU,CAAd,EAAiB;+BACU4L,aAAaxN,IAAb,CAAkBgO,OAAlB,CAAzB;;;;QAIED,cAAcC,OAAd,EAAuBpM,KAAvB,EAA8BqM,sBAA9B,CAAJ,EAA2D;UACrDa,IAAJ,CAASd,OAAT;;;WAGK1D,GAAP;GAhCsB,EAiCrB,EAjCqB,CAAxB;;SAmCUgE,QAAV,UAAuBC,IAAvB,GAA8BE,gBAAgBC,OAAhB,GAA0B5P,IAA1B,CAA+B,GAA/B,CAA9B;;;AC3EF;;AAEA,IAAMiQ,kBAAkB,IAAIlQ,MAAJ,CAAW,QAAX,CAAxB;AACA,AAAe,SAASmQ,cAAT,CAAwBtE,IAAxB,EAA8B;SACpCqE,gBAAgB/O,IAAhB,CAAqB0K,IAArB,CAAP;;;ACJa,SAASuE,cAAT,CAAwBvM,OAAxB,EAA6C;kBAAZwM,KAAY,uEAAJ,EAAI;;qBACnDxM,QAAQiI,IAAR,GACQ9F,KADR,CACc,KADd,EAEQmH,KAFR,CAEc,CAFd,EAEiBkD,KAFjB,EAGQpQ,IAHR,CAGa,GAHb,CAAP;;;ACQF;;;;;AAKA,AAAe,SAASqQ,aAAT,CAAuBC,UAAvB,EAAmCC,QAAnC,EAA6C5O,CAA7C,EAAgD;MACzD,CAAC2O,WAAW7C,MAAX,GAAoBvJ,MAAzB,EAAiC;WACxBoM,UAAP;;;MAGIE,wBAAwB1D,KAAKE,GAAL,CAAS,EAAT,EAAauD,WAAW,IAAxB,CAA9B;MACME,cAAc9O,EAAE,aAAF,CAApB;;aAEW8L,MAAX,GAAoBxJ,QAApB,GAA+BnC,IAA/B,CAAoC,UAACgB,KAAD,EAAQ2G,OAAR,EAAoB;QAChDiH,WAAW/O,EAAE8H,OAAF,CAAjB;;QAEI7B,4BAA0B1G,IAA1B,CAA+BuI,QAAQpF,OAAvC,CAAJ,EAAqD;aAC5C,IAAP;;;QAGIsM,eAAetE,SAASqE,QAAT,CAArB;QACIC,YAAJ,EAAkB;UACZD,aAAaJ,UAAjB,EAA6B;oBACfxL,MAAZ,CAAmB4L,QAAnB;OADF,MAEO;YACDE,eAAe,CAAnB;YACMC,UAAUC,YAAYJ,QAAZ,CAAhB;;;;YAIIG,UAAU,IAAd,EAAoB;0BACF,EAAhB;;;;;YAKEA,WAAW,GAAf,EAAoB;0BACF,EAAhB;;;;;YAKEH,SAASvO,IAAT,CAAc,OAAd,MAA2BmO,WAAWnO,IAAX,CAAgB,OAAhB,CAA/B,EAAyD;0BACvCoO,WAAW,GAA3B;;;YAGIQ,WAAWJ,eAAeC,YAAhC;;YAEIG,YAAYP,qBAAhB,EAAuC;iBAC9BC,YAAY3L,MAAZ,CAAmB4L,QAAnB,CAAP;SADF,MAEO,IAAIjH,QAAQpF,OAAR,KAAoB,GAAxB,EAA6B;cAC5B2M,iBAAiBN,SAAS9E,IAAT,EAAvB;cACMqF,uBAAuBtE,WAAWqE,cAAX,CAA7B;;cAEIC,uBAAuB,EAAvB,IAA6BJ,UAAU,IAA3C,EAAiD;mBACxCJ,YAAY3L,MAAZ,CAAmB4L,QAAnB,CAAP;WADF,MAEO,IAAIO,wBAAwB,EAAxB,IAA8BJ,YAAY,CAA1C,IACDX,eAAec,cAAf,CADH,EACmC;mBACjCP,YAAY3L,MAAZ,CAAmB4L,QAAnB,CAAP;;;;;;WAMD,IAAP;GAnDF;;SAsDOD,WAAP;;;ACxEF;;AAEA,AAAe,SAASS,gBAAT,CAA0BvP,CAA1B,EAA6B;MACtC2O,mBAAJ;MACIC,WAAW,CAAf;;IAEE,SAAF,EAAazO,IAAb,CAAkB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;;QAE7B4F,4BAA0B1G,IAA1B,CAA+Bc,KAAKqC,OAApC,CAAJ,EAAkD;;;;QAI5CpC,QAAQN,EAAEK,IAAF,CAAd;QACMoK,QAAQC,SAASpK,KAAT,CAAd;;QAEImK,QAAQmE,QAAZ,EAAsB;iBACTnE,KAAX;mBACanK,KAAb;;GAXJ;;;;MAiBI,CAACqO,UAAL,EAAiB;WACR3O,EAAE,MAAF,KAAaA,EAAE,GAAF,EAAOwP,KAAP,EAApB;;;eAGWd,cAAcC,UAAd,EAA0BC,QAA1B,EAAoC5O,CAApC,CAAb;;SAEO2O,UAAP;;;ACtBF,SAASc,mBAAT,CAA6BnP,KAA7B,EAAoCN,CAApC,EAAuC0P,MAAvC,EAA+C;;;;;MAKzCpP,MAAMqP,QAAN,CAAe,qBAAf,CAAJ,EAA2C;;;;MAIrC1N,UAAUuK,gBAAgBlM,MAAM2J,IAAN,EAAhB,CAAhB;;MAEIW,YAAY3I,OAAZ,IAAuB,EAA3B,EAA+B;QACvB2N,SAAS5P,EAAE,GAAF,EAAOM,KAAP,EAAciC,MAA7B;QACMsN,aAAa7P,EAAE,OAAF,EAAWM,KAAX,EAAkBiC,MAArC;;;QAGIsN,aAAcD,SAAS,CAA3B,EAA+B;YACvBlO,MAAN;;;;QAIIpC,gBAAgB2C,QAAQM,MAA9B;QACMuN,WAAW9P,EAAE,KAAF,EAASM,KAAT,EAAgBiC,MAAjC;;;;QAIIjD,gBAAgB,EAAhB,IAAsBwQ,aAAa,CAAvC,EAA0C;YAClCpO,MAAN;;;;QAIIwN,UAAUC,YAAY7O,KAAZ,CAAhB;;;;;QAKIoP,SAAS,EAAT,IAAeR,UAAU,GAAzB,IAAgC5P,gBAAgB,EAApD,EAAwD;YAChDoC,MAAN;;;;;;QAMEgO,UAAU,EAAV,IAAgBR,UAAU,GAA9B,EAAmC;;;;UAI3BxM,UAAUpC,MAAM/B,GAAN,CAAU,CAAV,EAAamE,OAA7B;UACMqN,aAAarN,YAAY,IAAZ,IAAoBA,YAAY,IAAnD;UACIqN,UAAJ,EAAgB;YACRC,eAAe1P,MAAM2P,IAAN,EAArB;YACID,gBAAgBxD,gBAAgBwD,aAAa/F,IAAb,EAAhB,EAAqCsB,KAArC,CAA2C,CAAC,CAA5C,MAAmD,GAAvE,EAA4E;;;;;YAKxE7J,MAAN;;;;QAIIwO,cAAclQ,EAAE,QAAF,EAAYM,KAAZ,EAAmBiC,MAAvC;;;QAGI2N,cAAc,CAAd,IAAmB5Q,gBAAgB,GAAvC,EAA4C;YACpCoC,MAAN;;;;;;;;;;;;;AAaN,AAAe,SAASyO,SAAT,CAAmBjH,QAAnB,EAA6BlJ,CAA7B,EAAgC;IAC3C0F,wBAAF,EAA4BwD,QAA5B,EAAsC/I,IAAtC,CAA2C,UAACgB,KAAD,EAAQd,IAAR,EAAiB;QACpDC,QAAQN,EAAEK,IAAF,CAAd;QACIqP,SAAShF,SAASpK,KAAT,CAAb;QACI,CAACoP,MAAL,EAAa;eACF/D,eAAerL,KAAf,EAAsBN,CAAtB,CAAT;eACSM,KAAT,EAAgBN,CAAhB,EAAmB0P,MAAnB;;;;QAIEA,SAAS,CAAb,EAAgB;YACRhO,MAAN;KADF,MAEO;;0BAEepB,KAApB,EAA2BN,CAA3B,EAA8B0P,MAA9B;;GAbJ;;SAiBO1P,CAAP;;;ACrGa,SAASoQ,YAAT,CAAsBlH,QAAtB,EAAgClJ,CAAhC,EAA+C;MAAZqQ,KAAY,uEAAJ,EAAI;;IAC1DzK,eAAF,EAAmBsD,QAAnB,EAA6B/I,IAA7B,CAAkC,UAACgB,KAAD,EAAQmP,MAAR,EAAmB;QAC7CC,UAAUvQ,EAAEsQ,MAAF,CAAhB;;;;;QAKItQ,EAAEuQ,OAAF,EAAWrH,QAAX,EAAqBsH,OAArB,CAA6B,GAA7B,EAAkCjO,MAAlC,KAA6C,CAAjD,EAAoD;aAC3CgO,QAAQ7O,MAAR,EAAP;;;;QAIE8K,gBAAgBxM,EAAEsQ,MAAF,EAAUrG,IAAV,EAAhB,MAAsCoG,KAA1C,EAAiD;aACxCE,QAAQ7O,MAAR,EAAP;;;;;QAKE8I,UAAUxK,EAAEsQ,MAAF,CAAV,IAAuB,CAA3B,EAA8B;aACrBC,QAAQ7O,MAAR,EAAP;;;WAGK6O,OAAP;GArBF;;SAwBOvQ,CAAP;;;AC5BF;;;AAEA,AAAe,SAASyQ,eAAT,CAAyBrH,OAAzB,EAAkCpJ,CAAlC,EAAqC;;;;MAI9C0I,cAAc1I,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;MACI0I,cAAc1I,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;;SAEOA,CAAP;;;ACTF,SAAS0Q,UAAT,CAAoB1Q,CAApB,EAAuB2Q,OAAvB,EAAgCnQ,IAAhC,EAAsCoQ,QAAtC,EAAgD;UACxCpQ,IAAN,QAAeoQ,QAAf,EAAyBzQ,IAAzB,CAA8B,UAACC,CAAD,EAAIC,IAAJ,EAAa;QACnCb,MAAMa,KAAKW,OAAL,CAAaR,IAAb,CAAZ;QACMqQ,cAAcnR,IAAIjB,OAAJ,CAAYkS,OAAZ,EAAqBnR,GAArB,CAApB;;SAEKwB,OAAL,CAAaR,IAAb,IAAqBqQ,WAArB;GAJF;;;AAQF,AAAe,SAASC,iBAAT,CAA2BF,QAA3B,EAAqC5Q,CAArC,EAAwCR,GAAxC,EAA6C;GACzD,MAAD,EAAS,KAAT,EAAgByB,OAAhB,CAAwB;WAAQyP,WAAW1Q,CAAX,EAAcR,GAAd,EAAmBgB,IAAnB,EAAyBoQ,QAAzB,CAAR;GAAxB;;SAEOA,QAAP;;;ACbK,SAAS5F,UAAT,CAAoBf,IAApB,EAA0B;SACxBA,KAAKC,IAAL,GACK1G,OADL,CACa,MADb,EACqB,GADrB,EAEKjB,MAFZ;;;;;;AAQF,AAAO,SAAS4M,WAAT,CAAqB7O,KAArB,EAA4B;MAC3ByQ,kBAAkB/F,WAAW1K,MAAM2J,IAAN,EAAX,CAAxB;;MAEM+G,WAAW1Q,MAAMiB,IAAN,CAAW,GAAX,EAAgB0I,IAAhB,EAAjB;MACMgH,aAAajG,WAAWgG,QAAX,CAAnB;;MAEID,kBAAkB,CAAtB,EAAyB;WAChBE,aAAaF,eAApB;GADF,MAEO,IAAIA,oBAAoB,CAApB,IAAyBE,aAAa,CAA1C,EAA6C;WAC3C,CAAP;;;SAGK,CAAP;;;ACpBF;;;AAEA,AAAe,SAASC,eAAT,CACblR,CADa,EAEbmR,SAFa,EAGbC,WAHa,EAKb;MADAjB,SACA,uEADY,IACZ;;MACMkB,aAAaF,UAAU1P,MAAV,CAAiB;WAAQ2P,YAAYE,OAAZ,CAAoBC,IAApB,MAA8B,CAAC,CAAvC;GAAjB,CAAnB;;;;;;;;UAEWA,IAHX;;UAIQnQ,OAAO,MAAb;UACMb,QAAQ,OAAd;;UAEMiR,QAAQxR,YAAUoB,IAAV,UAAmBmQ,IAAnB,QAAd;;;;;UAKME,SACJD,MAAMtM,GAAN,CAAU,UAAC/D,KAAD,EAAQd,IAAR;eAAiBL,EAAEK,IAAF,EAAQG,IAAR,CAAaD,KAAb,CAAjB;OAAV,EACMmR,OADN,GAEMjQ,MAFN,CAEa;eAAQwI,SAAS,EAAjB;OAFb,CADF;;;;;;UASIwH,OAAOlP,MAAP,KAAkB,CAAtB,EAAyB;YACnBoP,kBAAJ;;;YAGIxB,SAAJ,EAAe;sBACDyB,UAAUH,OAAO,CAAP,CAAV,EAAqBzR,CAArB,CAAZ;SADF,MAEO;sBACOyR,OAAO,CAAP,CAAZ;;;;aAGKE;;;;;sCA5BQN,UAAnB,4GAA+B;;;;;;;;;;;;;;;;;;;;;;SAiCxB,IAAP;;;AC3CF,SAASQ,UAAT,CAAoBvR,KAApB,EAA2BwR,WAA3B,EAAwC;;;MAGlCxR,MAAMgC,QAAN,GAAiBC,MAAjB,GAA0BuP,WAA9B,EAA2C;WAClC,KAAP;;;MAGEC,cAAczR,KAAd,CAAJ,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;;;;AAMF,AAAe,SAAS0R,oBAAT,CACbhS,CADa,EAEbiS,SAFa,EAKb;MAFAH,WAEA,uEAFc,CAEd;MADAI,QACA,uEADW,IACX;;;;;;sCACuBD,SAAvB,4GAAkC;UAAvB9M,QAAuB;;UAC1BqM,QAAQxR,EAAEmF,QAAF,CAAd;;;;UAIIqM,MAAMjP,MAAN,KAAiB,CAArB,EAAwB;YAChBjC,QAAQN,EAAEwR,MAAM,CAAN,CAAF,CAAd;;YAEIK,WAAWvR,KAAX,EAAkBwR,WAAlB,CAAJ,EAAoC;cAC9B7P,gBAAJ;cACIiQ,QAAJ,EAAc;sBACF5R,MAAM2J,IAAN,EAAV;WADF,MAEO;sBACK3J,MAAM6R,IAAN,EAAV;;;cAGElQ,OAAJ,EAAa;mBACJA,OAAP;;;;;;;;;;;;;;;;;;;;SAMD,IAAP;;;AChDF;AACA,AAAe,SAAS2P,SAAT,CAAmB3H,IAAnB,EAAyBjK,CAAzB,EAA4B;;;MAGnCoS,YAAYpS,aAAWiK,IAAX,cAA0BA,IAA1B,EAAlB;SACOmI,cAAc,EAAd,GAAmBnI,IAAnB,GAA0BmI,SAAjC;;;ACLa,SAASL,aAAT,CAAuBzR,KAAvB,EAA8B;MACrCwC,UAAUxC,MAAMwC,OAAN,GAAgB4O,OAAhB,EAAhB;MACMW,gBAAgBvP,QAAQvB,IAAR,CAAa,UAACuK,MAAD,EAAY;QACvCxE,aAAgBwE,OAAO9K,OAAP,CAAesR,KAA/B,SAAwCxG,OAAO9K,OAAP,CAAeqG,EAA7D;WACOC,WAAWpF,QAAX,CAAoB,SAApB,CAAP;GAFoB,CAAtB;;SAKOmQ,kBAAkBE,SAAzB;;;ACPF;;;;AAIA,AAAe,SAASC,gBAAT,CAA0BlS,KAA1B,EAAiC;SACvCA,MAAM2J,IAAN,GAAaC,IAAb,GAAoB3H,MAApB,IAA8B,GAArC;;;ACHa,SAASkQ,WAAT,CAAqBzS,CAArB,EAAwB;SAC9BA,EAAEwG,cAAF,EAAkBjE,MAAlB,GAA2B,CAAlC;;;ACHF;AACA,AAAO,IAAMmQ,kBAAkB,wCAAxB;;;;AAIP,AAAO,IAAMC,eAAe,IAAIvU,MAAJ,CAAW,aAAX,EAA0B,GAA1B,CAArB;AACP,AAYA,AASA;AACA,AAAO,IAAMwU,iBAAiB,WAAvB;AACP,AAAO,IAAMC,kBAAkB,WAAxB;AACP,AAAO,IAAMC,uBAAuB,4BAA7B;AACP,AAAO,IAAMC,yBAAyB,oBAA/B;AACP,AAAO,IAAMC,wBAAwB,QAA9B;AACP,IAAMC,SAAS,CACb,KADa,EAEb,KAFa,EAGb,KAHa,EAIb,KAJa,EAKb,KALa,EAMb,KANa,EAOb,KAPa,EAQb,KARa,EASb,KATa,EAUb,KAVa,EAWb,KAXa,EAYb,KAZa,CAAf;AAcA,IAAMC,YAAYD,OAAO5U,IAAP,CAAY,GAAZ,CAAlB;AACA,IAAM8U,aAAa,qCAAnB;AACA,IAAMC,aAAa,wCAAnB;AACA,AAAO,IAAMC,oBACX,IAAIjV,MAAJ,OAAe+U,UAAf,WAA+BC,UAA/B,wBAA4DF,SAA5D,QAA0E,IAA1E,CADK;;;;;AAMP,AAAO,IAAMI,qBAAqB,gBAA3B;;AAEP,AAAO,IAAMC,oBACX,IAAInV,MAAJ,CAAW,2BAAX,EAAwC,GAAxC,CADK;;ACxDP;;AAEA,AAAe,SAASoV,WAAT,CAAqBC,MAArB,EAA6B;SACnCA,OAAOjQ,OAAP,CAAekP,eAAf,EAAgC,IAAhC,EAAsCxI,IAAtC,EAAP;;;ACHa,SAASvI,OAAT,CAAe+R,YAAf,EAA6B;iBAC3BA,aAAaxJ,IAAb,EAAf;MACIyJ,SAASC,QAAT,CAAkBF,YAAlB,CAAJ,EAAqC;WAC5BA,YAAP;;;SAGK,IAAP;;;ACHF;;AAEA,AAAe,SAASG,QAAT,CAAkBC,GAAlB,QAAuC;MAAd9T,CAAc,QAAdA,CAAc;MAAX+T,OAAW,QAAXA,OAAW;;;MAEhDD,IAAIvR,MAAJ,GAAa,IAAb,IAAqBuR,IAAIvR,MAAJ,GAAa,CAAtC,EAAyC,OAAO,IAAP;;;MAGrCwR,WAAWvF,eAAeuF,OAAf,EAAwB,EAAxB,MAAgCvF,eAAesF,GAAf,EAAoB,EAApB,CAA/C,EAAwE,OAAO,IAAP;;MAElEE,UAAUpC,UAAUkC,GAAV,EAAe9T,CAAf,CAAhB;;;;MAII2S,aAAapT,IAAb,CAAkByU,OAAlB,CAAJ,EAAgC,OAAO,IAAP;;SAEzBA,QAAQ9J,IAAR,EAAP;;;ACnBF;;;;AAIA,AASA,AAAO,SAAS+J,eAAT,CAAyBC,UAAzB,EAAqC;SACnC,CAACA,WAAWrJ,KAAX,CAAiBwI,iBAAjB,KAAuC,EAAxC,EACWhV,IADX,CACgB,GADhB,EAEWmF,OAFX,CAEmBwP,qBAFnB,EAE0C,GAF1C,EAGWxP,OAHX,CAGmBuP,sBAHnB,EAG2C,UAH3C,EAIWvP,OAJX,CAImBsP,oBAJnB,EAIyC,IAJzC,EAKW5I,IALX,EAAP;;;;;AAUF,AAAe,SAASiK,kBAAT,CAA4BD,UAA5B,EAAwC;;MAEjDtB,eAAerT,IAAf,CAAoB2U,UAApB,KAAmCrB,gBAAgBtT,IAAhB,CAAqB2U,UAArB,CAAvC,EAAyE;iBAC1DnL,SAASmL,UAAT,EAAqB,EAArB,CAAb;;;MAGEE,OAAOC,OAAO,IAAIC,IAAJ,CAASJ,UAAT,CAAP,CAAX;;MAEI,CAACE,KAAKG,OAAL,EAAL,EAAqB;iBACNN,gBAAgBC,UAAhB,CAAb;WACOG,OAAO,IAAIC,IAAJ,CAASJ,UAAT,CAAP,CAAP;;;SAGKE,KAAKG,OAAL,KAAiBH,KAAKI,WAAL,EAAjB,GAAsC,IAA7C;;;ACzBF;;AACA,AAAe,SAASC,gBAAT,CACbrL,OADa,QASb;MANEpJ,CAMF,QANEA,CAMF;mCALE0U,kBAKF;MALEA,kBAKF,yCALuB,IAKvB;wBAJErE,KAIF;MAJEA,KAIF,8BAJU,EAIV;sBAHE7Q,GAGF;MAHEA,GAGF,4BAHQ,EAGR;iCAFEmV,cAEF;MAFEA,cAEF,uCAFmB,IAEnB;;;;kBAGgBvL,OAAhB,EAAyBpJ,CAAzB;;;;;MAKI2U,cAAJ,EAAoB1L,YAAYG,OAAZ,EAAqBpJ,CAArB;;;;;aAKToJ,OAAX,EAAoBpJ,CAApB;;;;gBAIcoJ,OAAd,EAAuBpJ,CAAvB;;;;;aAKWoJ,OAAX,EAAoBpJ,CAApB;;;eAGaoJ,OAAb,EAAsBpJ,CAAtB,EAAyBqQ,KAAzB;;;oBAGkBjH,OAAlB,EAA2BpJ,CAA3B,EAA8BR,GAA9B;;;kBAGgB4J,OAAhB;;;;;;MAMIuL,cAAJ,EAAoBxE,UAAU/G,OAAV,EAAmBpJ,CAAnB,EAAsB0U,kBAAtB;;;cAGRtL,OAAZ,EAAqBpJ,CAArB;;SAEOoJ,OAAP;;;AC5Da,SAASwL,UAAT,CAAoBvE,KAApB,QAAuC;MAAV7Q,GAAU,QAAVA,GAAU;MAALQ,CAAK,QAALA,CAAK;;;;MAGhDsT,mBAAmB/T,IAAnB,CAAwB8Q,KAAxB,CAAJ,EAAoC;YAC1BwE,kBAAkBxE,KAAlB,EAAyB7Q,GAAzB,CAAR;;;;;MAKE6Q,MAAM9N,MAAN,GAAe,GAAnB,EAAwB;;QAEhBuS,KAAK9U,EAAE,IAAF,CAAX;QACI8U,GAAGvS,MAAH,KAAc,CAAlB,EAAqB;cACXuS,GAAG7K,IAAH,EAAR;;;;;SAKG2H,UAAUvB,KAAV,EAAiBrQ,CAAjB,EAAoBkK,IAApB,EAAP;;;ACfF,SAAS6K,sBAAT,CAAgCC,UAAhC,EAA4C/K,IAA5C,EAAkD;;;;MAI5C+K,WAAWzS,MAAX,IAAqB,CAAzB,EAA4B;;;;;UAIpB0S,aAAaD,WAAWpL,MAAX,CAAkB,UAACC,GAAD,EAAMqL,SAAN,EAAoB;YACnDA,SAAJ,IAAiBrL,IAAIqL,SAAJ,IAAiBrL,IAAIqL,SAAJ,IAAiB,CAAlC,GAAsC,CAAvD;eACOrL,GAAP;OAFiB,EAGhB,EAHgB,CAAnB;;kCAME,iBAAgBoL,UAAhB,EACQrL,MADR,CACe,UAACC,GAAD,EAAMjB,GAAN,EAAc;YAChBiB,IAAI,CAAJ,IAASoL,WAAWrM,GAAX,CAAb,EAA8B;iBACrB,CAACA,GAAD,EAAMqM,WAAWrM,GAAX,CAAN,CAAP;;;eAGKiB,GAAP;OANT,EAOU,CAAC,CAAD,EAAI,CAAJ,CAPV,CAVwB;;;;UASnBsL,OATmB;UASVC,SATU;;;;;;;UAuBtBA,aAAa,CAAb,IAAkBD,QAAQ5S,MAAR,IAAkB,CAAxC,EAA2C;qBAC5B0H,KAAK7F,KAAL,CAAW+Q,OAAX,CAAb;;;UAGIE,YAAY,CAACL,WAAW,CAAX,CAAD,EAAgBA,WAAWzJ,KAAX,CAAiB,CAAC,CAAlB,CAAhB,CAAlB;UACM+J,aAAaD,UAAUzL,MAAV,CAAiB,UAACC,GAAD,EAAMjM,GAAN;eAAciM,IAAItH,MAAJ,GAAa3E,IAAI2E,MAAjB,GAA0BsH,GAA1B,GAAgCjM,GAA9C;OAAjB,EAAoE,EAApE,CAAnB;;UAEI0X,WAAW/S,MAAX,GAAoB,EAAxB,EAA4B;;aACnB+S;;;;;WAGFrL;;;;;;;SAGF,IAAP;;;AAGF,SAASsL,oBAAT,CAA8BP,UAA9B,EAA0CxV,GAA1C,EAA+C;;;;;;;mBAO5BE,IAAIC,KAAJ,CAAUH,GAAV,CAP4B;;MAOrCsO,IAPqC,cAOrCA,IAPqC;;MAQvC0H,cAAc1H,KAAKtK,OAAL,CAAa+P,iBAAb,EAAgC,EAAhC,CAApB;;MAEMkC,YAAYT,WAAW,CAAX,EAActH,WAAd,GAA4BlK,OAA5B,CAAoC,GAApC,EAAyC,EAAzC,CAAlB;MACMkS,iBAAiBC,MAAMC,WAAN,CAAkBH,SAAlB,EAA6BD,WAA7B,CAAvB;;MAEIE,iBAAiB,GAAjB,IAAwBD,UAAUlT,MAAV,GAAmB,CAA/C,EAAkD;WACzCyS,WAAWzJ,KAAX,CAAiB,CAAjB,EAAoBlN,IAApB,CAAyB,EAAzB,CAAP;;;MAGIwX,UAAUb,WAAWzJ,KAAX,CAAiB,CAAC,CAAlB,EAAqB,CAArB,EAAwBmC,WAAxB,GAAsClK,OAAtC,CAA8C,GAA9C,EAAmD,EAAnD,CAAhB;MACMsS,eAAeH,MAAMC,WAAN,CAAkBC,OAAlB,EAA2BL,WAA3B,CAArB;;MAEIM,eAAe,GAAf,IAAsBD,QAAQtT,MAAR,IAAkB,CAA5C,EAA+C;WACtCyS,WAAWzJ,KAAX,CAAiB,CAAjB,EAAoB,CAAC,CAArB,EAAwBlN,IAAxB,CAA6B,EAA7B,CAAP;;;SAGK,IAAP;;;;;AAKF,AAAe,SAASwW,iBAAT,CAA2BxE,KAA3B,EAA4C;MAAV7Q,GAAU,uEAAJ,EAAI;;;;MAGnDwV,aAAa3E,MAAMjM,KAAN,CAAYkP,kBAAZ,CAAnB;MACI0B,WAAWzS,MAAX,KAAsB,CAA1B,EAA6B;WACpB8N,KAAP;;;MAGE0F,WAAWhB,uBAAuBC,UAAvB,EAAmC3E,KAAnC,CAAf;MACI0F,QAAJ,EAAc,OAAOA,QAAP;;aAEHR,qBAAqBP,UAArB,EAAiCxV,GAAjC,CAAX;MACIuW,QAAJ,EAAc,OAAOA,QAAP;;;;SAIP1F,KAAP;;;AC1FF,IAAM2F,WAAW;UACPxC,WADO;kBAECyC,OAFD;OAGVpC,QAHU;kBAICM,kBAJD;WAKN+B,gBALM;SAMRtB;CANT,CAUA,AAEA,AACA,AACA,AACA,AACA,AACA,AACA;;ACfA;;;;;;;;;;;AAWA,AAAe,SAASuB,eAAT,CAAyBnW,CAAzB,EAA4BoW,IAA5B,EAAkC;;;;;;;MAO3CA,KAAKlP,uBAAT,EAAkC;QAC5BA,wBAAwBlH,CAAxB,CAAJ;;;MAGEyI,oBAAoBzI,CAApB,CAAJ;MACIoM,aAAapM,CAAb,EAAgBoW,KAAKrK,WAArB,CAAJ;MACMsK,gBAAgB9G,iBAAiBvP,CAAjB,CAAtB;;SAEOqW,aAAP;;;AC5BF,IAAMC,0BAA0B;eACjB;6BACc,IADd;iBAEE,IAFF;wBAGS;GAJQ;;;;;;;;;;;;;;;;;;;;;SAAA,yBA0BGF,IA1BH,EA0BS;QAA7BpW,CAA6B,QAA7BA,CAA6B;QAA1BmS,IAA0B,QAA1BA,IAA0B;QAApB9B,KAAoB,QAApBA,KAAoB;QAAb7Q,GAAa,QAAbA,GAAa;;wBACzB,KAAK+W,WAAjB,EAAiCH,IAAjC;;QAEIpW,KAAKmC,QAAQC,IAAR,CAAa+P,IAAb,CAAT;;;;QAII9R,OAAO,KAAKmW,cAAL,CAAoBxW,CAApB,EAAuBqQ,KAAvB,EAA8B7Q,GAA9B,EAAmC4W,IAAnC,CAAX;;QAEI5D,iBAAiBnS,IAAjB,CAAJ,EAA4B;aACnB,KAAKoW,kBAAL,CAAwBpW,IAAxB,EAA8BL,CAA9B,CAAP;;;;;;;;;;wCAKgB,iBAAgBoW,IAAhB,EAAsB3U,MAAtB,CAA6B;eAAK2U,KAAKM,CAAL,MAAY,IAAjB;OAA7B,CAAlB,4GAAuE;YAA5D9N,GAA4D;;aAChEA,GAAL,IAAY,KAAZ;YACIzG,QAAQC,IAAR,CAAa+P,IAAb,CAAJ;;eAEO,KAAKqE,cAAL,CAAoBxW,CAApB,EAAuBqQ,KAAvB,EAA8B7Q,GAA9B,EAAmC4W,IAAnC,CAAP;;YAEI5D,iBAAiBnS,IAAjB,CAAJ,EAA4B;;;;;;;;;;;;;;;;;;;WAKvB,KAAKoW,kBAAL,CAAwBpW,IAAxB,EAA8BL,CAA9B,CAAP;GApD4B;;;;gBAAA,0BAwDfA,CAxDe,EAwDZqQ,KAxDY,EAwDL7Q,GAxDK,EAwDA4W,IAxDA,EAwDM;WAC3BF,iBACGC,gBAAgBnW,CAAhB,EAAmBoW,IAAnB,CADH,EAEL;UAAA;0BAEsBA,KAAK1B,kBAF3B;kBAAA;;KAFK,CAAP;GAzD4B;;;;;;oBAAA,8BAsEXrU,IAtEW,EAsELL,CAtEK,EAsEF;QACtB,CAACK,IAAL,EAAW;aACF,IAAP;;;WAGKmM,gBAAgBxM,EAAEmS,IAAF,CAAO9R,IAAP,CAAhB,CAAP;;;;;;;CA3EJ,CAqFA;;AC7FA;;;;;;;AAOA,AAAO,IAAMsW,yBAAyB,CACpC,iBADoC,EAEpC,UAFoC,EAGpC,SAHoC,EAIpC,UAJoC,EAKpC,OALoC,CAA/B;;;;AAUP,AAAO,IAAMC,uBAAuB,CAClC,UADkC,CAA7B;;;;;;;;;AAWP,AAAO,IAAMC,yBAAyB,CACpC,sBADoC,EAEpC,kBAFoC,EAGpC,kBAHoC,EAIpC,YAJoC,EAKpC,mBALoC,EAMpC,cANoC,CAA/B;;AASP,AAAO,IAAMC,uBAAuB,CAClC,YADkC,EAElC,cAFkC,EAGlC,cAHkC,EAIlC,aAJkC,EAKlC,aALkC,EAMlC,aANkC,EAOlC,aAPkC,EAQlC,eARkC,EASlC,eATkC,EAUlC,iBAVkC,EAWlC,UAXkC,EAYlC,YAZkC,EAalC,IAbkC,EAclC,iBAdkC,EAelC,OAfkC,CAA7B;;ACxBP,IAAMC,wBAAwB;SAAA,yBACG;QAArB/W,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbwX,SAAa,QAAbA,SAAa;;;;QAGzB3G,cAAJ;;YAEQa,gBAAgBlR,CAAhB,EAAmB2W,sBAAnB,EAA2CK,SAA3C,CAAR;QACI3G,KAAJ,EAAW,OAAOuE,WAAWvE,KAAX,EAAkB,EAAE7Q,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;;YAIHgS,qBAAqBhS,CAArB,EAAwB6W,sBAAxB,CAAR;QACIxG,KAAJ,EAAW,OAAOuE,WAAWvE,KAAX,EAAkB,EAAE7Q,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;YAGHkR,gBAAgBlR,CAAhB,EAAmB4W,oBAAnB,EAAyCI,SAAzC,CAAR;QACI3G,KAAJ,EAAW,OAAOuE,WAAWvE,KAAX,EAAkB,EAAE7Q,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;YAGHgS,qBAAqBhS,CAArB,EAAwB8W,oBAAxB,CAAR;QACIzG,KAAJ,EAAW,OAAOuE,WAAWvE,KAAX,EAAkB,EAAE7Q,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;WAGJ,EAAP;;CAvBJ,CA2BA;;ACxCA;;;;;;AAMA,AAAO,IAAMiX,mBAAmB,CAC9B,KAD8B,EAE9B,OAF8B,EAG9B,WAH8B,EAI9B,eAJ8B,EAK9B,YAL8B,EAM9B,WAN8B,EAO9B,SAP8B,CAAzB;;AAUP,AAAO,IAAMC,oBAAoB,GAA1B;;;;;;;;;AASP,AAAO,IAAMC,mBAAmB,CAC9B,sBAD8B,EAE9B,mBAF8B,EAG9B,oBAH8B,EAI9B,mBAJ8B,EAK9B,oBAL8B,EAM9B,qBAN8B,EAO9B,aAP8B,EAQ9B,iBAR8B,EAS9B,oBAT8B,EAU9B,qBAV8B,EAW9B,eAX8B,EAY9B,YAZ8B,EAa9B,YAb8B,EAc9B,cAd8B,EAe9B,cAf8B,EAgB9B,yBAhB8B,EAiB9B,qBAjB8B,EAkB9B,qBAlB8B,EAmB9B,SAnB8B,EAoB9B,SApB8B,EAqB9B,gBArB8B,EAsB9B,gBAtB8B,EAuB9B,SAvB8B,CAAzB;;;;AA4BP,IAAMC,WAAW,aAAjB;AACA,AAAO,IAAMC,sBAAsB,CACjC,CAAC,SAAD,EAAYD,QAAZ,CADiC,EAEjC,CAAC,SAAD,EAAYA,QAAZ,CAFiC,CAA5B;;ACzCP,IAAME,yBAAyB;SAAA,yBACH;QAAhBtX,CAAgB,QAAhBA,CAAgB;QAAbgX,SAAa,QAAbA,SAAa;;QACpBvD,eAAJ;;;;aAISvC,gBAAgBlR,CAAhB,EAAmBiX,gBAAnB,EAAqCD,SAArC,CAAT;QACIvD,UAAUA,OAAOlR,MAAP,GAAgB2U,iBAA9B,EAAiD;aACxC1D,YAAYC,MAAZ,CAAP;;;;aAIOzB,qBAAqBhS,CAArB,EAAwBmX,gBAAxB,EAA0C,CAA1C,CAAT;QACI1D,UAAUA,OAAOlR,MAAP,GAAgB2U,iBAA9B,EAAiD;aACxC1D,YAAYC,MAAZ,CAAP;;;;;;;;;;wCAK8B4D,mBAAhC,4GAAqD;;;;;YAAzClS,QAAyC;YAA/BoS,KAA+B;;YAC7ClX,OAAOL,EAAEmF,QAAF,CAAb;YACI9E,KAAKkC,MAAL,KAAgB,CAApB,EAAuB;cACf0H,OAAO5J,KAAK4J,IAAL,EAAb;cACIsN,MAAMhY,IAAN,CAAW0K,IAAX,CAAJ,EAAsB;mBACbuJ,YAAYvJ,IAAZ,CAAP;;;;;;;;;;;;;;;;;;;WAKC,IAAP;;CA7BJ,CAiCA;;AC9CA;;;;AAIA,AAAO,IAAMuN,2BAA2B,CACtC,wBADsC,EAEtC,aAFsC,EAGtC,SAHsC,EAItC,gBAJsC,EAKtC,WALsC,EAMtC,cANsC,EAOtC,UAPsC,EAQtC,UARsC,EAStC,SATsC,EAUtC,eAVsC,EAWtC,UAXsC,EAYtC,cAZsC,EAatC,qBAbsC,EActC,cAdsC,EAetC,SAfsC,EAgBtC,MAhBsC,CAAjC;;;;;AAsBP,AAAO,IAAMC,2BAA2B,CACtC,4BADsC,EAEtC,oBAFsC,EAGtC,0BAHsC,EAItC,kBAJsC,EAKtC,oBALsC,EAMtC,kBANsC,EAOtC,iBAPsC,EAQtC,aARsC,EAStC,eATsC,EAUtC,qBAVsC,EAWtC,mBAXsC,EAYtC,cAZsC,EAatC,aAbsC,EActC,YAdsC,EAetC,kBAfsC,EAgBtC,WAhBsC,EAiBtC,UAjBsC,CAAjC;;;;;AAuBP,IAAMC,kBAAkB,mDAAxB;AACA,AAAO,IAAMC,yBAAyB;;AAEpC,IAAIvZ,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAFoC;;;;AAMpC,IAAIA,MAAJ,CAAW,6BAAX,EAA0C,GAA1C,CANoC;;AAQpC,IAAIA,MAAJ,iBAAyBsZ,eAAzB,kBAAuD,GAAvD,CARoC,CAA/B;;ACrCP,IAAME,gCAAgC;SAAA,yBACL;QAArB5X,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbwX,SAAa,QAAbA,SAAa;;QACzBa,sBAAJ;;;;oBAIgB3G,gBAAgBlR,CAAhB,EAAmBwX,wBAAnB,EAA6CR,SAA7C,EAAwD,KAAxD,CAAhB;QACIa,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;;oBAIH7F,qBAAqBhS,CAArB,EAAwByX,wBAAxB,CAAhB;QACII,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;oBAGHpL,eAAejN,GAAf,EAAoBmY,sBAApB,CAAhB;QACIE,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;WAEZ,IAAP;;CAlBJ,CAsBA;;ACnCA;;;;;;;;;;;;;;;;;AAiBA,IAAMC,sBAAsB;;SAAA,qBAEhB;WACD,IAAP;;CAHJ,CAOA;;ACxBA;;;AAGA,AAAO,IAAMC,2BAA2B,CACtC,UADsC,EAEtC,eAFsC,EAGtC,WAHsC,CAAjC;;AAMP,AAAO,IAAMC,2BAA2B,CACtC,qBADsC,CAAjC;;AAIP,AAAO,IAAMC,gCAAgC,CAC3C,QAD2C,EAE3C,YAF2C,EAG3C,OAH2C,EAI3C,OAJ2C,EAK3C,UAL2C,CAAtC;AAOP,AAAO,IAAMC,mCAAmC,IAAI9Z,MAAJ,CAAW6Z,8BAA8B5Z,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAM8Z,gCAAgC,CAC3C,QAD2C,EAE3C,QAF2C,EAG3C,OAH2C,EAI3C,UAJ2C,EAK3C,UAL2C,EAM3C,MAN2C,EAO3C,IAP2C,EAQ3C,YAR2C,EAS3C,MAT2C,EAU3C,QAV2C,EAW3C,QAX2C,EAY3C,KAZ2C,EAa3C,QAb2C,EAc3C,SAd2C,EAe3C,QAf2C,EAgB3C,SAhB2C,EAiB3C,SAjB2C,EAkB3C,QAlB2C,EAmB3C,OAnB2C,EAoB3C,UApB2C,EAqB3C,SArB2C,EAsB3C,OAtB2C,EAuB3C,OAvB2C,EAwB3C,KAxB2C,EAyB3C,aAzB2C,CAAtC;AA2BP,AAAO,IAAMC,mCAAmC,IAAIha,MAAJ,CAAW+Z,8BAA8B9Z,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAMga,SAAS,gBAAf;AACP,AAAO,IAAMC,SAAS,kBAAf;;AC3CP,SAASC,MAAT,CAAgBjY,KAAhB,EAAuB;UACXA,MAAME,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCF,MAAME,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;;AAIF,AAAO,SAASgY,aAAT,CAAuBhZ,GAAvB,EAA4B;QAC3BA,IAAI0K,IAAJ,EAAN;MACIO,QAAQ,CAAZ;;MAEIyN,iCAAiC3Y,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;MAGE4Y,iCAAiC7Y,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;;;MAKE6Y,OAAO9Y,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;MAGE8Y,OAAO/Y,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;;;SAKKiL,KAAP;;;;AAIF,AAAO,SAASgO,SAAT,CAAmB/T,IAAnB,EAAyB;MAC1BA,KAAKlE,IAAL,CAAU,KAAV,CAAJ,EAAsB;WACb,CAAP;;;SAGK,CAAP;;;;;AAKF,AAAO,SAASkY,cAAT,CAAwBhU,IAAxB,EAA8B;MAC/B+F,QAAQ,CAAZ;MACMkO,aAAajU,KAAK5B,OAAL,CAAa,QAAb,EAAuB0M,KAAvB,EAAnB;;MAEImJ,WAAWpW,MAAX,KAAsB,CAA1B,EAA6B;aAClB,EAAT;;;MAGIM,UAAU6B,KAAKoH,MAAL,EAAhB;MACI8M,iBAAJ;MACI/V,QAAQN,MAAR,KAAmB,CAAvB,EAA0B;eACbM,QAAQiJ,MAAR,EAAX;;;GAGDjJ,OAAD,EAAU+V,QAAV,EAAoB3X,OAApB,CAA4B,UAACX,KAAD,EAAW;QACjC6F,iBAAe5G,IAAf,CAAoBgZ,OAAOjY,KAAP,CAApB,CAAJ,EAAwC;eAC7B,EAAT;;GAFJ;;SAMOmK,KAAP;;;;;AAKF,AAAO,SAASoO,cAAT,CAAwBnU,IAAxB,EAA8B;MAC/B+F,QAAQ,CAAZ;MACMsE,WAAWrK,KAAKiD,IAAL,EAAjB;MACMG,UAAUiH,SAASxQ,GAAT,CAAa,CAAb,CAAhB;;MAEIuJ,WAAWA,QAAQpF,OAAR,KAAoB,YAAnC,EAAiD;aACtC,EAAT;;;MAGEyD,iBAAe5G,IAAf,CAAoBgZ,OAAOxJ,QAAP,CAApB,CAAJ,EAA2C;aAChC,EAAT;;;SAGKtE,KAAP;;;AAGF,AAAO,SAASqO,iBAAT,CAA2BpU,IAA3B,EAAiC;MAClC+F,QAAQ,CAAZ;;MAEMlH,QAAQoH,WAAWjG,KAAKlE,IAAL,CAAU,OAAV,CAAX,CAAd;MACMsI,SAAS6B,WAAWjG,KAAKlE,IAAL,CAAU,QAAV,CAAX,CAAf;MACM8C,MAAMoB,KAAKlE,IAAL,CAAU,KAAV,CAAZ;;;MAGI+C,SAASA,SAAS,EAAtB,EAA0B;aACf,EAAT;;;;MAIEuF,UAAUA,UAAU,EAAxB,EAA4B;aACjB,EAAT;;;MAGEvF,SAASuF,MAAT,IAAmB,CAACxF,IAAIpB,QAAJ,CAAa,QAAb,CAAxB,EAAgD;QACxC6W,OAAOxV,QAAQuF,MAArB;QACIiQ,OAAO,IAAX,EAAiB;;eACN,GAAT;KADF,MAEO;eACI5N,KAAK6N,KAAL,CAAWD,OAAO,IAAlB,CAAT;;;;SAIGtO,KAAP;;;AAGF,AAAO,SAASwO,eAAT,CAAyBC,KAAzB,EAAgC/X,KAAhC,EAAuC;SACpC+X,MAAM3W,MAAN,GAAe,CAAhB,GAAqBpB,KAA5B;;;AC1GF;;;;;;;;AAQA,IAAMgY,+BAA+B;SAAA,yBACA;QAAzBnZ,CAAyB,QAAzBA,CAAyB;QAAtBiC,OAAsB,QAAtBA,OAAsB;QAAb+U,SAAa,QAAbA,SAAa;;QAC7BoC,iBAAJ;;;;;;QAMMC,WACJnI,gBACElR,CADF,EAEE+X,wBAFF,EAGEf,SAHF,EAIE,KAJF,CADF;;QAQIqC,QAAJ,EAAc;iBACDpD,QAAWoD,QAAX,CAAX;;UAEID,QAAJ,EAAc,OAAOA,QAAP;;;;;;QAMVE,OAAOtZ,EAAE,KAAF,EAASiC,OAAT,EAAkByP,OAAlB,EAAb;QACM6H,YAAY,EAAlB;;SAEKtY,OAAL,CAAa,UAACF,GAAD,EAAMI,KAAN,EAAgB;UACrBuD,OAAO1E,EAAEe,GAAF,CAAb;UACMuC,MAAMoB,KAAKlE,IAAL,CAAU,KAAV,CAAZ;;UAEI,CAAC8C,GAAL,EAAU;;UAENmH,QAAQ+N,cAAclV,GAAd,CAAZ;eACSmV,UAAU/T,IAAV,CAAT;eACSgU,eAAehU,IAAf,CAAT;eACSmU,eAAenU,IAAf,CAAT;eACSoU,kBAAkBpU,IAAlB,CAAT;eACSuU,gBAAgBK,IAAhB,EAAsBnY,KAAtB,CAAT;;gBAEUmC,GAAV,IAAiBmH,KAAjB;KAbF;;gCAiBE,iBAAgB8O,SAAhB,EAA2B3P,MAA3B,CAAkC,UAACC,GAAD,EAAMjB,GAAN;aAChC2Q,UAAU3Q,GAAV,IAAiBiB,IAAI,CAAJ,CAAjB,GAA0B,CAACjB,GAAD,EAAM2Q,UAAU3Q,GAAV,CAAN,CAA1B,GAAkDiB,GADlB;KAAlC,EAEE,CAAC,IAAD,EAAO,CAAP,CAFF,CA5C+B;;;;QA2C1B2P,MA3C0B;QA2ClB5K,QA3CkB;;;QAgD7BA,WAAW,CAAf,EAAkB;iBACLqH,QAAWuD,MAAX,CAAX;;UAEIJ,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;wCAKOpB,wBAAvB,4GAAiD;YAAtC7S,QAAsC;;YACzC7E,QAAQN,EAAEmF,QAAF,EAAYqK,KAAZ,EAAd;YACMlM,MAAMhD,MAAME,IAAN,CAAW,KAAX,CAAZ;YACI8C,GAAJ,EAAS;qBACI2S,QAAW3S,GAAX,CAAX;cACI8V,QAAJ,EAAc,OAAOA,QAAP;;;YAGVK,OAAOnZ,MAAME,IAAN,CAAW,MAAX,CAAb;YACIiZ,IAAJ,EAAU;qBACGxD,QAAWwD,IAAX,CAAX;cACIL,QAAJ,EAAc,OAAOA,QAAP;;;YAGV7Y,QAAQD,MAAME,IAAN,CAAW,OAAX,CAAd;YACID,KAAJ,EAAW;qBACE0V,QAAW1V,KAAX,CAAX;cACI6Y,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;;;;;;;;;WAIX,IAAP;;CA9EJ,CAkFA;;ACzGe,SAASM,eAAT,CAAyBjP,KAAzB,EAAgCkP,UAAhC,EAA4CF,IAA5C,EAAkD;;;;;;MAM3DhP,QAAQ,CAAZ,EAAe;QACPmP,aAAa,IAAIC,QAAQC,eAAZ,CAA4B,IAA5B,EAAkCH,UAAlC,EAA8CF,IAA9C,EAAoDM,KAApD,EAAnB;;;;;;;QAOMC,cAAc,MAAMJ,UAA1B;QACMK,eAAe,EAAE,OAAOD,cAAc,GAArB,CAAF,CAArB;WACOvP,QAAQwP,YAAf;;;SAGK,CAAP;;;ACnBa,SAASC,aAAT,CAAuBlJ,QAAvB,EAAiC5D,OAAjC,EAA0C;;;;;MAKnD3C,QAAQ,CAAZ;;MAEIwC,YAAY1N,IAAZ,CAAiByR,SAAS9G,IAAT,EAAjB,CAAJ,EAAuC;QAC/BiQ,gBAAgBpR,SAASiI,QAAT,EAAmB,EAAnB,CAAtB;;;;QAIImJ,gBAAgB,CAApB,EAAuB;cACb,CAAC,EAAT;KADF,MAEO;cACGhP,KAAKE,GAAL,CAAS,CAAT,EAAY,KAAK8O,aAAjB,CAAR;;;;;;QAME/M,WAAWA,WAAW+M,aAA1B,EAAyC;eAC9B,EAAT;;;;SAIG1P,KAAP;;;AC5Ba,SAAS2P,eAAT,CAAyBhN,OAAzB,EAAkCiN,IAAlC,EAAwC;;;;MAIjDjN,WAAW,CAACiN,IAAhB,EAAsB;WACb,EAAP;;;SAGK,CAAP;;;ACRK,IAAMC,aAAW,IAAjB;;;;AAIP,AAAO,IAAM7T,0BAAwB,CACnC,OADmC,EAEnC,SAFmC,EAGnC,SAHmC,EAInC,SAJmC,EAKnC,QALmC,EAMnC,OANmC,EAOnC,OAPmC,EAQnC,OARmC,EASnC,KATmC,EAUnC,OAVmC,EAWnC,MAXmC,EAYnC,QAZmC,EAanC,KAbmC,EAcnC,iBAdmC,CAA9B;AAgBP,AAAO,IAAMC,6BAA2B,IAAItI,MAAJ,CAAWqI,wBAAsBpI,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;;;AAKP,AAAO,IAAMkc,sBAAoB,IAAInc,MAAJ,CAAW,4CAAX,EAAyD,GAAzD,CAA1B;;;;AAIP,AAAO,IAAMoc,qBAAmB,IAAIpc,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAzB;;;;AAIP,AAAO,IAAMqc,sBAAoB,IAAIrc,MAAJ,CAAW,yBAAX,EAAsC,GAAtC,CAA1B,CAEP;;ACjCe,SAASsc,oBAAT,CAA8BjB,IAA9B,EAAoC;;MAE7C/S,2BAAyBnH,IAAzB,CAA8Bka,IAA9B,CAAJ,EAAyC;WAChC,CAAC,EAAR;;;SAGK,CAAP;;;ACAF,SAASkB,SAAT,CAAiBC,KAAjB,EAAwB;UACZA,MAAMpa,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCoa,MAAMpa,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;AAGF,AAAe,SAASkY,gBAAT,CAAwBkC,KAAxB,EAA+B;;;;MAIxC/X,UAAU+X,MAAM9O,MAAN,EAAd;MACI+O,gBAAgB,KAApB;MACIC,gBAAgB,KAApB;MACIrQ,QAAQ,CAAZ;;cAEW/M,MAAM,CAAN,EAAS,CAAT,CAAX,EAAwBuD,OAAxB,CAAgC,YAAM;QAChC4B,QAAQN,MAAR,KAAmB,CAAvB,EAA0B;;;;QAIpBwY,aAAaJ,UAAQ9X,OAAR,EAAiB,GAAjB,CAAnB;;;;QAII,CAACgY,aAAD,IAAkBlU,QAAQpH,IAAR,CAAawb,UAAb,CAAtB,EAAgD;sBAC9B,IAAhB;eACS,EAAT;;;;;;QAME,CAACD,aAAD,IAAkBvU,kBAAkBhH,IAAlB,CAAuBwb,UAAvB,CAAlB,IACErU,2BAAyBnH,IAAzB,CAA8Bwb,UAA9B,CADN,EACiD;UAC3C,CAAC1U,kBAAkB9G,IAAlB,CAAuBwb,UAAvB,CAAL,EAAyC;wBACvB,IAAhB;iBACS,EAAT;;;;cAIMlY,QAAQiJ,MAAR,EAAV;GAzBF;;SA4BOrB,KAAP;;;AC/Ca,SAASuQ,aAAT,CAAuBC,QAAvB,EAAiC;;;MAG1CR,oBAAkBlb,IAAlB,CAAuB0b,QAAvB,CAAJ,EAAsC;WAC7B,CAAC,GAAR;;;SAGK,CAAP;;;ACFa,SAASC,WAAT,CACbzB,IADa,EAEbE,UAFa,EAGbwB,OAHa,EAIb1b,SAJa,EAKbuR,QALa,EAMboK,YANa,EAOb;;MAEIA,aAAa7Z,IAAb,CAAkB;WAAOkY,SAASja,GAAhB;GAAlB,MAA2C+S,SAA/C,EAA0D;WACjD,KAAP;;;;;MAKE,CAACkH,IAAD,IAASA,SAASE,UAAlB,IAAgCF,SAAS0B,OAA7C,EAAsD;WAC7C,KAAP;;;MAGMrd,QAZR,GAYqB2B,SAZrB,CAYQ3B,QAZR;;mBAa+B4B,IAAIC,KAAJ,CAAU8Z,IAAV,CAb/B;;MAakB4B,QAblB,cAaQvd,QAbR;;;;MAgBIud,aAAavd,QAAjB,EAA2B;WAClB,KAAP;;;;;MAKIwd,WAAW7B,KAAKjW,OAAL,CAAa2X,OAAb,EAAsB,EAAtB,CAAjB;MACI,CAACb,WAAS/a,IAAT,CAAc+b,QAAd,CAAL,EAA8B;WACrB,KAAP;;;;;MAKE5U,2BAAyBnH,IAAzB,CAA8ByR,QAA9B,CAAJ,EAA6C;WACpC,KAAP;;;;MAIEA,SAASzO,MAAT,GAAkB,EAAtB,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;ACpDa,SAASgZ,YAAT,CAAsB9B,IAAtB,EAA4B+B,SAA5B,EAAuC;;;;;MAKhD,CAACA,UAAUjc,IAAV,CAAeka,IAAf,CAAL,EAA2B;WAClB,CAAC,EAAR;;;SAGK,CAAP;;;ACPa,SAASgC,iBAAT,CAA2BR,QAA3B,EAAqC;;MAE9CV,oBAAkBhb,IAAlB,CAAuB0b,QAAvB,CAAJ,EAAsC;WAC7B,EAAP;;;SAGK,CAAP;;;ACHa,SAASS,aAAT,CAAuBT,QAAvB,EAAiC;;MAE1CT,mBAAiBjb,IAAjB,CAAsB0b,QAAtB,CAAJ,EAAqC;;;;;QAK/BV,oBAAkBhb,IAAlB,CAAuB0b,QAAvB,CAAJ,EAAsC;aAC7B,CAAC,EAAR;;;;SAIG,CAAP;;;ACIK,SAASU,aAAT,CAAuBR,OAAvB,EAAgC;SAC9B,IAAI/c,MAAJ,OAAe+c,OAAf,EAA0B,GAA1B,CAAP;;;AAGF,SAASR,OAAT,CAAiBC,KAAjB,EAAwB5J,QAAxB,EAAkC;UACtBA,YAAY4J,MAAM3Q,IAAN,EAAtB,WAAsC2Q,MAAMpa,IAAN,CAAW,OAAX,KAAuB,EAA7D,WAAmEoa,MAAMpa,IAAN,CAAW,IAAX,KAAoB,EAAvF;;;AAGF,AAAe,SAASob,UAAT,OAOZ;MANDC,KAMC,QANDA,KAMC;MALDlC,UAKC,QALDA,UAKC;MAJDwB,OAIC,QAJDA,OAIC;MAHD1b,SAGC,QAHDA,SAGC;MAFDO,CAEC,QAFDA,CAEC;+BADDob,YACC;MADDA,YACC,qCADc,EACd;;cACW3b,aAAaC,IAAIC,KAAJ,CAAUga,UAAV,CAAzB;MACM6B,YAAYG,cAAcR,OAAd,CAAlB;MACMd,OAAO5H,YAAYzS,CAAZ,CAAb;;;;;;;;;MASM8b,cAAcD,MAAMjS,MAAN,CAAa,UAACmS,aAAD,EAAgBC,IAAhB,EAAyB;;;;QAIlDvC,OAAOpM,aAAa2O,KAAKhb,OAAL,CAAayY,IAA1B,CAAb;QACMmB,QAAQ5a,EAAEgc,IAAF,CAAd;QACMhL,WAAW4J,MAAM3Q,IAAN,EAAjB;;QAEI,CAACiR,YAAYzB,IAAZ,EAAkBE,UAAlB,EAA8BwB,OAA9B,EAAuC1b,SAAvC,EAAkDuR,QAAlD,EAA4DoK,YAA5D,CAAL,EAAgF;aACvEW,aAAP;;;;QAIE,CAACA,cAActC,IAAd,CAAL,EAA0B;oBACVA,IAAd,IAAsB;eACb,CADa;0BAAA;;OAAtB;KADF,MAMO;oBACSA,IAAd,EAAoBzI,QAApB,GAAkC+K,cAActC,IAAd,EAAoBzI,QAAtD,SAAkEA,QAAlE;;;QAGIiL,eAAeF,cAActC,IAAd,CAArB;QACMwB,WAAWN,QAAQC,KAAR,EAAe5J,QAAf,CAAjB;QACM5D,UAAUF,eAAeuM,IAAf,CAAhB;;QAEIhP,QAAQ8Q,aAAa9B,IAAb,EAAmB+B,SAAnB,CAAZ;aACSC,kBAAkBR,QAAlB,CAAT;aACSS,cAAcT,QAAd,CAAT;aACSD,cAAcC,QAAd,CAAT;aACSvC,iBAAekC,KAAf,CAAT;aACSF,qBAAqBjB,IAArB,CAAT;aACSW,gBAAgBhN,OAAhB,EAAyBiN,IAAzB,CAAT;aACSH,cAAclJ,QAAd,EAAwB5D,OAAxB,CAAT;aACSsM,gBAAgBjP,KAAhB,EAAuBkP,UAAvB,EAAmCF,IAAnC,CAAT;;iBAEahP,KAAb,GAAqBA,KAArB;;WAEOsR,aAAP;GAvCkB,EAwCjB,EAxCiB,CAApB;;SA0CO,iBAAgBD,WAAhB,EAA6BvZ,MAA7B,KAAwC,CAAxC,GAA4C,IAA5C,GAAmDuZ,WAA1D;;;AClFF;;AAEA,IAAMI,8BAA8B;SAAA,yBACgB;QAAxClc,CAAwC,QAAxCA,CAAwC;QAArCR,GAAqC,QAArCA,GAAqC;QAAhCC,SAAgC,QAAhCA,SAAgC;iCAArB2b,YAAqB;QAArBA,YAAqB,qCAAN,EAAM;;gBACpC3b,aAAaC,IAAIC,KAAJ,CAAUH,GAAV,CAAzB;;QAEMma,aAAatM,aAAa7N,GAAb,CAAnB;QACM2b,UAAUxN,eAAenO,GAAf,EAAoBC,SAApB,CAAhB;;QAEMoc,QAAQ7b,EAAE,SAAF,EAAa0R,OAAb,EAAd;;QAEMyK,cAAcP,WAAW;kBAAA;4BAAA;sBAAA;0BAAA;UAAA;;KAAX,CAApB;;;QAUI,CAACO,WAAL,EAAkB,OAAO,IAAP;;;;QAIZC,UAAU,iBAAgBD,WAAhB,EAA6BvS,MAA7B,CAAoC,UAACC,GAAD,EAAMmS,IAAN,EAAe;UAC3DK,aAAaF,YAAYH,IAAZ,CAAnB;aACOK,WAAW5R,KAAX,GAAmBZ,IAAIY,KAAvB,GAA+B4R,UAA/B,GAA4CxS,GAAnD;KAFc,EAGb,EAAEY,OAAO,CAAC,GAAV,EAHa,CAAhB;;;;QAOI2R,QAAQ3R,KAAR,IAAiB,EAArB,EAAyB;aAChB2R,QAAQ3C,IAAf;;;WAGK,IAAP;;CAlCJ,CAuCA;;ACjDO,IAAM6C,2BAA2B,CACtC,QADsC,CAAjC;;ACKP,SAASC,WAAT,CAAqB/c,GAArB,EAA0B;MAClBC,YAAYC,IAAIC,KAAJ,CAAUH,GAAV,CAAlB;MACQ1B,QAFgB,GAEH2B,SAFG,CAEhB3B,QAFgB;;SAGjBA,QAAP;;;AAGF,SAASiE,MAAT,CAAgBvC,GAAhB,EAAqB;SACZ;YAAA;YAEG+c,YAAY/c,GAAZ;GAFV;;;AAMF,IAAMgd,sBAAsB;SAAA,yBACK;QAArBxc,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbwX,SAAa,QAAbA,SAAa;;QACvByF,aAAazc,EAAE,qBAAF,CAAnB;QACIyc,WAAWla,MAAX,KAAsB,CAA1B,EAA6B;UACrBkX,OAAOgD,WAAWjc,IAAX,CAAgB,MAAhB,CAAb;UACIiZ,IAAJ,EAAU;eACD1X,OAAO0X,IAAP,CAAP;;;;QAIEiD,UAAUxL,gBAAgBlR,CAAhB,EAAmBsc,wBAAnB,EAA6CtF,SAA7C,CAAhB;QACI0F,OAAJ,EAAa;aACJ3a,OAAO2a,OAAP,CAAP;;;WAGK3a,OAAOvC,GAAP,CAAP;;CAfJ,CAoBA;;ACtCO,IAAMmd,yBAAyB,CACpC,gBADoC,EAEpC,qBAFoC,CAA/B;;ACSA,SAAShb,OAAT,CAAeM,OAAf,EAAwBjC,CAAxB,EAA4C;MAAjB4c,SAAiB,uEAAL,GAAK;;YACvC3a,QAAQuB,OAAR,CAAgB,UAAhB,EAA4B,GAA5B,EAAiC0G,IAAjC,EAAV;SACO2S,UAAU5a,OAAV,EAAmB2a,SAAnB,EAA8B,EAAEE,SAAS,UAAX,EAA9B,CAAP;;;AAGF,IAAMC,0BAA0B;SAAA,yBACK;QAAzB/c,CAAyB,QAAzBA,CAAyB;QAAtBiC,OAAsB,QAAtBA,OAAsB;QAAb+U,SAAa,QAAbA,SAAa;;QAC3BjD,UAAU7C,gBAAgBlR,CAAhB,EAAmB2c,sBAAnB,EAA2C3F,SAA3C,CAAhB;QACIjD,OAAJ,EAAa;aACJpS,QAAMiQ,UAAUmC,OAAV,EAAmB/T,CAAnB,CAAN,CAAP;;;QAGI4c,YAAY,GAAlB;QACMI,eAAe/a,QAAQsJ,KAAR,CAAc,CAAd,EAAiBqR,YAAY,CAA7B,CAArB;WACOjb,QAAM3B,EAAEgd,YAAF,EAAgB/S,IAAhB,EAAN,EAA8BjK,CAA9B,EAAiC4c,SAAjC,CAAP;;CATJ,CAaA;;ACvBA,IAAMK,4BAA4B;SAAA,yBACX;QAAXhb,OAAW,QAAXA,OAAW;;QACbjC,IAAImC,QAAQC,IAAR,CAAaH,OAAb,CAAV;;QAEMgI,OAAOuC,gBAAgBxM,EAAE,KAAF,EAASwP,KAAT,GAAiBvF,IAAjB,EAAhB,CAAb;WACOA,KAAK7F,KAAL,CAAW,IAAX,EAAiB7B,MAAxB;;CALJ,CASA;;ACCA,IAAM2a,mBAAmB;;UAEf,GAFe;SAGhBnG,sBAAsBoG,OAHN;kBAIPvF,8BAA8BuF,OAJvB;UAKf7F,uBAAuB6F,OALR;WAMd7G,wBAAwB6G,OAAxB,CAAgCC,IAAhC,CAAqC9G,uBAArC,CANc;kBAOP6C,6BAA6BgE,OAPtB;OAQlBrF,oBAAoBqF,OARF;iBASRjB,4BAA4BiB,OATpB;kBAUPX,oBAAoBW,OAVb;WAWdJ,wBAAwBI,OAXV;cAYXF,0BAA0BE,OAZf;aAaZ;QAAG9M,KAAH,QAAGA,KAAH;WAAegN,gBAAgBC,YAAhB,CAA6BjN,KAA7B,CAAf;GAbY;;SAAA,mBAef7R,OAfe,EAeN;QACP2T,IADO,GACE3T,OADF,CACP2T,IADO;;;QAGXA,IAAJ,EAAU;UACFnS,IAAImC,QAAQC,IAAR,CAAa+P,IAAb,CAAV;cACQnS,CAAR,GAAYA,CAAZ;;;QAGIqQ,QAAQ,KAAKA,KAAL,CAAW7R,OAAX,CAAd;QACM+e,iBAAiB,KAAKA,cAAL,CAAoB/e,OAApB,CAAvB;QACMiV,SAAS,KAAKA,MAAL,CAAYjV,OAAZ,CAAf;QACMyD,UAAU,KAAKA,OAAL,cAAkBzD,OAAlB,IAA2B6R,YAA3B,IAAhB;QACMmN,iBAAiB,KAAKA,cAAL,cAAyBhf,OAAzB,IAAkCyD,gBAAlC,IAAvB;QACM6R,MAAM,KAAKA,GAAL,cAActV,OAAd,IAAuByD,gBAAvB,IAAZ;QACMwb,gBAAgB,KAAKA,aAAL,CAAmBjf,OAAnB,CAAtB;QACMuV,UAAU,KAAKA,OAAL,cAAkBvV,OAAlB,IAA2ByD,gBAA3B,IAAhB;QACMyb,aAAa,KAAKA,UAAL,cAAqBlf,OAArB,IAA8ByD,gBAA9B,IAAnB;QACM0b,YAAY,KAAKA,SAAL,CAAe,EAAEtN,YAAF,EAAf,CAAlB;;0BACwB,KAAKuN,cAAL,CAAoBpf,OAApB,CAlBT;;QAkBPgB,GAlBO,mBAkBPA,GAlBO;QAkBFqe,MAlBE,mBAkBFA,MAlBE;;;WAoBR;kBAAA;oBAAA;sBAGWN,kBAAkB,IAH7B;cAAA;oCAAA;sBAAA;kCAAA;cAAA;oBAAA;sBAAA;4BAAA;;KAAP;;CAnCJ,CAoDA;;AC7De,SAASO,YAAT,CAAsBte,GAAtB,EAA2BC,SAA3B,EAAsC;cACvCA,aAAaC,IAAIC,KAAJ,CAAUH,GAAV,CAAzB;mBACqBC,SAF8B;MAE3C3B,QAF2C,cAE3CA,QAF2C;;MAG7CigB,aAAajgB,SAASsG,KAAT,CAAe,GAAf,EAAoBmH,KAApB,CAA0B,CAAC,CAA3B,EAA8BlN,IAA9B,CAAmC,GAAnC,CAAnB;;SAEOsG,WAAW7G,QAAX,KAAwB6G,WAAWoZ,UAAX,CAAxB,IAAkDb,gBAAzD;;;ACNF;AACA,AAAO,SAASc,gBAAT,CAA0BpN,QAA1B,EAAoC5Q,CAApC,QAAkD;MAAT2B,KAAS,QAATA,KAAS;;MACnD,CAACA,KAAL,EAAY,OAAOiP,QAAP;;IAEVjP,MAAMtD,IAAN,CAAW,GAAX,CAAF,EAAmBuS,QAAnB,EAA6BlP,MAA7B;;SAEOkP,QAAP;;;;AAIF,AAAO,SAASqN,iBAAT,CAA2BrN,QAA3B,EAAqC5Q,CAArC,SAAwD;MAAdke,UAAc,SAAdA,UAAc;;MACzD,CAACA,UAAL,EAAiB,OAAOtN,QAAP;;mBAEDsN,UAAhB,EAA4Bjd,OAA5B,CAAoC,UAAC2H,GAAD,EAAS;QACrCuV,WAAWne,EAAE4I,GAAF,EAAOgI,QAAP,CAAjB;QACMrQ,QAAQ2d,WAAWtV,GAAX,CAAd;;;QAGI,OAAOrI,KAAP,KAAiB,QAArB,EAA+B;eACpBJ,IAAT,CAAc,UAACgB,KAAD,EAAQd,IAAR,EAAiB;sBACfL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0Bke,WAAWtV,GAAX,CAA1B;OADF;KADF,MAIO,IAAI,OAAOrI,KAAP,KAAiB,UAArB,EAAiC;;eAE7BJ,IAAT,CAAc,UAACgB,KAAD,EAAQd,IAAR,EAAiB;YACvB0B,SAASxB,MAAMP,EAAEK,IAAF,CAAN,EAAeL,CAAf,CAAf;;YAEI,OAAO+B,MAAP,KAAkB,QAAtB,EAAgC;wBAChB/B,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0B+B,MAA1B;;OAJJ;;GAXJ;;SAqBO6O,QAAP;;;AAGF,SAASwN,oBAAT,CAA8Bpe,CAA9B,EAAiCiS,SAAjC,EAA4C;SACnCA,UAAU1Q,IAAV,CAAe,UAAC4D,QAAD,EAAc;QAC9BkZ,MAAMC,OAAN,CAAcnZ,QAAd,CAAJ,EAA6B;qCACTA,QADS;;UACpBoZ,CADoB;UACjB/d,IADiB;;aAEpBR,EAAEue,CAAF,EAAKhc,MAAL,KAAgB,CAAhB,IAAqBvC,EAAEue,CAAF,EAAK/d,IAAL,CAAUA,IAAV,CAArB,IAAwCR,EAAEue,CAAF,EAAK/d,IAAL,CAAUA,IAAV,EAAgB0J,IAAhB,OAA2B,EAA1E;;;WAGKlK,EAAEmF,QAAF,EAAY5C,MAAZ,KAAuB,CAAvB,IAA4BvC,EAAEmF,QAAF,EAAY8E,IAAZ,GAAmBC,IAAnB,OAA8B,EAAjE;GANK,CAAP;;;AAUF,AAAO,SAASsU,MAAT,CAAgBpI,IAAhB,EAAsB;MACnBpW,CADmB,GAC8BoW,IAD9B,CACnBpW,CADmB;MAChBoB,IADgB,GAC8BgV,IAD9B,CAChBhV,IADgB;MACVqd,cADU,GAC8BrI,IAD9B,CACVqI,cADU;0BAC8BrI,IAD9B,CACMsI,WADN;MACMA,WADN,qCACoB,KADpB;;;MAGvB,CAACD,cAAL,EAAqB,OAAO,IAAP;;;;MAIjB,OAAOA,cAAP,KAA0B,QAA9B,EAAwC,OAAOA,cAAP;;MAEhCxM,SATmB,GASkBwM,cATlB,CASnBxM,SATmB;8BASkBwM,cATlB,CASR9J,cATQ;MASRA,cATQ,yCASS,IATT;;;MAWrBgK,mBAAmBP,qBAAqBpe,CAArB,EAAwBiS,SAAxB,CAAzB;;MAEI,CAAC0M,gBAAL,EAAuB,OAAO,IAAP;;;;;;;;MAQnBD,WAAJ,EAAiB;QACX9N,WAAW5Q,EAAE2e,gBAAF,CAAf;;;aAGSC,IAAT,CAAc5e,EAAE,aAAF,CAAd;eACW4Q,SAAS9E,MAAT,EAAX;;eAEWmS,kBAAkBrN,QAAlB,EAA4B5Q,CAA5B,EAA+Bye,cAA/B,CAAX;eACWT,iBAAiBpN,QAAjB,EAA2B5Q,CAA3B,EAA8Bye,cAA9B,CAAX;;eAEWzI,SAAS5U,IAAT,EAAewP,QAAf,eAA8BwF,IAA9B,IAAoCzB,8BAApC,IAAX;;WAEO3U,EAAEmS,IAAF,CAAOvB,QAAP,CAAP;;;MAGE7O,eAAJ;;;;MAIIsc,MAAMC,OAAN,CAAcK,gBAAd,CAAJ,EAAqC;2CACVA,gBADU;;QAC5BxZ,QAD4B;QAClB3E,IADkB;;aAE1BR,EAAEmF,QAAF,EAAY3E,IAAZ,CAAiBA,IAAjB,EAAuB0J,IAAvB,EAAT;GAFF,MAGO;aACIlK,EAAE2e,gBAAF,EAAoB1U,IAApB,GAA2BC,IAA3B,EAAT;;;;;MAKEyK,cAAJ,EAAoB;WACXqB,SAAS5U,IAAT,EAAeW,MAAf,EAAuBqU,IAAvB,CAAP;;;SAGKrU,MAAP;;;AAGF,SAAS8c,aAAT,CAAuBzI,IAAvB,EAA6B;MACnBhV,IADmB,GACkBgV,IADlB,CACnBhV,IADmB;MACb0d,SADa,GACkB1I,IADlB,CACb0I,SADa;uBACkB1I,IADlB,CACF2I,QADE;MACFA,QADE,kCACS,IADT;;;MAGrBhd,SAASyc,oBAAYpI,IAAZ,IAAkBqI,gBAAgBK,UAAU1d,IAAV,CAAlC,IAAf;;;MAGIW,MAAJ,EAAY;WACHA,MAAP;;;;;MAKEgd,QAAJ,EAAc,OAAO7B,iBAAiB9b,IAAjB,EAAuBgV,IAAvB,CAAP;;SAEP,IAAP;;;AAGF,IAAM4I,gBAAgB;SAAA,qBACwB;QAApCF,SAAoC,uEAAxB5B,gBAAwB;QAAN9G,IAAM;gBACFA,IADE;QAClC6I,WADkC,SAClCA,WADkC;QACrBC,cADqB,SACrBA,cADqB;;;QAGtCJ,UAAUjB,MAAV,KAAqB,GAAzB,EAA8B,OAAOiB,UAAU3B,OAAV,CAAkB/G,IAAlB,CAAP;;wBAGzBA,IADL;;;;QAKI6I,WAAJ,EAAiB;UACThd,WAAU4c,2BACXzI,IADW,IACLhV,MAAM,SADD,EACYsd,aAAa,IADzB,EAC+BrO,OAAO6O;SADtD;aAGO;;OAAP;;QAII7O,QAAQwO,2BAAmBzI,IAAnB,IAAyBhV,MAAM,OAA/B,IAAd;QACMmc,iBAAiBsB,2BAAmBzI,IAAnB,IAAyBhV,MAAM,gBAA/B,IAAvB;QACMqS,SAASoL,2BAAmBzI,IAAnB,IAAyBhV,MAAM,QAA/B,IAAf;QACMqc,gBAAgBoB,2BAAmBzI,IAAnB,IAAyBhV,MAAM,eAA/B,IAAtB;QACMa,UAAU4c,2BACXzI,IADW,IACLhV,MAAM,SADD,EACYsd,aAAa,IADzB,EAC+BrO;OAD/C;QAGMmN,iBAAiBqB,2BAAmBzI,IAAnB,IAAyBhV,MAAM,gBAA/B,EAAiDa,gBAAjD,IAAvB;QACM8R,UAAU8K,2BAAmBzI,IAAnB,IAAyBhV,MAAM,SAA/B,EAA0Ca,gBAA1C,IAAhB;QACM6R,MAAM+K,2BAAmBzI,IAAnB,IAAyBhV,MAAM,KAA/B,EAAsCa,gBAAtC,EAA+C8R,gBAA/C,IAAZ;QACM2J,aAAamB,2BAAmBzI,IAAnB,IAAyBhV,MAAM,YAA/B,EAA6Ca,gBAA7C,IAAnB;QACM0b,YAAYkB,2BAAmBzI,IAAnB,IAAyBhV,MAAM,WAA/B,EAA4CiP,YAA5C,IAAlB;;gBAEEwO,2BAAmBzI,IAAnB,IAAyBhV,MAAM,gBAA/B,QAAsD,EAAE5B,KAAK,IAAP,EAAaqe,QAAQ,IAArB,EA/Bd;;QA8BlCre,GA9BkC,SA8BlCA,GA9BkC;QA8B7Bqe,MA9B6B,SA8B7BA,MA9B6B;;;WAiCnC;kBAAA;sBAAA;oBAAA;oCAAA;oCAAA;cAAA;kCAAA;cAAA;oBAAA;sBAAA;4BAAA;;KAAP;;CAlCJ,CAmDA;;AC3KA;wDAAe;QAEXJ,aAFW,SAEXA,aAFW;QAGXtL,IAHW,SAGXA,IAHW;QAIXnS,CAJW,SAIXA,CAJW;QAKXgX,SALW,SAKXA,SALW;QAMXjV,MANW,SAMXA,MANW;QAOXod,SAPW,SAOXA,SAPW;QAQX9O,KARW,SAQXA,KARW;QASX7Q,GATW,SASXA,GATW;;;;;;;iBAAA,GAaD,CAbC;wBAAA,GAcQ,CAAC6N,aAAa7N,GAAb,CAAD,CAdR;;;;;;kBAkBNie,iBAAiB2B,QAAQ,EAlBnB;;;;;qBAmBF,CAAT;;mBACUxd,SAASyd,MAAT,CAAgB5B,aAAhB,CApBC;;;aAAA;;mBAqBJzd,EAAEmS,IAAF,EAAP;;yBArBW,GAuBW;mBACfsL,aADe;wBAAA;kBAAA;kCAAA;2BAKP,IALO;8BAMJpN,KANI;;aAvBX;0BAAA,GAiCY2O,cAAc7B,OAAd,CAAsBgC,SAAtB,EAAiCG,aAAjC,CAjCZ;;;yBAmCEjR,IAAb,CAAkBoP,aAAlB;kCAEK1b,MADL;sCAGMA,OAAOE,OADX,yCAGamd,KAHb,uBAIIG,eAAetd,OAJnB;;;4BAQcsd,eAAe9B,aAA/B;;;;;sBA9CW,GAiDMP,iBAAiBQ,UAAjB,CAA4B,EAAEzb,mBAAiBF,OAAOE,OAAxB,WAAF,EAA5B,CAjDN;0DAmDRF,MAnDQ;2BAoDEqd,KApDF;8BAqDKA,KArDL;;;;;;;;;;GAAf;;WAA8BI,eAA9B;;;;SAA8BA,eAA9B;;;ACMA,IAAMC,UAAU;OAAA,iBACFjgB,GADE,EACG2S,IADH,EACoB;;;QAAXiE,IAAW,uEAAJ,EAAI;;;;;;;;oCAI5BA,IAJ4B,CAE9BsJ,aAF8B;2BAAA,uCAEd,IAFc;+BAI5BtJ,IAJ4B,CAG9B2I,QAH8B;sBAAA,kCAGnB,IAHmB;uBAAA,GAMdrf,IAAIC,KAAJ,CAAUH,GAAV,CANc;;kBAQ3B3B,YAAY4B,SAAZ,CAR2B;;;;;+CASvB1B,OAAO8B,MATgB;;;uBAAA,GAYdie,aAAate,GAAb,EAAkBC,SAAlB,CAZc;;;;qBAehBmC,SAASyd,MAAT,CAAgB7f,GAAhB,EAAqB2S,IAArB,EAA2B1S,SAA3B,CAfgB;;;eAAA;;mBAkB5BO,EAAEb,KAlB0B;;;;;+CAmBvBa,CAnBuB;;;;qBAsBzBA,EAAEmS,IAAF,EAAP;;;;uBAtBgC,GA0BdnS,EAAE,MAAF,EAAUkF,GAAV,CAAc,UAAC9E,CAAD,EAAIC,IAAJ;uBAAaL,EAAEK,IAAF,EAAQG,IAAR,CAAa,MAAb,CAAb;eAAd,EAAiDkR,OAAjD,EA1Bc;oBAAA,GA4BnBsN,cAAc7B,OAAd,CAAsBgC,SAAtB,EAAiC,EAAE3f,QAAF,EAAO2S,UAAP,EAAanS,IAAb,EAAgBgX,oBAAhB,EAA2BvX,oBAA3B,EAAsCsf,kBAAtC,EAAjC,CA5BmB;wBA6BChd,MA7BD;mBAAA,WA6BxBsO,KA7BwB;2BAAA,WA6BjBoN,aA7BiB;;;;oBAgC5BiC,iBAAiBjC,aAhCW;;;;;;qBAiCf+B,gBACb;oCAAA;4CAAA;0BAAA;oBAAA;oCAAA;8BAAA;4BAAA;;eADa,CAjCe;;;oBAAA;;;;;oCA+CzBzd,MADL;6BAEe,CAFf;gCAGkB;;;;+CAIbA,MArDyB;;;;;;;;;GADpB;;;;;eAAA,yBA2DMvC,GA3DN,EA2DW;;;;;;;;;qBACVoC,SAASyd,MAAT,CAAgB7f,GAAhB,CADU;;;;;;;;;;;;;CA3D3B,CAiEA;;"}
\ No newline at end of file
+{"version":3,"file":null,"sources":["../src/utils/range.js","../src/utils/validate-url.js","../src/utils/errors.js","../src/resource/utils/constants.js","../src/resource/utils/fetch-resource.js","../src/resource/utils/dom/normalize-meta-tags.js","../src/resource/utils/dom/constants.js","../src/resource/utils/dom/convert-lazy-loaded-images.js","../src/resource/utils/dom/clean.js","../src/resource/index.js","../src/extractors/custom/nymag.com/index.js","../src/extractors/custom/blogspot.com/index.js","../src/extractors/custom/wikipedia.org/index.js","../src/extractors/custom/twitter.com/index.js","../src/extractors/custom/www.nytimes.com/index.js","../src/extractors/custom/www.theatlantic.com/index.js","../src/extractors/custom/www.newyorker.com/index.js","../src/extractors/custom/www.wired.com/index.js","../src/extractors/custom/www.msn.com/index.js","../src/extractors/custom/www.yahoo.com/index.js","../src/extractors/custom/www.buzzfeed.com/index.js","../src/extractors/custom/fandom.wikia.com/index.js","../src/extractors/custom/www.littlethings.com/index.js","../src/extractors/custom/www.politico.com/index.js","../src/extractors/custom/deadspin.com/index.js","../src/extractors/custom/www.broadwayworld.com/index.js","../src/extractors/custom/www.apartmenttherapy.com/index.js","../src/extractors/all.js","../src/utils/dom/constants.js","../src/utils/dom/strip-unlikely-candidates.js","../src/utils/dom/brs-to-ps.js","../src/utils/dom/paragraphize.js","../src/utils/dom/convert-to-paragraphs.js","../src/utils/dom/convert-node-to.js","../src/utils/dom/clean-images.js","../src/utils/dom/mark-to-keep.js","../src/utils/dom/strip-junk-tags.js","../src/utils/dom/clean-h-ones.js","../src/utils/dom/clean-attributes.js","../src/utils/dom/remove-empty.js","../src/extractors/generic/content/scoring/constants.js","../src/extractors/generic/content/scoring/get-weight.js","../src/extractors/generic/content/scoring/get-score.js","../src/extractors/generic/content/scoring/score-commas.js","../src/extractors/generic/content/scoring/score-length.js","../src/extractors/generic/content/scoring/score-paragraph.js","../src/extractors/generic/content/scoring/set-score.js","../src/extractors/generic/content/scoring/add-score.js","../src/extractors/generic/content/scoring/add-to-parent.js","../src/extractors/generic/content/scoring/get-or-init-score.js","../src/extractors/generic/content/scoring/score-node.js","../src/extractors/generic/content/scoring/score-content.js","../src/utils/text/normalize-spaces.js","../src/utils/text/extract-from-url.js","../src/utils/text/constants.js","../src/utils/text/page-num-from-url.js","../src/utils/text/remove-anchor.js","../src/utils/text/article-base-url.js","../src/utils/text/has-sentence-end.js","../src/utils/text/excerpt-content.js","../src/extractors/generic/content/scoring/merge-siblings.js","../src/extractors/generic/content/scoring/find-top-candidate.js","../src/utils/dom/clean-tags.js","../src/utils/dom/clean-headers.js","../src/utils/dom/rewrite-top-level.js","../src/utils/dom/make-links-absolute.js","../src/utils/dom/link-density.js","../src/utils/dom/extract-from-meta.js","../src/utils/dom/extract-from-selectors.js","../src/utils/dom/strip-tags.js","../src/utils/dom/within-comment.js","../src/utils/dom/node-is-sufficient.js","../src/utils/dom/is-wordpress.js","../src/cleaners/constants.js","../src/cleaners/author.js","../src/cleaners/lead-image-url.js","../src/cleaners/dek.js","../src/cleaners/date-published.js","../src/cleaners/content.js","../src/cleaners/title.js","../src/cleaners/resolve-split-title.js","../src/cleaners/index.js","../src/extractors/generic/content/extract-best-node.js","../src/extractors/generic/content/extractor.js","../src/extractors/generic/title/constants.js","../src/extractors/generic/title/extractor.js","../src/extractors/generic/author/constants.js","../src/extractors/generic/author/extractor.js","../src/extractors/generic/date-published/constants.js","../src/extractors/generic/date-published/extractor.js","../src/extractors/generic/dek/extractor.js","../src/extractors/generic/lead-image-url/constants.js","../src/extractors/generic/lead-image-url/score-image.js","../src/extractors/generic/lead-image-url/extractor.js","../src/extractors/generic/next-page-url/scoring/utils/score-similarity.js","../src/extractors/generic/next-page-url/scoring/utils/score-link-text.js","../src/extractors/generic/next-page-url/scoring/utils/score-page-in-link.js","../src/extractors/generic/next-page-url/scoring/constants.js","../src/extractors/generic/next-page-url/scoring/utils/score-extraneous-links.js","../src/extractors/generic/next-page-url/scoring/utils/score-by-parents.js","../src/extractors/generic/next-page-url/scoring/utils/score-prev-link.js","../src/extractors/generic/next-page-url/scoring/utils/should-score.js","../src/extractors/generic/next-page-url/scoring/utils/score-base-url.js","../src/extractors/generic/next-page-url/scoring/utils/score-next-link-text.js","../src/extractors/generic/next-page-url/scoring/utils/score-cap-links.js","../src/extractors/generic/next-page-url/scoring/score-links.js","../src/extractors/generic/next-page-url/extractor.js","../src/extractors/generic/url/constants.js","../src/extractors/generic/url/extractor.js","../src/extractors/generic/excerpt/constants.js","../src/extractors/generic/excerpt/extractor.js","../src/extractors/generic/word-count/extractor.js","../src/extractors/generic/index.js","../src/extractors/get-extractor.js","../src/extractors/root-extractor.js","../src/extractors/collect-all-pages.js","../src/mercury.js"],"sourcesContent":["export default function* range(start = 1, end = 1) {\n while (start <= end) {\n yield start += 1;\n }\n}\n","// extremely simple url validation as a first step\nexport default function validateUrl({ hostname }) {\n // If this isn't a valid url, return an error message\n return !!hostname;\n}\n","const Errors = {\n badUrl: {\n error: true,\n messages: 'The url parameter passed does not look like a valid URL. Please check your data and try again.',\n },\n};\n\nexport default Errors;\n","export const REQUEST_HEADERS = {\n 'User-Agent': 'Readability - http://readability.com/about/',\n};\n\n// The number of milliseconds to attempt to fetch a resource before timing out.\nexport const FETCH_TIMEOUT = 10000;\n\n// Content types that we do not extract content from\nconst BAD_CONTENT_TYPES = [\n 'audio/mpeg',\n 'image/gif',\n 'image/jpeg',\n 'image/jpg',\n];\n\nexport const BAD_CONTENT_TYPES_RE = new RegExp(`^(${BAD_CONTENT_TYPES.join('|')})$`, 'i');\n\n\n// Use this setting as the maximum size an article can be\n// for us to attempt parsing. Defaults to 5 MB.\nexport const MAX_CONTENT_LENGTH = 5242880;\n\n// Turn the global proxy on or off\n// Proxying is not currently enabled in Python source\n// so not implementing logic in port.\nexport const PROXY_DOMAINS = false;\nexport const REQUESTS_PROXIES = {\n http: 'http://38.98.105.139:33333',\n https: 'http://38.98.105.139:33333',\n};\n\nexport const DOMAINS_TO_PROXY = [\n 'nih.gov',\n 'gutenberg.org',\n];\n","import URL from 'url';\nimport request from 'request';\nimport { Errors } from 'utils';\n\nimport {\n REQUEST_HEADERS,\n FETCH_TIMEOUT,\n BAD_CONTENT_TYPES_RE,\n MAX_CONTENT_LENGTH,\n} from './constants';\n\nfunction get(options) {\n return new Promise((resolve, reject) => {\n request(options, (err, response, body) => {\n if (err) {\n reject(err);\n } else {\n resolve({ body, response });\n }\n });\n });\n}\n\n// Evaluate a response to ensure it's something we should be keeping.\n// This does not validate in the sense of a response being 200 level or\n// not. Validation here means that we haven't found reason to bail from\n// further processing of this url.\n\nexport function validateResponse(response, parseNon2xx = false) {\n // Check if we got a valid status code\n if (response.statusMessage !== 'OK') {\n if (!response.statusCode) {\n throw new Error(\n `Unable to fetch content. Original exception was ${response.error}`\n );\n } else if (!parseNon2xx) {\n throw new Error(\n `Resource returned a response status code of ${response.statusCode} and resource was instructed to reject non-2xx level status codes.`\n );\n }\n }\n\n const {\n 'content-type': contentType,\n 'content-length': contentLength,\n } = response.headers;\n\n // Check that the content is not in BAD_CONTENT_TYPES\n if (BAD_CONTENT_TYPES_RE.test(contentType)) {\n throw new Error(\n `Content-type for this resource was ${contentType} and is not allowed.`\n );\n }\n\n // Check that the content length is below maximum\n if (contentLength > MAX_CONTENT_LENGTH) {\n throw new Error(\n `Content for this resource was too large. Maximum content length is ${MAX_CONTENT_LENGTH}.`\n );\n }\n\n return true;\n}\n\n// Grabs the last two pieces of the URL and joins them back together\n// This is to get the 'livejournal.com' from 'erotictrains.livejournal.com'\nexport function baseDomain({ host }) {\n return host.split('.').slice(-2).join('.');\n}\n\n// Set our response attribute to the result of fetching our URL.\n// TODO: This should gracefully handle timeouts and raise the\n// proper exceptions on the many failure cases of HTTP.\n// TODO: Ensure we are not fetching something enormous. Always return\n// unicode content for HTML, with charset conversion.\n\nexport default async function fetchResource(url, parsedUrl) {\n parsedUrl = parsedUrl || URL.parse(encodeURI(url));\n\n const options = {\n url: parsedUrl,\n headers: { ...REQUEST_HEADERS },\n timeout: FETCH_TIMEOUT,\n // Don't set encoding; fixes issues\n // w/gzipped responses\n encoding: null,\n // Accept cookies\n jar: true,\n // Accept and decode gzip\n gzip: true,\n // Follow any redirect\n followAllRedirects: true,\n };\n\n const { response, body } = await get(options);\n\n try {\n validateResponse(response);\n return { body, response };\n } catch (e) {\n return Errors.badUrl;\n }\n}\n","function convertMetaProp($, from, to) {\n $(`meta[${from}]`).each((_, node) => {\n const $node = $(node);\n\n const value = $node.attr(from);\n $node.attr(to, value);\n $node.removeAttr(from);\n });\n\n return $;\n}\n\n// For ease of use in extracting from meta tags,\n// replace the \"content\" attribute on meta tags with the\n// \"value\" attribute.\n//\n// In addition, normalize 'property' attributes to 'name' for ease of\n// querying later. See, e.g., og or twitter meta tags.\n\nexport default function normalizeMetaTags($) {\n $ = convertMetaProp($, 'content', 'value');\n $ = convertMetaProp($, 'property', 'name');\n return $;\n}\n","export const IS_LINK = new RegExp('https?://', 'i');\nexport const IS_IMAGE = new RegExp('.(png|gif|jpe?g)', 'i');\n\nexport const TAGS_TO_REMOVE = [\n 'script',\n 'style',\n 'form',\n].join(',');\n","import {\n IS_LINK,\n IS_IMAGE,\n} from './constants';\n\n// Convert all instances of images with potentially\n// lazy loaded images into normal images.\n// Many sites will have img tags with no source, or an image tag with a src\n// attribute that a is a placeholer. We need to be able to properly fill in\n// the src attribute so the images are no longer lazy loaded.\nexport default function convertLazyLoadedImages($) {\n $('img').each((_, img) => {\n Reflect.ownKeys(img.attribs).forEach((attr) => {\n const value = img.attribs[attr];\n\n if (attr !== 'src' && IS_LINK.test(value) &&\n IS_IMAGE.test(value)) {\n $(img).attr('src', value);\n }\n });\n });\n\n return $;\n}\n","import { TAGS_TO_REMOVE } from './constants';\n\nfunction isComment(index, node) {\n return node.type === 'comment';\n}\n\nfunction cleanComments($) {\n $.root().find('*')\n .contents()\n .filter(isComment)\n .remove();\n\n return $;\n}\n\nexport default function clean($) {\n $(TAGS_TO_REMOVE).remove();\n\n $ = cleanComments($);\n return $;\n}\n","import cheerio from 'cheerio';\n\nimport { fetchResource } from './utils';\nimport {\n normalizeMetaTags,\n convertLazyLoadedImages,\n clean,\n} from './utils/dom';\n\nconst Resource = {\n\n // Create a Resource.\n //\n // :param url: The URL for the document we should retrieve.\n // :param response: If set, use as the response rather than\n // attempting to fetch it ourselves. Expects a\n // string.\n async create(url, preparedResponse, parsedUrl) {\n let result;\n\n if (preparedResponse) {\n const validResponse = {\n statusMessage: 'OK',\n statusCode: 200,\n headers: {\n 'content-type': 'text/html',\n 'content-length': 500,\n },\n };\n\n result = { body: preparedResponse, response: validResponse };\n } else {\n result = await fetchResource(url, parsedUrl);\n }\n\n if (result.error) {\n return result;\n }\n\n return this.generateDoc(result);\n },\n\n generateDoc({ body: content, response }) {\n const { 'content-type': contentType } = response.headers;\n\n // TODO: Implement is_text function from\n // https://github.com/ReadabilityHoldings/readability/blob/8dc89613241d04741ebd42fa9fa7df1b1d746303/readability/utils/text.py#L57\n if (!contentType.includes('html') &&\n !contentType.includes('text')) {\n throw new Error('Content does not appear to be text.');\n }\n\n let $ = cheerio.load(content, { normalizeWhitespace: true });\n\n if ($.root().children().length === 0) {\n throw new Error('No children, likely a bad parse.');\n }\n\n $ = normalizeMetaTags($);\n $ = convertLazyLoadedImages($);\n $ = clean($);\n\n return $;\n },\n};\n\nexport default Resource;\n","export const NYMagExtractor = {\n domain: 'nymag.com',\n content: {\n // Order by most likely. Extractor will stop on first occurrence\n selectors: [\n 'div.article-content',\n 'section.body',\n 'article.article',\n ],\n\n // Selectors to remove from the extracted content\n clean: [\n '.ad',\n '.single-related-story',\n ],\n\n // Object of tranformations to make on matched elements\n // Each key is the selector, each value is the tag to\n // transform to.\n // If a function is given, it should return a string\n // to convert to or nothing (in which case it will not perform\n // the transformation.\n transforms: {\n // Convert h1s to h2s\n h1: 'h2',\n\n // Convert lazy-loaded noscript images to figures\n noscript: ($node) => {\n const $children = $node.children();\n if ($children.length === 1 && $children.get(0).tagName === 'img') {\n return 'figure';\n }\n\n return null;\n },\n },\n },\n\n title: {\n selectors: [\n 'h1.lede-feature-title',\n 'h1.headline-primary',\n 'h1',\n ],\n },\n\n author: {\n selectors: [\n '.by-authors',\n '.lede-feature-author',\n ],\n },\n\n dek: {\n selectors: [\n '.lede-feature-teaser',\n ],\n },\n\n date_published: {\n selectors: [\n ['time.article-timestamp[datetime]', 'datetime'],\n 'time.article-timestamp',\n ],\n },\n};\n","export const BloggerExtractor = {\n domain: 'blogspot.com',\n content: {\n // Blogger is insane and does not load its content\n // initially in the page, but it's all there\n // in noscript\n selectors: [\n '.post-content noscript',\n ],\n\n // Selectors to remove from the extracted content\n clean: [\n ],\n\n // Convert the noscript tag to a div\n transforms: {\n noscript: 'div',\n },\n },\n\n author: {\n selectors: [\n '.post-author-name',\n ],\n },\n\n title: {\n selectors: [\n '.post h2.title',\n ],\n },\n\n date_published: {\n selectors: [\n 'span.publishdate',\n ],\n },\n};\n","export const WikipediaExtractor = {\n domain: 'wikipedia.org',\n content: {\n selectors: [\n '#mw-content-text',\n ],\n\n defaultCleaner: false,\n\n // transform top infobox to an image with caption\n transforms: {\n '.infobox img': ($node) => {\n const $parent = $node.parents('.infobox');\n // Only prepend the first image in .infobox\n if ($parent.children('img').length === 0) {\n $parent.prepend($node);\n }\n },\n '.infobox caption': 'figcaption',\n '.infobox': 'figure',\n },\n\n // Selectors to remove from the extracted content\n clean: [\n '.mw-editsection',\n 'figure tr, figure td, figure tbody',\n '#toc',\n '.navbox',\n ],\n\n },\n\n author: 'Wikipedia Contributors',\n\n title: {\n selectors: [\n 'h2.title',\n ],\n },\n\n date_published: {\n selectors: [\n '#footer-info-lastmod',\n ],\n },\n\n};\n","export const TwitterExtractor = {\n domain: 'twitter.com',\n\n content: {\n transforms: {\n // We're transforming essentially the whole page here.\n // Twitter doesn't have nice selectors, so our initial\n // selector grabs the whole page, then we're re-writing\n // it to fit our needs before we clean it up.\n '.permalink[role=main]': ($node, $) => {\n const tweets = $node.find('.tweet');\n const $tweetContainer = $('<div id=\"TWEETS_GO_HERE\"></div>');\n $tweetContainer.append(tweets);\n $node.replaceWith($tweetContainer);\n },\n\n // Twitter wraps @ with s, which\n // renders as a strikethrough\n s: 'span',\n },\n\n selectors: [\n '.permalink[role=main]',\n ],\n\n defaultCleaner: false,\n\n clean: [\n '.stream-item-footer',\n 'button',\n '.tweet-details-fixer',\n ],\n },\n\n author: {\n selectors: [\n '.tweet.permalink-tweet .username',\n ],\n },\n\n date_published: {\n selectors: [\n ['.permalink-tweet ._timestamp[data-time-ms]', 'data-time-ms'],\n // '.tweet.permalink-tweet .metadata',\n ],\n },\n\n};\n","export const NYTimesExtractor = {\n title: {\n selectors: [\n '.g-headline',\n 'h1.headline',\n ],\n },\n\n author: {\n selectors: [\n ['meta[name=\"author\"]', 'value'],\n '.g-byline',\n '.byline',\n ],\n },\n\n content: {\n selectors: [\n 'div.g-blocks',\n 'article#story',\n ],\n\n defaultCleaner: false,\n\n transforms: {\n 'img.g-lazy': ($node) => {\n let src = $node.attr('src');\n // const widths = $node.attr('data-widths')\n // .slice(1)\n // .slice(0, -1)\n // .split(',');\n // if (widths.length) {\n // width = widths.slice(-1);\n // } else {\n // width = '900';\n // }\n const width = 640;\n\n src = src.replace('{{size}}', width);\n $node.attr('src', src);\n },\n },\n\n clean: [\n '.ad',\n 'header#story-header',\n '.story-body-1 .lede.video',\n '.visually-hidden',\n '#newsletter-promo',\n '.promo',\n '.comments-button',\n '.hidden',\n ],\n },\n\n date_published: null,\n\n lead_image_url: null,\n\n dek: null,\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\nexport const TheAtlanticExtractor = {\n domain: 'www.theatlantic.com',\n title: {\n selectors: [\n 'h1.hed',\n ],\n },\n\n author: {\n selectors: [\n 'article#article .article-cover-extra .metadata .byline a',\n ],\n },\n\n content: {\n selectors: [\n '.article-body',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: null,\n\n lead_image_url: null,\n\n dek: null,\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const NewYorkerExtractor = {\n domain: 'www.newyorker.com',\n title: {\n selectors: [\n 'h1.title',\n ],\n },\n\n author: {\n selectors: [\n '.contributors',\n ],\n },\n\n content: {\n selectors: [\n 'div#articleBody',\n 'div.articleBody',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: {\n selectors: [\n ['meta[name=\"article:published_time\"]', 'value'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const WiredExtractor = {\n domain: 'www.wired.com',\n title: {\n selectors: [\n 'h1.post-title',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n 'a[rel=\"author\"]',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n 'article.content',\n // enter content selectors\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n '.visually-hidden',\n\n ],\n },\n\n date_published: {\n selectors: [\n ['meta[itemprop=\"datePublished\"]', 'value'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const MSNExtractor = {\n domain: 'www.msn.com',\n title: {\n selectors: [\n 'h1',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n 'span.authorname-txt',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n 'div.richtext',\n // enter content selectors\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n 'span.caption',\n\n ],\n },\n\n date_published: {\n selectors: [\n 'span.time',\n ],\n },\n\n lead_image_url: {\n selectors: [\n\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const YahooExtractor = {\n domain: 'www.yahoo.com',\n title: {\n selectors: [\n 'header.canvas-header',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n 'span.provider-name',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n // enter content selectors\n '.content-canvas',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n '.figure-caption',\n\n ],\n },\n\n date_published: {\n selectors: [\n ['time.date[datetime]', 'datetime'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n // enter dek selectors\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const BuzzfeedExtractor = {\n domain: 'www.buzzfeed.com',\n title: {\n selectors: [\n 'h1[id=\"post-title\"]',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n 'a[data-action=\"user/username\"]', 'byline__author',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n '#buzz_sub_buzz',\n // enter content selectors\n ],\n\n defaultCleaner: false,\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: {\n h2: 'b',\n },\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n '.instapaper_ignore',\n '.suplist_list_hide .buzz_superlist_item .buzz_superlist_number_inline',\n '.share-box',\n ],\n },\n\n date_published: {\n selectors: [\n '.buzz-datetime',\n // enter author selectors\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const WikiaExtractor = {\n domain: 'fandom.wikia.com',\n title: {\n selectors: [\n 'h1.entry-title',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n '.author vcard', '.fn',\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n '.grid-content',\n '.entry-content',\n // enter content selectors\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: {\n selectors: [\n ['meta[name=\"article:published_time\"]', 'value'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const LittleThingsExtractor = {\n domain: 'www.littlethings.com',\n title: {\n selectors: [\n 'h1.post-title',\n // enter title selectors\n ],\n },\n\n author: {\n selectors: [\n ['meta[name=\"author\"]', 'value'],\n // enter author selectors\n ],\n },\n\n content: {\n selectors: [\n // enter content selectors\n '.mainContentIntro',\n '.content-wrapper',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const PoliticoExtractor = {\n domain: 'www.politico.com',\n title: {\n selectors: [\n // enter title selectors\n ['meta[name=\"og:title\"]', 'value'],\n ],\n },\n\n author: {\n selectors: [\n '.story-main-content .byline .vcard',\n ],\n },\n\n content: {\n selectors: [\n // enter content selectors\n '.story-main-content',\n '.content-group', '.story-core',\n '.story-text',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: [\n ],\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n 'figcaption',\n ],\n },\n\n date_published: {\n selectors: [\n ['.story-main-content .timestamp time[datetime]', 'datetime'],\n\n ],\n },\n\n lead_image_url: {\n selectors: [\n // enter lead_image_url selectors\n ['meta[name=\"og:image\"]', 'value'],\n\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"description\"]', 'value'],\n ],\n },\n\n next_page_url: null,\n\n excerpt: null,\n};\n","export const DeadspinExtractor = {\n domain: 'deadspin.com',\n title: {\n selectors: [\n 'h1.headline',\n ],\n },\n\n author: {\n selectors: [\n '.author',\n ],\n },\n\n content: {\n selectors: [\n '.post-content',\n '.entry-content',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: {\n 'iframe.lazyload[data-recommend-id^=\"youtube://\"]': ($node) => {\n const youtubeId = $node.attr('id').split('youtube-')[1];\n $node.attr('src', `https://www.youtube.com/embed/${youtubeId}`);\n },\n },\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n ],\n },\n\n date_published: {\n selectors: [\n ['time.updated[datetime]', 'datetime'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n // enter selectors\n ],\n },\n\n next_page_url: {\n selectors: [\n // enter selectors\n ],\n },\n\n excerpt: {\n selectors: [\n // enter selectors\n ],\n },\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const BroadwayWorldExtractor = {\n domain: 'www.broadwayworld.com',\n title: {\n selectors: [\n 'h1.article-title',\n ],\n },\n\n author: {\n selectors: [\n 'span[itemprop=author]',\n ],\n },\n\n content: {\n selectors: [\n 'div[itemprop=articlebody]',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: {\n },\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: {\n selectors: [\n ['meta[itemprop=datePublished]', 'value'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=\"og:description\"]', 'value'],\n ],\n },\n\n next_page_url: {\n selectors: [\n // enter selectors\n ],\n },\n\n excerpt: {\n selectors: [\n // enter selectors\n ],\n },\n};\n","// Rename CustomExtractor\n// to fit your publication\n// (e.g., NYTimesExtractor)\nexport const ApartmentTherapyExtractor = {\n domain: 'www.apartmenttherapy.com',\n title: {\n selectors: [\n 'h1.headline',\n ],\n },\n\n author: {\n selectors: [\n '.PostByline__name',\n ],\n },\n\n content: {\n selectors: [\n 'div.post__content',\n ],\n\n // Is there anything in the content you selected that needs transformed\n // before it's consumable content? E.g., unusual lazy loaded images\n transforms: {\n 'div[data-render-react-id=\"images/LazyPicture\"]': ($node, $) => {\n const data = JSON.parse($node.attr('data-props'));\n const { src } = data.sources[0];\n const $img = $('<img />').attr('src', src);\n $node.replaceWith($img);\n },\n },\n\n // Is there anything that is in the result that shouldn't be?\n // The clean selectors will remove anything that matches from\n // the result\n clean: [\n\n ],\n },\n\n date_published: {\n selectors: [\n ['.PostByline__timestamp[datetime]', 'datetime'],\n ],\n },\n\n lead_image_url: {\n selectors: [\n ['meta[name=\"og:image\"]', 'value'],\n ],\n },\n\n dek: {\n selectors: [\n ['meta[name=description]', 'value'],\n ],\n },\n\n next_page_url: {\n selectors: [\n // enter selectors\n ],\n },\n\n excerpt: {\n selectors: [\n // enter selectors\n ],\n },\n};\n","import { NYMagExtractor } from './custom/nymag.com';\nimport { BloggerExtractor } from './custom/blogspot.com';\nimport { WikipediaExtractor } from './custom/wikipedia.org';\nimport { TwitterExtractor } from './custom/twitter.com';\nimport { NYTimesExtractor } from './custom/www.nytimes.com';\nimport { TheAtlanticExtractor } from './custom/www.theatlantic.com';\nimport { NewYorkerExtractor } from './custom/www.newyorker.com';\nimport { WiredExtractor } from './custom/www.wired.com';\nimport { MSNExtractor } from './custom/www.msn.com';\nimport { YahooExtractor } from './custom/www.yahoo.com';\nimport { BuzzfeedExtractor } from './custom/www.buzzfeed.com';\nimport { WikiaExtractor } from './custom/fandom.wikia.com';\nimport { LittleThingsExtractor } from './custom/www.littlethings.com';\nimport { PoliticoExtractor } from './custom/www.politico.com';\nimport { DeadspinExtractor } from './custom/deadspin.com';\nimport { BroadwayWorldExtractor } from './custom/www.broadwayworld.com';\nimport { ApartmentTherapyExtractor } from './custom/www.apartmenttherapy.com';\n\nconst Extractors = {\n 'nymag.com': NYMagExtractor,\n 'blogspot.com': BloggerExtractor,\n 'wikipedia.org': WikipediaExtractor,\n 'twitter.com': TwitterExtractor,\n 'www.nytimes.com': NYTimesExtractor,\n 'www.theatlantic.com': TheAtlanticExtractor,\n 'www.newyorker.com': NewYorkerExtractor,\n 'www.wired.com': WiredExtractor,\n 'www.msn.com': MSNExtractor,\n 'www.yahoo.com': YahooExtractor,\n 'www.buzzfeed.com': BuzzfeedExtractor,\n 'fandom.wikia.com': WikiaExtractor,\n 'www.littlethings.com': LittleThingsExtractor,\n 'www.politico.com': PoliticoExtractor,\n 'deadspin.com': DeadspinExtractor,\n 'www.broadwayworld.com': BroadwayWorldExtractor,\n 'www.apartmenttherapy.com': ApartmentTherapyExtractor,\n};\n\nexport default Extractors;\n","// Spacer images to be removed\nexport const SPACER_RE = new RegExp('trans|transparent|spacer|blank', 'i');\n\n// The class we will use to mark elements we want to keep\n// but would normally remove\nexport const KEEP_CLASS = 'mercury-parser-keep';\n\nexport const KEEP_SELECTORS = [\n 'iframe[src^=\"https://www.youtube.com\"]',\n 'iframe[src^=\"http://www.youtube.com\"]',\n 'iframe[src^=\"https://player.vimeo\"]',\n 'iframe[src^=\"http://player.vimeo\"]',\n];\n\n// A list of tags to strip from the output if we encounter them.\nexport const STRIP_OUTPUT_TAGS = [\n 'title',\n 'script',\n 'noscript',\n 'link',\n 'style',\n 'hr',\n 'embed',\n 'iframe',\n 'object',\n];\n\n// cleanAttributes\nexport const REMOVE_ATTRS = ['style', 'align'];\nexport const REMOVE_ATTR_SELECTORS = REMOVE_ATTRS.map(selector => `[${selector}]`);\nexport const REMOVE_ATTR_LIST = REMOVE_ATTRS.join(',');\nexport const WHITELIST_ATTRS = ['src', 'srcset', 'href', 'class', 'id', 'alt'];\nexport const WHITELIST_ATTRS_RE = new RegExp(`^(${WHITELIST_ATTRS.join('|')})$`, 'i');\n\n// removeEmpty\nexport const REMOVE_EMPTY_TAGS = ['p'];\nexport const REMOVE_EMPTY_SELECTORS = REMOVE_EMPTY_TAGS.map(tag => `${tag}:empty`).join(',');\n\n// cleanTags\nexport const CLEAN_CONDITIONALLY_TAGS = ['ul', 'ol', 'table', 'div', 'button', 'form'].join(',');\n\n// cleanHeaders\nconst HEADER_TAGS = ['h2', 'h3', 'h4', 'h5', 'h6'];\nexport const HEADER_TAG_LIST = HEADER_TAGS.join(',');\n\n\n// // CONTENT FETCHING CONSTANTS ////\n\n// A list of strings that can be considered unlikely candidates when\n// extracting content from a resource. These strings are joined together\n// and then tested for existence using re:test, so may contain simple,\n// non-pipe style regular expression queries if necessary.\nexport const UNLIKELY_CANDIDATES_BLACKLIST = [\n 'ad-break',\n 'adbox',\n 'advert',\n 'addthis',\n 'agegate',\n 'aux',\n 'blogger-labels',\n 'combx',\n 'comment',\n 'conversation',\n 'disqus',\n 'entry-unrelated',\n 'extra',\n 'foot',\n // 'form', // This is too generic, has too many false positives\n 'header',\n 'hidden',\n 'loader',\n 'login', // Note: This can hit 'blogindex'.\n 'menu',\n 'meta',\n 'nav',\n 'outbrain',\n 'pager',\n 'pagination',\n 'predicta', // readwriteweb inline ad box\n 'presence_control_external', // lifehacker.com container full of false positives\n 'popup',\n 'printfriendly',\n 'related',\n 'remove',\n 'remark',\n 'rss',\n 'share',\n 'shoutbox',\n 'sidebar',\n 'sociable',\n 'sponsor',\n 'taboola',\n 'tools',\n];\n\n// A list of strings that can be considered LIKELY candidates when\n// extracting content from a resource. Essentially, the inverse of the\n// blacklist above - if something matches both blacklist and whitelist,\n// it is kept. This is useful, for example, if something has a className\n// of \"rss-content entry-content\". It matched 'rss', so it would normally\n// be removed, however, it's also the entry content, so it should be left\n// alone.\n//\n// These strings are joined together and then tested for existence using\n// re:test, so may contain simple, non-pipe style regular expression queries\n// if necessary.\nexport const UNLIKELY_CANDIDATES_WHITELIST = [\n 'and',\n 'article',\n 'body',\n 'blogindex',\n 'column',\n 'content',\n 'entry-content-asset',\n 'format', // misuse of form\n 'hfeed',\n 'hentry',\n 'hatom',\n 'main',\n 'page',\n 'posts',\n 'shadow',\n];\n\n// A list of tags which, if found inside, should cause a <div /> to NOT\n// be turned into a paragraph tag. Shallow div tags without these elements\n// should be turned into <p /> tags.\nexport const DIV_TO_P_BLOCK_TAGS = [\n 'a',\n 'blockquote',\n 'dl',\n 'div',\n 'img',\n 'p',\n 'pre',\n 'table',\n].join(',');\n\n// A list of tags that should be ignored when trying to find the top candidate\n// for a document.\nexport const NON_TOP_CANDIDATE_TAGS = [\n 'br',\n 'b',\n 'i',\n 'label',\n 'hr',\n 'area',\n 'base',\n 'basefont',\n 'input',\n 'img',\n 'link',\n 'meta',\n];\n\nexport const NON_TOP_CANDIDATE_TAGS_RE =\n new RegExp(`^(${NON_TOP_CANDIDATE_TAGS.join('|')})$`, 'i');\n\n// A list of selectors that specify, very clearly, either hNews or other\n// very content-specific style content, like Blogger templates.\n// More examples here: http://microformats.org/wiki/blog-post-formats\nexport const HNEWS_CONTENT_SELECTORS = [\n ['.hentry', '.entry-content'],\n ['entry', '.entry-content'],\n ['.entry', '.entry_content'],\n ['.post', '.postbody'],\n ['.post', '.post_body'],\n ['.post', '.post-body'],\n];\n\nexport const PHOTO_HINTS = [\n 'figure',\n 'photo',\n 'image',\n 'caption',\n];\nexport const PHOTO_HINTS_RE = new RegExp(PHOTO_HINTS.join('|'), 'i');\n\n\n// A list of strings that denote a positive scoring for this content as being\n// an article container. Checked against className and id.\n//\n// TODO: Perhaps have these scale based on their odds of being quality?\nexport const POSITIVE_SCORE_HINTS = [\n 'article',\n 'articlecontent',\n 'instapaper_body',\n 'blog',\n 'body',\n 'content',\n 'entry-content-asset',\n 'entry',\n 'hentry',\n 'main',\n 'Normal',\n 'page',\n 'pagination',\n 'permalink',\n 'post',\n 'story',\n 'text',\n '[-_]copy', // usatoday\n '\\\\Bcopy',\n];\n\n// The above list, joined into a matching regular expression\nexport const POSITIVE_SCORE_RE = new RegExp(POSITIVE_SCORE_HINTS.join('|'), 'i');\n\n// Readability publisher-specific guidelines\nexport const READABILITY_ASSET = new RegExp('entry-content-asset', 'i');\n\n// A list of strings that denote a negative scoring for this content as being\n// an article container. Checked against className and id.\n//\n// TODO: Perhaps have these scale based on their odds of being quality?\nexport const NEGATIVE_SCORE_HINTS = [\n 'adbox',\n 'advert',\n 'author',\n 'bio',\n 'bookmark',\n 'bottom',\n 'byline',\n 'clear',\n 'com-',\n 'combx',\n 'comment',\n 'comment\\\\B',\n 'contact',\n 'copy',\n 'credit',\n 'crumb',\n 'date',\n 'deck',\n 'excerpt',\n 'featured', // tnr.com has a featured_content which throws us off\n 'foot',\n 'footer',\n 'footnote',\n 'graf',\n 'head',\n 'info',\n 'infotext', // newscientist.com copyright\n 'instapaper_ignore',\n 'jump',\n 'linebreak',\n 'link',\n 'masthead',\n 'media',\n 'meta',\n 'modal',\n 'outbrain', // slate.com junk\n 'promo',\n 'pr_', // autoblog - press release\n 'related',\n 'respond',\n 'roundcontent', // lifehacker restricted content warning\n 'scroll',\n 'secondary',\n 'share',\n 'shopping',\n 'shoutbox',\n 'side',\n 'sidebar',\n 'sponsor',\n 'stamp',\n 'sub',\n 'summary',\n 'tags',\n 'tools',\n 'widget',\n];\n// The above list, joined into a matching regular expression\nexport const NEGATIVE_SCORE_RE = new RegExp(NEGATIVE_SCORE_HINTS.join('|'), 'i');\n\n// XPath to try to determine if a page is wordpress. Not always successful.\nexport const IS_WP_SELECTOR = 'meta[name=generator][value^=WordPress]';\n\n// Match a digit. Pretty clear.\nexport const DIGIT_RE = new RegExp('[0-9]');\n\n// A list of words that, if found in link text or URLs, likely mean that\n// this link is not a next page link.\nexport const EXTRANEOUS_LINK_HINTS = [\n 'print',\n 'archive',\n 'comment',\n 'discuss',\n 'e-mail',\n 'email',\n 'share',\n 'reply',\n 'all',\n 'login',\n 'sign',\n 'single',\n 'adx',\n 'entry-unrelated',\n];\nexport const EXTRANEOUS_LINK_HINTS_RE = new RegExp(EXTRANEOUS_LINK_HINTS.join('|'), 'i');\n\n// Match any phrase that looks like it could be page, or paging, or pagination\nexport const PAGE_RE = new RegExp('pag(e|ing|inat)', 'i');\n\n// Match any link text/classname/id that looks like it could mean the next\n// page. Things like: next, continue, >, >>, » but not >|, »| as those can\n// mean last page.\n// export const NEXT_LINK_TEXT_RE = new RegExp('(next|weiter|continue|>([^\\|]|$)|»([^\\|]|$))', 'i');\nexport const NEXT_LINK_TEXT_RE = /(next|weiter|continue|>([^\\|]|$)|»([^\\|]|$))/i;\n\n// Match any link text/classname/id that looks like it is an end link: things\n// like \"first\", \"last\", \"end\", etc.\nexport const CAP_LINK_TEXT_RE = new RegExp('(first|last|end)', 'i');\n\n// Match any link text/classname/id that looks like it means the previous\n// page.\nexport const PREV_LINK_TEXT_RE = new RegExp('(prev|earl|old|new|<|«)', 'i');\n\n// Match 2 or more consecutive <br> tags\nexport const BR_TAGS_RE = new RegExp('(<br[^>]*>[ \\n\\r\\t]*){2,}', 'i');\n\n// Match 1 BR tag.\nexport const BR_TAG_RE = new RegExp('<br[^>]*>', 'i');\n\n// A list of all of the block level tags known in HTML5 and below. Taken from\n// http://bit.ly/qneNIT\nexport const BLOCK_LEVEL_TAGS = [\n 'article',\n 'aside',\n 'blockquote',\n 'body',\n 'br',\n 'button',\n 'canvas',\n 'caption',\n 'col',\n 'colgroup',\n 'dd',\n 'div',\n 'dl',\n 'dt',\n 'embed',\n 'fieldset',\n 'figcaption',\n 'figure',\n 'footer',\n 'form',\n 'h1',\n 'h2',\n 'h3',\n 'h4',\n 'h5',\n 'h6',\n 'header',\n 'hgroup',\n 'hr',\n 'li',\n 'map',\n 'object',\n 'ol',\n 'output',\n 'p',\n 'pre',\n 'progress',\n 'section',\n 'table',\n 'tbody',\n 'textarea',\n 'tfoot',\n 'th',\n 'thead',\n 'tr',\n 'ul',\n 'video',\n];\nexport const BLOCK_LEVEL_TAGS_RE = new RegExp(`^(${BLOCK_LEVEL_TAGS.join('|')})$`, 'i');\n\n\n// The removal is implemented as a blacklist and whitelist, this test finds\n// blacklisted elements that aren't whitelisted. We do this all in one\n// expression-both because it's only one pass, and because this skips the\n// serialization for whitelisted nodes.\nconst candidatesBlacklist = UNLIKELY_CANDIDATES_BLACKLIST.join('|');\nexport const CANDIDATES_BLACKLIST = new RegExp(candidatesBlacklist, 'i');\n\nconst candidatesWhitelist = UNLIKELY_CANDIDATES_WHITELIST.join('|');\nexport const CANDIDATES_WHITELIST = new RegExp(candidatesWhitelist, 'i');\n\nexport const UNLIKELY_RE = new RegExp(`!(${candidatesWhitelist})|(${candidatesBlacklist})`, 'i');\n\n\nexport const PARAGRAPH_SCORE_TAGS = new RegExp('^(p|li|span|pre)$', 'i');\nexport const CHILD_CONTENT_TAGS = new RegExp('^(td|blockquote|ol|ul|dl)$', 'i');\nexport const BAD_TAGS = new RegExp('^(address|form)$', 'i');\n\nexport const HTML_OR_BODY_RE = new RegExp('^(html|body)$', 'i');\n","import {\n CANDIDATES_WHITELIST,\n CANDIDATES_BLACKLIST,\n} from './constants';\n\nexport default function stripUnlikelyCandidates($) {\n // Loop through the provided document and remove any non-link nodes\n // that are unlikely candidates for article content.\n //\n // Links are ignored because there are very often links to content\n // that are identified as non-body-content, but may be inside\n // article-like content.\n //\n // :param $: a cheerio object to strip nodes from\n // :return $: the cleaned cheerio object\n $('*').not('a').each((index, node) => {\n const $node = $(node);\n const classes = $node.attr('class');\n const id = $node.attr('id');\n if (!id && !classes) return;\n\n const classAndId = `${classes || ''} ${id || ''}`;\n if (CANDIDATES_WHITELIST.test(classAndId)) {\n return;\n } else if (CANDIDATES_BLACKLIST.test(classAndId)) {\n $node.remove();\n }\n });\n\n return $;\n}\n","import { paragraphize } from './index';\n\n// ## NOTES:\n// Another good candidate for refactoring/optimizing.\n// Very imperative code, I don't love it. - AP\n\n\n// Given cheerio object, convert consecutive <br /> tags into\n// <p /> tags instead.\n//\n// :param $: A cheerio object\n\nexport default function brsToPs($) {\n let collapsing = false;\n $('br').each((index, element) => {\n const nextElement = $(element).next().get(0);\n\n if (nextElement && nextElement.tagName === 'br') {\n collapsing = true;\n $(element).remove();\n } else if (collapsing) {\n collapsing = false;\n // $(element).replaceWith('<p />')\n paragraphize(element, $, true);\n }\n });\n\n return $;\n}\n","import { BLOCK_LEVEL_TAGS_RE } from './constants';\n\n// Given a node, turn it into a P if it is not already a P, and\n// make sure it conforms to the constraints of a P tag (I.E. does\n// not contain any other block tags.)\n//\n// If the node is a <br />, it treats the following inline siblings\n// as if they were its children.\n//\n// :param node: The node to paragraphize; this is a raw node\n// :param $: The cheerio object to handle dom manipulation\n// :param br: Whether or not the passed node is a br\n\nexport default function paragraphize(node, $, br = false) {\n const $node = $(node);\n\n if (br) {\n let sibling = node.nextSibling;\n const p = $('<p></p>');\n\n // while the next node is text or not a block level element\n // append it to a new p node\n while (sibling && !(sibling.tagName && BLOCK_LEVEL_TAGS_RE.test(sibling.tagName))) {\n const nextSibling = sibling.nextSibling;\n $(sibling).appendTo(p);\n sibling = nextSibling;\n }\n\n $node.replaceWith(p);\n $node.remove();\n return $;\n }\n\n return $;\n}\n","import { brsToPs, convertNodeTo } from 'utils/dom';\n\nimport { DIV_TO_P_BLOCK_TAGS } from './constants';\n\nfunction convertDivs($) {\n $('div').each((index, div) => {\n const $div = $(div);\n const convertable = $div.children(DIV_TO_P_BLOCK_TAGS).length === 0;\n\n if (convertable) {\n convertNodeTo($div, $, 'p');\n }\n });\n\n return $;\n}\n\nfunction convertSpans($) {\n $('span').each((index, span) => {\n const $span = $(span);\n const convertable = $span.parents('p, div').length === 0;\n if (convertable) {\n convertNodeTo($span, $, 'p');\n }\n });\n\n return $;\n}\n\n// Loop through the provided doc, and convert any p-like elements to\n// actual paragraph tags.\n//\n// Things fitting this criteria:\n// * Multiple consecutive <br /> tags.\n// * <div /> tags without block level elements inside of them\n// * <span /> tags who are not children of <p /> or <div /> tags.\n//\n// :param $: A cheerio object to search\n// :return cheerio object with new p elements\n// (By-reference mutation, though. Returned just for convenience.)\n\nexport default function convertToParagraphs($) {\n $ = brsToPs($);\n $ = convertDivs($);\n $ = convertSpans($);\n\n return $;\n}\n","export default function convertNodeTo($node, $, tag = 'p') {\n const node = $node.get(0);\n if (!node) {\n return $;\n }\n const { attribs } = $node.get(0);\n const attribString = Reflect.ownKeys(attribs)\n .map(key => `${key}=${attribs[key]}`)\n .join(' ');\n\n $node.replaceWith(`<${tag} ${attribString}>${$node.contents()}</${tag}>`);\n return $;\n}\n","import { SPACER_RE } from './constants';\n\nfunction cleanForHeight($img, $) {\n const height = parseInt($img.attr('height'), 10);\n const width = parseInt($img.attr('width'), 10) || 20;\n\n // Remove images that explicitly have very small heights or\n // widths, because they are most likely shims or icons,\n // which aren't very useful for reading.\n if ((height || 20) < 10 || width < 10) {\n $img.remove();\n } else if (height) {\n // Don't ever specify a height on images, so that we can\n // scale with respect to width without screwing up the\n // aspect ratio.\n $img.removeAttr('height');\n }\n\n return $;\n}\n\n// Cleans out images where the source string matches transparent/spacer/etc\n// TODO This seems very aggressive - AP\nfunction removeSpacers($img, $) {\n if (SPACER_RE.test($img.attr('src'))) {\n $img.remove();\n }\n\n return $;\n}\n\nexport default function cleanImages($article, $) {\n $article.find('img').each((index, img) => {\n const $img = $(img);\n\n cleanForHeight($img, $);\n removeSpacers($img, $);\n });\n\n return $;\n}\n","import {\n KEEP_SELECTORS,\n KEEP_CLASS,\n} from './constants';\n\nexport default function markToKeep(article, $, tags = []) {\n if (tags.length === 0) {\n tags = KEEP_SELECTORS;\n }\n\n $(tags.join(','), article).addClass(KEEP_CLASS);\n\n return $;\n}\n","import {\n STRIP_OUTPUT_TAGS,\n KEEP_CLASS,\n} from './constants';\n\nexport default function stripJunkTags(article, $, tags = []) {\n if (tags.length === 0) {\n tags = STRIP_OUTPUT_TAGS;\n }\n\n // Remove matching elements, but ignore\n // any element with a class of mercury-parser-keep\n $(tags.join(','), article).not(`.${KEEP_CLASS}`).remove();\n\n // Remove the mercury-parser-keep class from result\n $(`.${KEEP_CLASS}`, article).removeClass(KEEP_CLASS);\n\n return $;\n}\n","import { convertNodeTo } from 'utils/dom';\n\n// H1 tags are typically the article title, which should be extracted\n// by the title extractor instead. If there's less than 3 of them (<3),\n// strip them. Otherwise, turn 'em into H2s.\nexport default function cleanHOnes(article, $) {\n const $hOnes = $('h1', article);\n\n if ($hOnes.length < 3) {\n $hOnes.each((index, node) => $(node).remove());\n } else {\n $hOnes.each((index, node) => {\n convertNodeTo($(node), $, 'h2');\n });\n }\n\n return $;\n}\n","import { WHITELIST_ATTRS_RE } from './constants';\n\nfunction removeAllButWhitelist($article) {\n // $('*', article).each((index, node) => {\n $article.find('*').each((index, node) => {\n node.attribs = Reflect.ownKeys(node.attribs).reduce((acc, attr) => {\n if (WHITELIST_ATTRS_RE.test(attr)) {\n return { ...acc, [attr]: node.attribs[attr] };\n }\n\n return acc;\n }, {});\n });\n\n return $article;\n}\n\n// function removeAttrs(article, $) {\n// REMOVE_ATTRS.forEach((attr) => {\n// $(`[${attr}]`, article).removeAttr(attr);\n// });\n// }\n\n// Remove attributes like style or align\nexport default function cleanAttributes($article) {\n // Grabbing the parent because at this point\n // $article will be wrapped in a div which will\n // have a score set on it.\n console.log('HMM', $article.parent().length);\n return removeAllButWhitelist(\n $article.parent().length ?\n $article.parent() : $article\n );\n}\n","export default function removeEmpty($article, $) {\n $article.find('p').each((index, p) => {\n const $p = $(p);\n if ($p.find('iframe, img').length === 0 && $p.text().trim() === '') $p.remove();\n });\n\n return $;\n}\n","// // CONTENT FETCHING CONSTANTS ////\n\n// A list of strings that can be considered unlikely candidates when\n// extracting content from a resource. These strings are joined together\n// and then tested for existence using re:test, so may contain simple,\n// non-pipe style regular expression queries if necessary.\nexport const UNLIKELY_CANDIDATES_BLACKLIST = [\n 'ad-break',\n 'adbox',\n 'advert',\n 'addthis',\n 'agegate',\n 'aux',\n 'blogger-labels',\n 'combx',\n 'comment',\n 'conversation',\n 'disqus',\n 'entry-unrelated',\n 'extra',\n 'foot',\n 'form',\n 'header',\n 'hidden',\n 'loader',\n 'login', // Note: This can hit 'blogindex'.\n 'menu',\n 'meta',\n 'nav',\n 'pager',\n 'pagination',\n 'predicta', // readwriteweb inline ad box\n 'presence_control_external', // lifehacker.com container full of false positives\n 'popup',\n 'printfriendly',\n 'related',\n 'remove',\n 'remark',\n 'rss',\n 'share',\n 'shoutbox',\n 'sidebar',\n 'sociable',\n 'sponsor',\n 'tools',\n];\n\n// A list of strings that can be considered LIKELY candidates when\n// extracting content from a resource. Essentially, the inverse of the\n// blacklist above - if something matches both blacklist and whitelist,\n// it is kept. This is useful, for example, if something has a className\n// of \"rss-content entry-content\". It matched 'rss', so it would normally\n// be removed, however, it's also the entry content, so it should be left\n// alone.\n//\n// These strings are joined together and then tested for existence using\n// re:test, so may contain simple, non-pipe style regular expression queries\n// if necessary.\nexport const UNLIKELY_CANDIDATES_WHITELIST = [\n 'and',\n 'article',\n 'body',\n 'blogindex',\n 'column',\n 'content',\n 'entry-content-asset',\n 'format', // misuse of form\n 'hfeed',\n 'hentry',\n 'hatom',\n 'main',\n 'page',\n 'posts',\n 'shadow',\n];\n\n// A list of tags which, if found inside, should cause a <div /> to NOT\n// be turned into a paragraph tag. Shallow div tags without these elements\n// should be turned into <p /> tags.\nexport const DIV_TO_P_BLOCK_TAGS = [\n 'a',\n 'blockquote',\n 'dl',\n 'div',\n 'img',\n 'p',\n 'pre',\n 'table',\n].join(',');\n\n// A list of tags that should be ignored when trying to find the top candidate\n// for a document.\nexport const NON_TOP_CANDIDATE_TAGS = [\n 'br',\n 'b',\n 'i',\n 'label',\n 'hr',\n 'area',\n 'base',\n 'basefont',\n 'input',\n 'img',\n 'link',\n 'meta',\n];\n\nexport const NON_TOP_CANDIDATE_TAGS_RE =\n new RegExp(`^(${NON_TOP_CANDIDATE_TAGS.join('|')})$`, 'i');\n\n// A list of selectors that specify, very clearly, either hNews or other\n// very content-specific style content, like Blogger templates.\n// More examples here: http://microformats.org/wiki/blog-post-formats\nexport const HNEWS_CONTENT_SELECTORS = [\n ['.hentry', '.entry-content'],\n ['entry', '.entry-content'],\n ['.entry', '.entry_content'],\n ['.post', '.postbody'],\n ['.post', '.post_body'],\n ['.post', '.post-body'],\n];\n\nexport const PHOTO_HINTS = [\n 'figure',\n 'photo',\n 'image',\n 'caption',\n];\nexport const PHOTO_HINTS_RE = new RegExp(PHOTO_HINTS.join('|'), 'i');\n\n\n// A list of strings that denote a positive scoring for this content as being\n// an article container. Checked against className and id.\n//\n// TODO: Perhaps have these scale based on their odds of being quality?\nexport const POSITIVE_SCORE_HINTS = [\n 'article',\n 'articlecontent',\n 'instapaper_body',\n 'blog',\n 'body',\n 'content',\n 'entry-content-asset',\n 'entry',\n 'hentry',\n 'main',\n 'Normal',\n 'page',\n 'pagination',\n 'permalink',\n 'post',\n 'story',\n 'text',\n '[-_]copy', // usatoday\n '\\\\Bcopy',\n];\n\n// The above list, joined into a matching regular expression\nexport const POSITIVE_SCORE_RE = new RegExp(POSITIVE_SCORE_HINTS.join('|'), 'i');\n\n// Readability publisher-specific guidelines\nexport const READABILITY_ASSET = new RegExp('entry-content-asset', 'i');\n\n// A list of strings that denote a negative scoring for this content as being\n// an article container. Checked against className and id.\n//\n// TODO: Perhaps have these scale based on their odds of being quality?\nexport const NEGATIVE_SCORE_HINTS = [\n 'adbox',\n 'advert',\n 'author',\n 'bio',\n 'bookmark',\n 'bottom',\n 'byline',\n 'clear',\n 'com-',\n 'combx',\n 'comment',\n 'comment\\\\B',\n 'contact',\n 'copy',\n 'credit',\n 'crumb',\n 'date',\n 'deck',\n 'excerpt',\n 'featured', // tnr.com has a featured_content which throws us off\n 'foot',\n 'footer',\n 'footnote',\n 'graf',\n 'head',\n 'info',\n 'infotext', // newscientist.com copyright\n 'instapaper_ignore',\n 'jump',\n 'linebreak',\n 'link',\n 'masthead',\n 'media',\n 'meta',\n 'modal',\n 'outbrain', // slate.com junk\n 'promo',\n 'pr_', // autoblog - press release\n 'related',\n 'respond',\n 'roundcontent', // lifehacker restricted content warning\n 'scroll',\n 'secondary',\n 'share',\n 'shopping',\n 'shoutbox',\n 'side',\n 'sidebar',\n 'sponsor',\n 'stamp',\n 'sub',\n 'summary',\n 'tags',\n 'tools',\n 'widget',\n];\n// The above list, joined into a matching regular expression\nexport const NEGATIVE_SCORE_RE = new RegExp(NEGATIVE_SCORE_HINTS.join('|'), 'i');\n\n// Match a digit. Pretty clear.\nexport const DIGIT_RE = new RegExp('[0-9]');\n\n// Match 2 or more consecutive <br> tags\nexport const BR_TAGS_RE = new RegExp('(<br[^>]*>[ \\n\\r\\t]*){2,}', 'i');\n\n// Match 1 BR tag.\nexport const BR_TAG_RE = new RegExp('<br[^>]*>', 'i');\n\n// A list of all of the block level tags known in HTML5 and below. Taken from\n// http://bit.ly/qneNIT\nexport const BLOCK_LEVEL_TAGS = [\n 'article',\n 'aside',\n 'blockquote',\n 'body',\n 'br',\n 'button',\n 'canvas',\n 'caption',\n 'col',\n 'colgroup',\n 'dd',\n 'div',\n 'dl',\n 'dt',\n 'embed',\n 'fieldset',\n 'figcaption',\n 'figure',\n 'footer',\n 'form',\n 'h1',\n 'h2',\n 'h3',\n 'h4',\n 'h5',\n 'h6',\n 'header',\n 'hgroup',\n 'hr',\n 'li',\n 'map',\n 'object',\n 'ol',\n 'output',\n 'p',\n 'pre',\n 'progress',\n 'section',\n 'table',\n 'tbody',\n 'textarea',\n 'tfoot',\n 'th',\n 'thead',\n 'tr',\n 'ul',\n 'video',\n];\nexport const BLOCK_LEVEL_TAGS_RE = new RegExp(`^(${BLOCK_LEVEL_TAGS.join('|')})$`, 'i');\n\n\n// The removal is implemented as a blacklist and whitelist, this test finds\n// blacklisted elements that aren't whitelisted. We do this all in one\n// expression-both because it's only one pass, and because this skips the\n// serialization for whitelisted nodes.\nconst candidatesBlacklist = UNLIKELY_CANDIDATES_BLACKLIST.join('|');\nexport const CANDIDATES_BLACKLIST = new RegExp(candidatesBlacklist, 'i');\n\nconst candidatesWhitelist = UNLIKELY_CANDIDATES_WHITELIST.join('|');\nexport const CANDIDATES_WHITELIST = new RegExp(candidatesWhitelist, 'i');\n\nexport const UNLIKELY_RE = new RegExp(`!(${candidatesWhitelist})|(${candidatesBlacklist})`, 'i');\n\n\nexport const PARAGRAPH_SCORE_TAGS = new RegExp('^(p|li|span|pre)$', 'i');\nexport const CHILD_CONTENT_TAGS = new RegExp('^(td|blockquote|ol|ul|dl)$', 'i');\nexport const BAD_TAGS = new RegExp('^(address|form)$', 'i');\n\nexport const HTML_OR_BODY_RE = new RegExp('^(html|body)$', 'i');\n","import {\n NEGATIVE_SCORE_RE,\n POSITIVE_SCORE_RE,\n PHOTO_HINTS_RE,\n READABILITY_ASSET,\n} from './constants';\n\n\n// Get the score of a node based on its className and id.\nexport default function getWeight(node) {\n const classes = node.attr('class');\n const id = node.attr('id');\n let score = 0;\n\n if (id) {\n // if id exists, try to score on both positive and negative\n if (POSITIVE_SCORE_RE.test(id)) {\n score += 25;\n }\n if (NEGATIVE_SCORE_RE.test(id)) {\n score -= 25;\n }\n }\n\n if (classes) {\n if (score === 0) {\n // if classes exist and id did not contribute to score\n // try to score on both positive and negative\n if (POSITIVE_SCORE_RE.test(classes)) {\n score += 25;\n }\n if (NEGATIVE_SCORE_RE.test(classes)) {\n score -= 25;\n }\n }\n\n // even if score has been set by id, add score for\n // possible photo matches\n // \"try to keep photos if we can\"\n if (PHOTO_HINTS_RE.test(classes)) {\n score += 10;\n }\n\n // add 25 if class matches entry-content-asset,\n // a class apparently instructed for use in the\n // Readability publisher guidelines\n // https://www.readability.com/developers/guidelines\n if (READABILITY_ASSET.test(classes)) {\n score += 25;\n }\n }\n\n return score;\n}\n\n","// returns the score of a node based on\n// the node's score attribute\n// returns null if no score set\nexport default function getScore($node) {\n return parseFloat($node.attr('score')) || null;\n}\n","// return 1 for every comma in text\nexport default function scoreCommas(text) {\n return (text.match(/,/g) || []).length;\n}\n\n","const idkRe = new RegExp('^(p|pre)$', 'i');\n\nexport default function scoreLength(textLength, tagName = 'p') {\n const chunks = textLength / 50;\n\n if (chunks > 0) {\n let lengthBonus;\n\n // No idea why p or pre are being tamped down here\n // but just following the source for now\n // Not even sure why tagName is included here,\n // since this is only being called from the context\n // of scoreParagraph\n if (idkRe.test(tagName)) {\n lengthBonus = chunks - 2;\n } else {\n lengthBonus = chunks - 1.25;\n }\n\n return Math.min(Math.max(lengthBonus, 0), 3);\n }\n\n return 0;\n}\n\n","import {\n scoreCommas,\n scoreLength,\n} from './index';\n\n// Score a paragraph using various methods. Things like number of\n// commas, etc. Higher is better.\nexport default function scoreParagraph(node) {\n let score = 1;\n const text = node.text().trim();\n const textLength = text.length;\n\n // If this paragraph is less than 25 characters, don't count it.\n if (textLength < 25) {\n return 0;\n }\n\n // Add points for any commas within this paragraph\n score += scoreCommas(text);\n\n // For every 50 characters in this paragraph, add another point. Up\n // to 3 points.\n score += scoreLength(textLength);\n\n // Articles can end with short paragraphs when people are being clever\n // but they can also end with short paragraphs setting up lists of junk\n // that we strip. This negative tweaks junk setup paragraphs just below\n // the cutoff threshold.\n if (text.slice(-1) === ':') {\n score -= 1;\n }\n\n return score;\n}\n\n","\nexport default function setScore($node, $, score) {\n $node.attr('score', score);\n return $node;\n}\n\n","import {\n getOrInitScore,\n setScore,\n} from './index';\n\nexport default function addScore($node, $, amount) {\n try {\n const score = getOrInitScore($node, $) + amount;\n setScore($node, $, score);\n } catch (e) {\n // Ignoring; error occurs in scoreNode\n }\n\n return $node;\n}\n","import { addScore } from './index';\n\n// Adds 1/4 of a child's score to its parent\nexport default function addToParent(node, $, score) {\n const parent = node.parent();\n if (parent) {\n addScore(parent, $, score * 0.25);\n }\n\n return node;\n}\n","import {\n getScore,\n scoreNode,\n getWeight,\n addToParent,\n} from './index';\n\n// gets and returns the score if it exists\n// if not, initializes a score based on\n// the node's tag type\nexport default function getOrInitScore($node, $, weightNodes = true) {\n let score = getScore($node);\n\n if (score) {\n return score;\n }\n\n score = scoreNode($node);\n\n if (weightNodes) {\n score += getWeight($node);\n }\n\n addToParent($node, $, score);\n\n return score;\n}\n\n","import { scoreParagraph } from './index';\nimport {\n PARAGRAPH_SCORE_TAGS,\n CHILD_CONTENT_TAGS,\n BAD_TAGS,\n} from './constants';\n\n// Score an individual node. Has some smarts for paragraphs, otherwise\n// just scores based on tag.\nexport default function scoreNode($node) {\n const { tagName } = $node.get(0);\n\n // TODO: Consider ordering by most likely.\n // E.g., if divs are a more common tag on a page,\n // Could save doing that regex test on every node – AP\n if (PARAGRAPH_SCORE_TAGS.test(tagName)) {\n return scoreParagraph($node);\n } else if (tagName === 'div') {\n return 5;\n } else if (CHILD_CONTENT_TAGS.test(tagName)) {\n return 3;\n } else if (BAD_TAGS.test(tagName)) {\n return -3;\n } else if (tagName === 'th') {\n return -5;\n }\n\n return 0;\n}\n","import { convertNodeTo } from 'utils/dom';\n\nimport { HNEWS_CONTENT_SELECTORS } from './constants';\nimport {\n scoreNode,\n setScore,\n getOrInitScore,\n addScore,\n} from './index';\n\nfunction convertSpans($node, $) {\n if ($node.get(0)) {\n const { tagName } = $node.get(0);\n\n if (tagName === 'span') {\n // convert spans to divs\n convertNodeTo($node, $, 'div');\n }\n }\n}\n\nfunction addScoreTo($node, $, score) {\n if ($node) {\n convertSpans($node, $);\n addScore($node, $, score);\n }\n}\n\nfunction scorePs($, weightNodes) {\n $('p, pre').not('[score]').each((index, node) => {\n // The raw score for this paragraph, before we add any parent/child\n // scores.\n let $node = $(node);\n $node = setScore($node, $, getOrInitScore($node, $, weightNodes));\n\n const $parent = $node.parent();\n const rawScore = scoreNode($node);\n\n addScoreTo($parent, $, rawScore, weightNodes);\n if ($parent) {\n // Add half of the individual content score to the\n // grandparent\n addScoreTo($parent.parent(), $, rawScore / 2, weightNodes);\n }\n });\n\n return $;\n}\n\n// score content. Parents get the full value of their children's\n// content score, grandparents half\nexport default function scoreContent($, weightNodes = true) {\n // First, look for special hNews based selectors and give them a big\n // boost, if they exist\n HNEWS_CONTENT_SELECTORS.forEach(([parentSelector, childSelector]) => {\n $(`${parentSelector} ${childSelector}`).each((index, node) => {\n addScore($(node).parent(parentSelector), $, 80);\n });\n });\n\n // Doubling this again\n // Previous solution caused a bug\n // in which parents weren't retaining\n // scores. This is not ideal, and\n // should be fixed.\n scorePs($, weightNodes);\n scorePs($, weightNodes);\n\n return $;\n}\n","const NORMALIZE_RE = /\\s{2,}/g;\n\nexport default function normalizeSpaces(text) {\n return text.replace(NORMALIZE_RE, ' ').trim();\n}\n","// Given a node type to search for, and a list of regular expressions,\n// look to see if this extraction can be found in the URL. Expects\n// that each expression in r_list will return group(1) as the proper\n// string to be cleaned.\n// Only used for date_published currently.\nexport default function extractFromUrl(url, regexList) {\n const matchRe = regexList.find(re => re.test(url));\n if (matchRe) {\n return matchRe.exec(url)[1];\n }\n\n return null;\n}\n","// An expression that looks to try to find the page digit within a URL, if\n// it exists.\n// Matches:\n// page=1\n// pg=1\n// p=1\n// paging=12\n// pag=7\n// pagination/1\n// paging/88\n// pa/83\n// p/11\n//\n// Does not match:\n// pg=102\n// page:2\nexport const PAGE_IN_HREF_RE = new RegExp('(page|paging|(p(a|g|ag)?(e|enum|ewanted|ing|ination)))?(=|/)([0-9]{1,3})', 'i');\n\nexport const HAS_ALPHA_RE = /[a-z]/i;\n\nexport const IS_ALPHA_RE = /^[a-z]+$/i;\nexport const IS_DIGIT_RE = /^[0-9]+$/i;\n","import { PAGE_IN_HREF_RE } from './constants';\n\nexport default function pageNumFromUrl(url) {\n const matches = url.match(PAGE_IN_HREF_RE);\n if (!matches) return null;\n\n const pageNum = parseInt(matches[6], 10);\n\n // Return pageNum < 100, otherwise\n // return null\n return pageNum < 100 ? pageNum : null;\n}\n","export default function removeAnchor(url) {\n return url.split('#')[0].replace(/\\/$/, '');\n}\n","import URL from 'url';\nimport {\n HAS_ALPHA_RE,\n IS_ALPHA_RE,\n IS_DIGIT_RE,\n PAGE_IN_HREF_RE,\n} from './constants';\n\nfunction isGoodSegment(segment, index, firstSegmentHasLetters) {\n let goodSegment = true;\n\n // If this is purely a number, and it's the first or second\n // url_segment, it's probably a page number. Remove it.\n if (index < 2 && IS_DIGIT_RE.test(segment) && segment.length < 3) {\n goodSegment = true;\n }\n\n // If this is the first url_segment and it's just \"index\",\n // remove it\n if (index === 0 && segment.toLowerCase() === 'index') {\n goodSegment = false;\n }\n\n // If our first or second url_segment is smaller than 3 characters,\n // and the first url_segment had no alphas, remove it.\n if (index < 2 && segment.length < 3 && !firstSegmentHasLetters) {\n goodSegment = false;\n }\n\n return goodSegment;\n}\n\n// Take a URL, and return the article base of said URL. That is, no\n// pagination data exists in it. Useful for comparing to other links\n// that might have pagination data within them.\nexport default function articleBaseUrl(url, parsed) {\n const parsedUrl = parsed || URL.parse(url);\n const { protocol, host, path } = parsedUrl;\n\n let firstSegmentHasLetters = false;\n const cleanedSegments = path.split('/')\n .reverse()\n .reduce((acc, rawSegment, index) => {\n let segment = rawSegment;\n\n // Split off and save anything that looks like a file type.\n if (segment.includes('.')) {\n const [possibleSegment, fileExt] = segment.split('.');\n if (IS_ALPHA_RE.test(fileExt)) {\n segment = possibleSegment;\n }\n }\n\n // If our first or second segment has anything looking like a page\n // number, remove it.\n if (PAGE_IN_HREF_RE.test(segment) && index < 2) {\n segment = segment.replace(PAGE_IN_HREF_RE, '');\n }\n\n // If we're on the first segment, check to see if we have any\n // characters in it. The first segment is actually the last bit of\n // the URL, and this will be helpful to determine if we're on a URL\n // segment that looks like \"/2/\" for example.\n if (index === 0) {\n firstSegmentHasLetters = HAS_ALPHA_RE.test(segment);\n }\n\n // If it's not marked for deletion, push it to cleaned_segments.\n if (isGoodSegment(segment, index, firstSegmentHasLetters)) {\n acc.push(segment);\n }\n\n return acc;\n }, []);\n\n return `${protocol}//${host}${cleanedSegments.reverse().join('/')}`;\n}\n","// Given a string, return True if it appears to have an ending sentence\n// within it, false otherwise.\nconst SENTENCE_END_RE = new RegExp('.( |$)');\nexport default function hasSentenceEnd(text) {\n return SENTENCE_END_RE.test(text);\n}\n\n","export default function excerptContent(content, words = 10) {\n return content.trim()\n .split(/\\s+/)\n .slice(0, words)\n .join(' ');\n}\n","import {\n textLength,\n linkDensity,\n} from 'utils/dom';\nimport { hasSentenceEnd } from 'utils/text';\n\nimport { NON_TOP_CANDIDATE_TAGS_RE } from './constants';\nimport { getScore } from './index';\n\n// Now that we have a top_candidate, look through the siblings of\n// it to see if any of them are decently scored. If they are, they\n// may be split parts of the content (Like two divs, a preamble and\n// a body.) Example:\n// http://articles.latimes.com/2009/oct/14/business/fi-bigtvs14\nexport default function mergeSiblings($candidate, topScore, $) {\n if (!$candidate.parent().length) {\n return $candidate;\n }\n\n const siblingScoreThreshold = Math.max(10, topScore * 0.25);\n const wrappingDiv = $('<div></div>');\n\n $candidate.parent().children().each((index, sibling) => {\n const $sibling = $(sibling);\n // Ignore tags like BR, HR, etc\n if (NON_TOP_CANDIDATE_TAGS_RE.test(sibling.tagName)) {\n return null;\n }\n\n const siblingScore = getScore($sibling);\n if (siblingScore) {\n if ($sibling === $candidate) {\n wrappingDiv.append($sibling);\n } else {\n let contentBonus = 0;\n const density = linkDensity($sibling);\n\n // If sibling has a very low link density,\n // give it a small bonus\n if (density < 0.05) {\n contentBonus += 20;\n }\n\n // If sibling has a high link density,\n // give it a penalty\n if (density >= 0.5) {\n contentBonus -= 20;\n }\n\n // If sibling node has the same class as\n // candidate, give it a bonus\n if ($sibling.attr('class') === $candidate.attr('class')) {\n contentBonus += topScore * 0.2;\n }\n\n const newScore = siblingScore + contentBonus;\n\n if (newScore >= siblingScoreThreshold) {\n return wrappingDiv.append($sibling);\n } else if (sibling.tagName === 'p') {\n const siblingContent = $sibling.text();\n const siblingContentLength = textLength(siblingContent);\n\n if (siblingContentLength > 80 && density < 0.25) {\n return wrappingDiv.append($sibling);\n } else if (siblingContentLength <= 80 && density === 0 &&\n hasSentenceEnd(siblingContent)) {\n return wrappingDiv.append($sibling);\n }\n }\n }\n }\n\n return null;\n });\n\n return wrappingDiv;\n}\n","import { NON_TOP_CANDIDATE_TAGS_RE } from './constants';\nimport { getScore } from './index';\nimport mergeSiblings from './merge-siblings';\n\n// After we've calculated scores, loop through all of the possible\n// candidate nodes we found and find the one with the highest score.\nexport default function findTopCandidate($) {\n let $candidate;\n let topScore = 0;\n\n $('[score]').each((index, node) => {\n // Ignore tags like BR, HR, etc\n if (NON_TOP_CANDIDATE_TAGS_RE.test(node.tagName)) {\n return;\n }\n\n const $node = $(node);\n const score = getScore($node);\n\n if (score > topScore) {\n topScore = score;\n $candidate = $node;\n }\n });\n\n // If we don't have a candidate, return the body\n // or whatever the first element is\n if (!$candidate) {\n return $('body') || $('*').first();\n }\n\n $candidate = mergeSiblings($candidate, topScore, $);\n\n return $candidate;\n}\n","import {\n getScore,\n setScore,\n getOrInitScore,\n scoreCommas,\n} from 'extractors/generic/content/scoring';\n\nimport { CLEAN_CONDITIONALLY_TAGS } from './constants';\nimport { normalizeSpaces } from '../text';\nimport { linkDensity } from './index';\n\nfunction removeUnlessContent($node, $, weight) {\n // Explicitly save entry-content-asset tags, which are\n // noted as valuable in the Publisher guidelines. For now\n // this works everywhere. We may want to consider making\n // this less of a sure-thing later.\n if ($node.hasClass('entry-content-asset')) {\n return;\n }\n\n const content = normalizeSpaces($node.text());\n\n if (scoreCommas(content) < 10) {\n const pCount = $('p', $node).length;\n const inputCount = $('input', $node).length;\n\n // Looks like a form, too many inputs.\n if (inputCount > (pCount / 3)) {\n $node.remove();\n return;\n }\n\n const contentLength = content.length;\n const imgCount = $('img', $node).length;\n\n // Content is too short, and there are no images, so\n // this is probably junk content.\n if (contentLength < 25 && imgCount === 0) {\n $node.remove();\n return;\n }\n\n const density = linkDensity($node);\n\n // Too high of link density, is probably a menu or\n // something similar.\n // console.log(weight, density, contentLength)\n if (weight < 25 && density > 0.2 && contentLength > 75) {\n $node.remove();\n return;\n }\n\n // Too high of a link density, despite the score being\n // high.\n if (weight >= 25 && density > 0.5) {\n // Don't remove the node if it's a list and the\n // previous sibling starts with a colon though. That\n // means it's probably content.\n const tagName = $node.get(0).tagName;\n const nodeIsList = tagName === 'ol' || tagName === 'ul';\n if (nodeIsList) {\n const previousNode = $node.prev();\n if (previousNode && normalizeSpaces(previousNode.text()).slice(-1) === ':') {\n return;\n }\n }\n\n $node.remove();\n return;\n }\n\n const scriptCount = $('script', $node).length;\n\n // Too many script tags, not enough content.\n if (scriptCount > 0 && contentLength < 150) {\n $node.remove();\n return;\n }\n }\n}\n\n// Given an article, clean it of some superfluous content specified by\n// tags. Things like forms, ads, etc.\n//\n// Tags is an array of tag name's to search through. (like div, form,\n// etc)\n//\n// Return this same doc.\nexport default function cleanTags($article, $) {\n $(CLEAN_CONDITIONALLY_TAGS, $article).each((index, node) => {\n const $node = $(node);\n let weight = getScore($node);\n if (!weight) {\n weight = getOrInitScore($node, $);\n setScore($node, $, weight);\n }\n\n // drop node if its weight is < 0\n if (weight < 0) {\n $node.remove();\n } else {\n // deteremine if node seems like content\n removeUnlessContent($node, $, weight);\n }\n });\n\n return $;\n}\n\n","import { getWeight } from 'extractors/generic/content/scoring';\n\nimport { HEADER_TAG_LIST } from './constants';\nimport { normalizeSpaces } from '../text';\n\nexport default function cleanHeaders($article, $, title = '') {\n $(HEADER_TAG_LIST, $article).each((index, header) => {\n const $header = $(header);\n // Remove any headers that appear before all other p tags in the\n // document. This probably means that it was part of the title, a\n // subtitle or something else extraneous like a datestamp or byline,\n // all of which should be handled by other metadata handling.\n if ($($header, $article).prevAll('p').length === 0) {\n return $header.remove();\n }\n\n // Remove any headers that match the title exactly.\n if (normalizeSpaces($(header).text()) === title) {\n return $header.remove();\n }\n\n // If this header has a negative weight, it's probably junk.\n // Get rid of it.\n if (getWeight($(header)) < 0) {\n return $header.remove();\n }\n\n return $header;\n });\n\n return $;\n}\n","import { convertNodeTo } from 'utils/dom';\n\n// Rewrite the tag name to div if it's a top level node like body or\n// html to avoid later complications with multiple body tags.\nexport default function rewriteTopLevel(article, $) {\n // I'm not using context here because\n // it's problematic when converting the\n // top-level/root node - AP\n $ = convertNodeTo($('html'), $, 'div');\n $ = convertNodeTo($('body'), $, 'div');\n\n return $;\n}\n","import URL from 'url';\n\nfunction absolutize($, rootUrl, attr, $content) {\n $(`[${attr}]`, $content).each((_, node) => {\n const url = node.attribs[attr];\n const absoluteUrl = URL.resolve(rootUrl, url);\n\n node.attribs[attr] = absoluteUrl;\n });\n}\n\nexport default function makeLinksAbsolute($content, $, url) {\n ['href', 'src'].forEach(attr => absolutize($, url, attr, $content));\n\n return $content;\n}\n","\nexport function textLength(text) {\n return text.trim()\n .replace(/\\s+/g, ' ')\n .length;\n}\n\n// Determines what percentage of the text\n// in a node is link text\n// Takes a node, returns a float\nexport function linkDensity($node) {\n const totalTextLength = textLength($node.text());\n\n const linkText = $node.find('a').text();\n const linkLength = textLength(linkText);\n\n if (totalTextLength > 0) {\n return linkLength / totalTextLength;\n } else if (totalTextLength === 0 && linkLength > 0) {\n return 1;\n }\n\n return 0;\n}\n","import { stripTags } from 'utils/dom';\n\n// Given a node type to search for, and a list of meta tag names to\n// search for, find a meta tag associated.\nexport default function extractFromMeta(\n $,\n metaNames,\n cachedNames,\n cleanTags = true\n) {\n const foundNames = metaNames.filter(name => cachedNames.indexOf(name) !== -1);\n\n for (const name of foundNames) {\n const type = 'name';\n const value = 'value';\n\n const nodes = $(`meta[${type}=\"${name}\"]`);\n\n // Get the unique value of every matching node, in case there\n // are two meta tags with the same name and value.\n // Remove empty values.\n const values =\n nodes.map((index, node) => $(node).attr(value))\n .toArray()\n .filter(text => text !== '');\n\n // If we have more than one value for the same name, we have a\n // conflict and can't trust any of them. Skip this name. If we have\n // zero, that means our meta tags had no values. Skip this name\n // also.\n if (values.length === 1) {\n let metaValue;\n // Meta values that contain HTML should be stripped, as they\n // weren't subject to cleaning previously.\n if (cleanTags) {\n metaValue = stripTags(values[0], $);\n } else {\n metaValue = values[0];\n }\n\n return metaValue;\n }\n }\n\n // If nothing is found, return null\n return null;\n}\n","import { withinComment } from 'utils/dom';\n\nfunction isGoodNode($node, maxChildren) {\n // If it has a number of children, it's more likely a container\n // element. Skip it.\n if ($node.children().length > maxChildren) {\n return false;\n }\n // If it looks to be within a comment, skip it.\n if (withinComment($node)) {\n return false;\n }\n\n return true;\n}\n\n// Given a a list of selectors find content that may\n// be extractable from the document. This is for flat\n// meta-information, like author, title, date published, etc.\nexport default function extractFromSelectors(\n $,\n selectors,\n maxChildren = 1,\n textOnly = true\n) {\n for (const selector of selectors) {\n const nodes = $(selector);\n\n // If we didn't get exactly one of this selector, this may be\n // a list of articles or comments. Skip it.\n if (nodes.length === 1) {\n const $node = $(nodes[0]);\n\n if (isGoodNode($node, maxChildren)) {\n let content;\n if (textOnly) {\n content = $node.text();\n } else {\n content = $node.html();\n }\n\n if (content) {\n return content;\n }\n }\n }\n }\n\n return null;\n}\n","// strips all tags from a string of text\nexport default function stripTags(text, $) {\n // Wrapping text in html element prevents errors when text\n // has no html\n const cleanText = $(`<span>${text}</span>`).text();\n return cleanText === '' ? text : cleanText;\n}\n","export default function withinComment($node) {\n const parents = $node.parents().toArray();\n const commentParent = parents.find((parent) => {\n const classAndId = `${parent.attribs.class} ${parent.attribs.id}`;\n return classAndId.includes('comment');\n });\n\n return commentParent !== undefined;\n}\n","// Given a node, determine if it's article-like enough to return\n// param: node (a cheerio node)\n// return: boolean\n\nexport default function nodeIsSufficient($node) {\n return $node.text().trim().length >= 100;\n}\n","import { IS_WP_SELECTOR } from './constants';\n\nexport default function isWordpress($) {\n return $(IS_WP_SELECTOR).length > 0;\n}\n","// CLEAN AUTHOR CONSTANTS\nexport const CLEAN_AUTHOR_RE = /^\\s*(posted |written )?by\\s*:?\\s*(.*)/i;\n // author = re.sub(r'^\\s*(posted |written )?by\\s*:?\\s*(.*)(?i)',\n\n// CLEAN DEK CONSTANTS\nexport const TEXT_LINK_RE = new RegExp('http(s)?://', 'i');\n// An ordered list of meta tag names that denote likely article deks.\n// From most distinct to least distinct.\n//\n// NOTE: There are currently no meta tags that seem to provide the right\n// content consistenty enough. Two options were:\n// - og:description\n// - dc.description\n// However, these tags often have SEO-specific junk in them that's not\n// header-worthy like a dek is. Excerpt material at best.\nexport const DEK_META_TAGS = [\n];\n\n// An ordered list of Selectors to find likely article deks. From\n// most explicit to least explicit.\n//\n// Should be more restrictive than not, as a failed dek can be pretty\n// detrimental to the aesthetics of an article.\nexport const DEK_SELECTORS = [\n '.entry-summary',\n];\n\n// CLEAN DATE PUBLISHED CONSTANTS\nexport const MS_DATE_STRING = /^\\d{13}$/i;\nexport const SEC_DATE_STRING = /^\\d{10}$/i;\nexport const CLEAN_DATE_STRING_RE = /^\\s*published\\s*:?\\s*(.*)/i;\nexport const TIME_MERIDIAN_SPACE_RE = /(.*\\d)(am|pm)(.*)/i;\nexport const TIME_MERIDIAN_DOTS_RE = /\\.m\\./i;\nconst months = [\n 'jan',\n 'feb',\n 'mar',\n 'apr',\n 'may',\n 'jun',\n 'jul',\n 'aug',\n 'sep',\n 'oct',\n 'nov',\n 'dec',\n];\nconst allMonths = months.join('|');\nconst timestamp1 = '[0-9]{1,2}:[0-9]{2,2}( ?[ap].?m.?)?';\nconst timestamp2 = '[0-9]{1,2}[/-][0-9]{1,2}[/-][0-9]{2,4}';\nexport const SPLIT_DATE_STRING =\n new RegExp(`(${timestamp1})|(${timestamp2})|([0-9]{1,4})|(${allMonths})`, 'ig');\n\n// CLEAN TITLE CONSTANTS\n// A regular expression that will match separating characters on a\n// title, that usually denote breadcrumbs or something similar.\nexport const TITLE_SPLITTERS_RE = /(: | - | \\| )/g;\n\nexport const DOMAIN_ENDINGS_RE =\n new RegExp('.com$|.net$|.org$|.co.uk$', 'g');\n","import { CLEAN_AUTHOR_RE } from './constants';\n\n// Take an author string (like 'By David Smith ') and clean it to\n// just the name(s): 'David Smith'.\nexport default function cleanAuthor(author) {\n return author.replace(CLEAN_AUTHOR_RE, '$2').trim();\n}\n","import validUrl from 'valid-url';\n\nexport default function clean(leadImageUrl) {\n leadImageUrl = leadImageUrl.trim();\n if (validUrl.isWebUri(leadImageUrl)) {\n return leadImageUrl;\n }\n\n return null;\n}\n","import { stripTags } from 'utils/dom';\nimport { excerptContent } from 'utils/text';\n\nimport { TEXT_LINK_RE } from './constants';\n\n// Take a dek HTML fragment, and return the cleaned version of it.\n// Return None if the dek wasn't good enough.\nexport default function cleanDek(dek, { $, excerpt }) {\n // Sanity check that we didn't get too short or long of a dek.\n if (dek.length > 1000 || dek.length < 5) return null;\n\n // Check that dek isn't the same as excerpt\n if (excerpt && excerptContent(excerpt, 10) === excerptContent(dek, 10)) return null;\n\n const dekText = stripTags(dek, $);\n\n // Plain text links shouldn't exist in the dek. If we have some, it's\n // not a good dek - bail.\n if (TEXT_LINK_RE.test(dekText)) return null;\n\n return dekText.trim();\n}\n","import moment from 'moment';\n// Is there a compelling reason to use moment here?\n// Mostly only being used for the isValid() method,\n// but could just check for 'Invalid Date' string.\n\nimport {\n MS_DATE_STRING,\n SEC_DATE_STRING,\n CLEAN_DATE_STRING_RE,\n SPLIT_DATE_STRING,\n TIME_MERIDIAN_SPACE_RE,\n TIME_MERIDIAN_DOTS_RE,\n} from './constants';\n\nexport function cleanDateString(dateString) {\n return (dateString.match(SPLIT_DATE_STRING) || [])\n .join(' ')\n .replace(TIME_MERIDIAN_DOTS_RE, 'm')\n .replace(TIME_MERIDIAN_SPACE_RE, '$1 $2 $3')\n .replace(CLEAN_DATE_STRING_RE, '$1')\n .trim();\n}\n\n// Take a date published string, and hopefully return a date out of\n// it. Return none if we fail.\nexport default function cleanDatePublished(dateString) {\n // If string is in milliseconds or seconds, convert to int\n if (MS_DATE_STRING.test(dateString) || SEC_DATE_STRING.test(dateString)) {\n dateString = parseInt(dateString, 10);\n }\n\n let date = moment(new Date(dateString));\n\n if (!date.isValid()) {\n dateString = cleanDateString(dateString);\n date = moment(new Date(dateString));\n }\n\n return date.isValid() ? date.toISOString() : null;\n}\n","import {\n cleanAttributes,\n cleanHeaders,\n cleanHOnes,\n cleanImages,\n cleanTags,\n removeEmpty,\n rewriteTopLevel,\n markToKeep,\n stripJunkTags,\n makeLinksAbsolute,\n} from 'utils/dom';\n\n// Clean our article content, returning a new, cleaned node.\nexport default function extractCleanNode(\n article,\n {\n $,\n cleanConditionally = true,\n title = '',\n url = '',\n defaultCleaner = true,\n }\n) {\n // Rewrite the tag name to div if it's a top level node like body or\n // html to avoid later complications with multiple body tags.\n rewriteTopLevel(article, $);\n\n // Drop small images and spacer images\n // Only do this is defaultCleaner is set to true;\n // this can sometimes be too aggressive.\n if (defaultCleaner) cleanImages(article, $);\n\n // Mark elements to keep that would normally be removed.\n // E.g., stripJunkTags will remove iframes, so we're going to mark\n // YouTube/Vimeo videos as elements we want to keep.\n markToKeep(article, $);\n\n // Drop certain tags like <title>, etc\n // This is -mostly- for cleanliness, not security.\n stripJunkTags(article, $);\n\n // H1 tags are typically the article title, which should be extracted\n // by the title extractor instead. If there's less than 3 of them (<3),\n // strip them. Otherwise, turn 'em into H2s.\n cleanHOnes(article, $);\n\n // Clean headers\n cleanHeaders(article, $, title);\n\n // Make links absolute\n makeLinksAbsolute(article, $, url);\n\n // We used to clean UL's and OL's here, but it was leading to\n // too many in-article lists being removed. Consider a better\n // way to detect menus particularly and remove them.\n // Also optionally running, since it can be overly aggressive.\n if (defaultCleaner) cleanTags(article, $, cleanConditionally);\n\n // Remove empty paragraph nodes\n removeEmpty(article, $);\n\n // Remove unnecessary attributes\n cleanAttributes(article, $);\n\n return article;\n}\n","import { stripTags } from 'utils/dom';\n\nimport { TITLE_SPLITTERS_RE } from './constants';\nimport { resolveSplitTitle } from './index';\n\nexport default function cleanTitle(title, { url, $ }) {\n // If title has |, :, or - in it, see if\n // we can clean it up.\n if (TITLE_SPLITTERS_RE.test(title)) {\n title = resolveSplitTitle(title, url);\n }\n\n // Final sanity check that we didn't get a crazy title.\n // if (title.length > 150 || title.length < 15) {\n if (title.length > 150) {\n // If we did, return h1 from the document if it exists\n const h1 = $('h1');\n if (h1.length === 1) {\n title = h1.text();\n }\n }\n\n // strip any html tags in the title text\n return stripTags(title, $).trim();\n}\n\n","import URL from 'url';\nimport wuzzy from 'wuzzy';\n\nimport {\n TITLE_SPLITTERS_RE,\n DOMAIN_ENDINGS_RE,\n} from './constants';\n\nfunction extractBreadcrumbTitle(splitTitle, text) {\n // This must be a very breadcrumbed title, like:\n // The Best Gadgets on Earth : Bits : Blogs : NYTimes.com\n // NYTimes - Blogs - Bits - The Best Gadgets on Earth\n if (splitTitle.length >= 6) {\n // Look to see if we can find a breadcrumb splitter that happens\n // more than once. If we can, we'll be able to better pull out\n // the title.\n const termCounts = splitTitle.reduce((acc, titleText) => {\n acc[titleText] = acc[titleText] ? acc[titleText] + 1 : 1;\n return acc;\n }, {});\n\n const [maxTerm, termCount] =\n Reflect.ownKeys(termCounts)\n .reduce((acc, key) => {\n if (acc[1] < termCounts[key]) {\n return [key, termCounts[key]];\n }\n\n return acc;\n }, [0, 0]);\n\n // We found a splitter that was used more than once, so it\n // is probably the breadcrumber. Split our title on that instead.\n // Note: max_term should be <= 4 characters, so that \" >> \"\n // will match, but nothing longer than that.\n if (termCount >= 2 && maxTerm.length <= 4) {\n splitTitle = text.split(maxTerm);\n }\n\n const splitEnds = [splitTitle[0], splitTitle.slice(-1)];\n const longestEnd = splitEnds.reduce((acc, end) => acc.length > end.length ? acc : end, '');\n\n if (longestEnd.length > 10) {\n return longestEnd;\n }\n\n return text;\n }\n\n return null;\n}\n\nfunction cleanDomainFromTitle(splitTitle, url) {\n // Search the ends of the title, looking for bits that fuzzy match\n // the URL too closely. If one is found, discard it and return the\n // rest.\n //\n // Strip out the big TLDs - it just makes the matching a bit more\n // accurate. Not the end of the world if it doesn't strip right.\n const { host } = URL.parse(url);\n const nakedDomain = host.replace(DOMAIN_ENDINGS_RE, '');\n\n const startSlug = splitTitle[0].toLowerCase().replace(' ', '');\n const startSlugRatio = wuzzy.levenshtein(startSlug, nakedDomain);\n\n if (startSlugRatio > 0.4 && startSlug.length > 5) {\n return splitTitle.slice(2).join('');\n }\n\n const endSlug = splitTitle.slice(-1)[0].toLowerCase().replace(' ', '');\n const endSlugRatio = wuzzy.levenshtein(endSlug, nakedDomain);\n\n if (endSlugRatio > 0.4 && endSlug.length >= 5) {\n return splitTitle.slice(0, -2).join('');\n }\n\n return null;\n}\n\n// Given a title with separators in it (colons, dashes, etc),\n// resolve whether any of the segments should be removed.\nexport default function resolveSplitTitle(title, url = '') {\n // Splits while preserving splitters, like:\n // ['The New New York', ' - ', 'The Washington Post']\n const splitTitle = title.split(TITLE_SPLITTERS_RE);\n if (splitTitle.length === 1) {\n return title;\n }\n\n let newTitle = extractBreadcrumbTitle(splitTitle, title);\n if (newTitle) return newTitle;\n\n newTitle = cleanDomainFromTitle(splitTitle, url);\n if (newTitle) return newTitle;\n\n // Fuzzy ratio didn't find anything, so this title is probably legit.\n // Just return it all.\n return title;\n}\n","import cleanAuthor from './author';\nimport cleanImage from './lead-image-url';\nimport cleanDek from './dek';\nimport cleanDatePublished from './date-published';\nimport cleanContent from './content';\nimport cleanTitle from './title';\n\nconst Cleaners = {\n author: cleanAuthor,\n lead_image_url: cleanImage,\n dek: cleanDek,\n date_published: cleanDatePublished,\n content: cleanContent,\n title: cleanTitle,\n};\n\n\nexport default Cleaners;\n\nexport { cleanAuthor };\nexport { cleanImage };\nexport { cleanDek };\nexport { cleanDatePublished };\nexport { cleanContent };\nexport { cleanTitle };\nexport { default as resolveSplitTitle } from './resolve-split-title';\n","import {\n stripUnlikelyCandidates,\n convertToParagraphs,\n} from 'utils/dom';\n\nimport {\n scoreContent,\n findTopCandidate,\n} from './scoring';\n\n// Using a variety of scoring techniques, extract the content most\n// likely to be article text.\n//\n// If strip_unlikely_candidates is True, remove any elements that\n// match certain criteria first. (Like, does this element have a\n// classname of \"comment\")\n//\n// If weight_nodes is True, use classNames and IDs to determine the\n// worthiness of nodes.\n//\n// Returns a cheerio object $\nexport default function extractBestNode($, opts) {\n // clone the node so we can get back to our\n // initial parsed state if needed\n // TODO Do I need this? – AP\n // let $root = $.root().clone()\n\n\n if (opts.stripUnlikelyCandidates) {\n $ = stripUnlikelyCandidates($);\n }\n\n $ = convertToParagraphs($);\n $ = scoreContent($, opts.weightNodes);\n const $topCandidate = findTopCandidate($);\n\n return $topCandidate;\n}\n","import cheerio from 'cheerio';\n\nimport { nodeIsSufficient } from 'utils/dom';\nimport { cleanContent } from 'cleaners';\nimport { normalizeSpaces } from 'utils/text';\n\nimport extractBestNode from './extract-best-node';\n\nconst GenericContentExtractor = {\n defaultOpts: {\n stripUnlikelyCandidates: true,\n weightNodes: true,\n cleanConditionally: true,\n },\n\n // Extract the content for this resource - initially, pass in our\n // most restrictive opts which will return the highest quality\n // content. On each failure, retry with slightly more lax opts.\n //\n // :param return_type: string. If \"node\", should return the content\n // as a cheerio node rather than as an HTML string.\n //\n // Opts:\n // stripUnlikelyCandidates: Remove any elements that match\n // non-article-like criteria first.(Like, does this element\n // have a classname of \"comment\")\n //\n // weightNodes: Modify an elements score based on whether it has\n // certain classNames or IDs. Examples: Subtract if a node has\n // a className of 'comment', Add if a node has an ID of\n // 'entry-content'.\n //\n // cleanConditionally: Clean the node to return of some\n // superfluous content. Things like forms, ads, etc.\n extract({ $, html, title, url }, opts) {\n opts = { ...this.defaultOpts, ...opts };\n\n $ = $ || cheerio.load(html);\n\n // Cascade through our extraction-specific opts in an ordered fashion,\n // turning them off as we try to extract content.\n let node = this.getContentNode($, title, url, opts);\n\n if (nodeIsSufficient(node)) {\n return this.cleanAndReturnNode(node, $);\n }\n\n // We didn't succeed on first pass, one by one disable our\n // extraction opts and try again.\n for (const key of Reflect.ownKeys(opts).filter(k => opts[k] === true)) {\n opts[key] = false;\n $ = cheerio.load(html);\n\n node = this.getContentNode($, title, url, opts);\n\n if (nodeIsSufficient(node)) {\n break;\n }\n }\n\n return this.cleanAndReturnNode(node, $);\n },\n\n // Get node given current options\n getContentNode($, title, url, opts) {\n return cleanContent(\n extractBestNode($, opts),\n {\n $,\n cleanConditionally: opts.cleanConditionally,\n title,\n url,\n });\n },\n\n // Once we got here, either we're at our last-resort node, or\n // we broke early. Make sure we at least have -something- before we\n // move forward.\n cleanAndReturnNode(node, $) {\n if (!node) {\n return null;\n }\n\n return normalizeSpaces($.html(node));\n\n // if return_type == \"html\":\n // return normalize_spaces(node_to_html(node))\n // else:\n // return node\n },\n\n};\n\nexport default GenericContentExtractor;\n","// TODO: It would be great if we could merge the meta and selector lists into\n// a list of objects, because we could then rank them better. For example,\n// .hentry .entry-title is far better suited than <meta title>.\n\n// An ordered list of meta tag names that denote likely article titles. All\n// attributes should be lowercase for faster case-insensitive matching. From\n// most distinct to least distinct.\nexport const STRONG_TITLE_META_TAGS = [\n 'tweetmeme-title',\n 'dc.title',\n 'rbtitle',\n 'headline',\n 'title',\n];\n\n// og:title is weak because it typically contains context that we don't like,\n// for example the source site's name. Gotta get that brand into facebook!\nexport const WEAK_TITLE_META_TAGS = [\n 'og:title',\n];\n\n// An ordered list of XPath Selectors to find likely article titles. From\n// most explicit to least explicit.\n//\n// Note - this does not use classes like CSS. This checks to see if the string\n// exists in the className, which is not as accurate as .className (which\n// splits on spaces/endlines), but for our purposes it's close enough. The\n// speed tradeoff is worth the accuracy hit.\nexport const STRONG_TITLE_SELECTORS = [\n '.hentry .entry-title',\n 'h1#articleHeader',\n 'h1.articleHeader',\n 'h1.article',\n '.instapaper_title',\n '#meebo-title',\n];\n\nexport const WEAK_TITLE_SELECTORS = [\n 'article h1',\n '#entry-title',\n '.entry-title',\n '#entryTitle',\n '#entrytitle',\n '.entryTitle',\n '.entrytitle',\n '#articleTitle',\n '.articleTitle',\n 'post post-title',\n 'h1.title',\n 'h2.article',\n 'h1',\n 'html head title',\n 'title',\n];\n","import { cleanTitle } from 'cleaners';\nimport {\n extractFromMeta,\n extractFromSelectors,\n} from 'utils/dom';\n\nimport {\n STRONG_TITLE_META_TAGS,\n WEAK_TITLE_META_TAGS,\n STRONG_TITLE_SELECTORS,\n WEAK_TITLE_SELECTORS,\n} from './constants';\n\nconst GenericTitleExtractor = {\n extract({ $, url, metaCache }) {\n // First, check to see if we have a matching meta tag that we can make\n // use of that is strongly associated with the headline.\n let title;\n\n title = extractFromMeta($, STRONG_TITLE_META_TAGS, metaCache);\n if (title) return cleanTitle(title, { url, $ });\n\n // Second, look through our content selectors for the most likely\n // article title that is strongly associated with the headline.\n title = extractFromSelectors($, STRONG_TITLE_SELECTORS);\n if (title) return cleanTitle(title, { url, $ });\n\n // Third, check for weaker meta tags that may match.\n title = extractFromMeta($, WEAK_TITLE_META_TAGS, metaCache);\n if (title) return cleanTitle(title, { url, $ });\n\n // Last, look for weaker selector tags that may match.\n title = extractFromSelectors($, WEAK_TITLE_SELECTORS);\n if (title) return cleanTitle(title, { url, $ });\n\n // If no matches, return an empty string\n return '';\n },\n};\n\nexport default GenericTitleExtractor;\n","// An ordered list of meta tag names that denote likely article authors. All\n// attributes should be lowercase for faster case-insensitive matching. From\n// most distinct to least distinct.\n//\n// Note: \"author\" is too often the -developer- of the page, so it is not\n// added here.\nexport const AUTHOR_META_TAGS = [\n 'byl',\n 'clmst',\n 'dc.author',\n 'dcsext.author',\n 'dc.creator',\n 'rbauthors',\n 'authors',\n];\n\nexport const AUTHOR_MAX_LENGTH = 300;\n\n// An ordered list of XPath Selectors to find likely article authors. From\n// most explicit to least explicit.\n//\n// Note - this does not use classes like CSS. This checks to see if the string\n// exists in the className, which is not as accurate as .className (which\n// splits on spaces/endlines), but for our purposes it's close enough. The\n// speed tradeoff is worth the accuracy hit.\nexport const AUTHOR_SELECTORS = [\n '.entry .entry-author',\n '.author.vcard .fn',\n '.author .vcard .fn',\n '.byline.vcard .fn',\n '.byline .vcard .fn',\n '.byline .by .author',\n '.byline .by',\n '.byline .author',\n '.post-author.vcard',\n '.post-author .vcard',\n 'a[rel=author]',\n '#by_author',\n '.by_author',\n '#entryAuthor',\n '.entryAuthor',\n '.byline a[href*=author]',\n '#author .authorname',\n '.author .authorname',\n '#author',\n '.author',\n '.articleauthor',\n '.ArticleAuthor',\n '.byline',\n];\n\n// An ordered list of Selectors to find likely article authors, with\n// regular expression for content.\nconst bylineRe = /^[\\n\\s]*By/i;\nexport const BYLINE_SELECTORS_RE = [\n ['#byline', bylineRe],\n ['.byline', bylineRe],\n];\n","import { cleanAuthor } from 'cleaners';\nimport {\n extractFromMeta,\n extractFromSelectors,\n} from 'utils/dom';\n\nimport {\n AUTHOR_META_TAGS,\n AUTHOR_MAX_LENGTH,\n AUTHOR_SELECTORS,\n BYLINE_SELECTORS_RE,\n} from './constants';\n\nconst GenericAuthorExtractor = {\n extract({ $, metaCache }) {\n let author;\n\n // First, check to see if we have a matching\n // meta tag that we can make use of.\n author = extractFromMeta($, AUTHOR_META_TAGS, metaCache);\n if (author && author.length < AUTHOR_MAX_LENGTH) {\n return cleanAuthor(author);\n }\n\n // Second, look through our selectors looking for potential authors.\n author = extractFromSelectors($, AUTHOR_SELECTORS, 2);\n if (author && author.length < AUTHOR_MAX_LENGTH) {\n return cleanAuthor(author);\n }\n\n // Last, use our looser regular-expression based selectors for\n // potential authors.\n for (const [selector, regex] of BYLINE_SELECTORS_RE) {\n const node = $(selector);\n if (node.length === 1) {\n const text = node.text();\n if (regex.test(text)) {\n return cleanAuthor(text);\n }\n }\n }\n\n return null;\n },\n};\n\nexport default GenericAuthorExtractor;\n\n","// An ordered list of meta tag names that denote\n// likely date published dates. All attributes\n// should be lowercase for faster case-insensitive matching.\n// From most distinct to least distinct.\nexport const DATE_PUBLISHED_META_TAGS = [\n 'article:published_time',\n 'displaydate',\n 'dc.date',\n 'dc.date.issued',\n 'rbpubdate',\n 'publish_date',\n 'pub_date',\n 'pagedate',\n 'pubdate',\n 'revision_date',\n 'doc_date',\n 'date_created',\n 'content_create_date',\n 'lastmodified',\n 'created',\n 'date',\n];\n\n// An ordered list of XPath Selectors to find\n// likely date published dates. From most explicit\n// to least explicit.\nexport const DATE_PUBLISHED_SELECTORS = [\n '.hentry .dtstamp.published',\n '.hentry .published',\n '.hentry .dtstamp.updated',\n '.hentry .updated',\n '.single .published',\n '.meta .published',\n '.meta .postDate',\n '.entry-date',\n '.byline .date',\n '.postmetadata .date',\n '.article_datetime',\n '.date-header',\n '.story-date',\n '.dateStamp',\n '#story .datetime',\n '.dateline',\n '.pubdate',\n];\n\n// An ordered list of compiled regular expressions to find likely date\n// published dates from the URL. These should always have the first\n// reference be a date string that is parseable by dateutil.parser.parse\nconst abbrevMonthsStr = '(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)';\nexport const DATE_PUBLISHED_URL_RES = [\n // /2012/01/27/ but not /2012/01/293\n new RegExp('/(20\\\\d{2}/\\\\d{2}/\\\\d{2})/', 'i'),\n // 20120127 or 20120127T but not 2012012733 or 8201201733\n // /[^0-9](20\\d{2}[01]\\d[0-3]\\d)([^0-9]|$)/i,\n // 2012-01-27\n new RegExp('(20\\\\d{2}-[01]\\\\d-[0-3]\\\\d)', 'i'),\n // /2012/jan/27/\n new RegExp(`/(20\\\\d{2}/${abbrevMonthsStr}/[0-3]\\\\d)/`, 'i'),\n];\n\n","import { cleanDatePublished } from 'cleaners';\nimport {\n extractFromMeta,\n extractFromSelectors,\n} from 'utils/dom';\nimport { extractFromUrl } from 'utils/text';\n\nimport {\n DATE_PUBLISHED_META_TAGS,\n DATE_PUBLISHED_SELECTORS,\n DATE_PUBLISHED_URL_RES,\n} from './constants';\n\nconst GenericDatePublishedExtractor = {\n extract({ $, url, metaCache }) {\n let datePublished;\n // First, check to see if we have a matching meta tag\n // that we can make use of.\n // Don't try cleaning tags from this string\n datePublished = extractFromMeta($, DATE_PUBLISHED_META_TAGS, metaCache, false);\n if (datePublished) return cleanDatePublished(datePublished);\n\n // Second, look through our selectors looking for potential\n // date_published's.\n datePublished = extractFromSelectors($, DATE_PUBLISHED_SELECTORS);\n if (datePublished) return cleanDatePublished(datePublished);\n\n // Lastly, look to see if a dately string exists in the URL\n datePublished = extractFromUrl(url, DATE_PUBLISHED_URL_RES);\n if (datePublished) return cleanDatePublished(datePublished);\n\n return null;\n },\n};\n\nexport default GenericDatePublishedExtractor;\n","// import {\n// DEK_META_TAGS,\n// DEK_SELECTORS,\n// DEK_URL_RES,\n// } from './constants';\n\n// import { cleanDek } from 'cleaners';\n\n// import {\n// extractFromMeta,\n// extractFromSelectors,\n// } from 'utils/dom';\n\n// Currently there is only one selector for\n// deks. We should simply return null here\n// until we have a more robust generic option.\n// Below is the original source for this, for reference.\nconst GenericDekExtractor = {\n // extract({ $, content, metaCache }) {\n extract() {\n return null;\n },\n};\n\nexport default GenericDekExtractor;\n\n// def extract_dek(self):\n// # First, check to see if we have a matching meta tag that we can make\n// # use of.\n// dek = self.extract_from_meta('dek', constants.DEK_META_TAGS)\n// if not dek:\n// # Second, look through our CSS/XPath selectors. This may return\n// # an HTML fragment.\n// dek = self.extract_from_selectors('dek',\n// constants.DEK_SELECTORS,\n// text_only=False)\n//\n// if dek:\n// # Make sure our dek isn't in the first few thousand characters\n// # of the content, otherwise it's just the start of the article\n// # and not a true dek.\n// content = self.extract_content()\n// content_chunk = normalize_spaces(strip_tags(content[:2000]))\n// dek_chunk = normalize_spaces(dek[:100]) # Already has no tags.\n//\n// # 80% or greater similarity means the dek was very similar to some\n// # of the starting content, so we skip it.\n// if fuzz.partial_ratio(content_chunk, dek_chunk) < 80:\n// return dek\n//\n// return None\n","// An ordered list of meta tag names that denote likely article leading images.\n// All attributes should be lowercase for faster case-insensitive matching.\n// From most distinct to least distinct.\nexport const LEAD_IMAGE_URL_META_TAGS = [\n 'og:image',\n 'twitter:image',\n 'image_src',\n];\n\nexport const LEAD_IMAGE_URL_SELECTORS = [\n 'link[rel=image_src]',\n];\n\nexport const POSITIVE_LEAD_IMAGE_URL_HINTS = [\n 'upload',\n 'wp-content',\n 'large',\n 'photo',\n 'wp-image',\n];\nexport const POSITIVE_LEAD_IMAGE_URL_HINTS_RE = new RegExp(POSITIVE_LEAD_IMAGE_URL_HINTS.join('|'), 'i');\n\nexport const NEGATIVE_LEAD_IMAGE_URL_HINTS = [\n 'spacer',\n 'sprite',\n 'blank',\n 'throbber',\n 'gradient',\n 'tile',\n 'bg',\n 'background',\n 'icon',\n 'social',\n 'header',\n 'hdr',\n 'advert',\n 'spinner',\n 'loader',\n 'loading',\n 'default',\n 'rating',\n 'share',\n 'facebook',\n 'twitter',\n 'theme',\n 'promo',\n 'ads',\n 'wp-includes',\n];\nexport const NEGATIVE_LEAD_IMAGE_URL_HINTS_RE = new RegExp(NEGATIVE_LEAD_IMAGE_URL_HINTS.join('|'), 'i');\n\nexport const GIF_RE = /\\.gif(\\?.*)?$/i;\nexport const JPG_RE = /\\.jpe?g(\\?.*)?$/i;\n","import {\n POSITIVE_LEAD_IMAGE_URL_HINTS_RE,\n NEGATIVE_LEAD_IMAGE_URL_HINTS_RE,\n GIF_RE,\n JPG_RE,\n} from './constants';\n\nimport { PHOTO_HINTS_RE } from '../content/scoring/constants';\n\nfunction getSig($node) {\n return `${$node.attr('class') || ''} ${$node.attr('id') || ''}`;\n}\n\n// Scores image urls based on a variety of heuristics.\nexport function scoreImageUrl(url) {\n url = url.trim();\n let score = 0;\n\n if (POSITIVE_LEAD_IMAGE_URL_HINTS_RE.test(url)) {\n score += 20;\n }\n\n if (NEGATIVE_LEAD_IMAGE_URL_HINTS_RE.test(url)) {\n score -= 20;\n }\n\n // TODO: We might want to consider removing this as\n // gifs are much more common/popular than they once were\n if (GIF_RE.test(url)) {\n score -= 10;\n }\n\n if (JPG_RE.test(url)) {\n score += 10;\n }\n\n // PNGs are neutral.\n\n return score;\n}\n\n// Alt attribute usually means non-presentational image.\nexport function scoreAttr($img) {\n if ($img.attr('alt')) {\n return 5;\n }\n\n return 0;\n}\n\n// Look through our parent and grandparent for figure-like\n// container elements, give a bonus if we find them\nexport function scoreByParents($img) {\n let score = 0;\n const $figParent = $img.parents('figure').first();\n\n if ($figParent.length === 1) {\n score += 25;\n }\n\n const $parent = $img.parent();\n let $gParent;\n if ($parent.length === 1) {\n $gParent = $parent.parent();\n }\n\n [$parent, $gParent].forEach(($node) => {\n if (PHOTO_HINTS_RE.test(getSig($node))) {\n score += 15;\n }\n });\n\n return score;\n}\n\n// Look at our immediate sibling and see if it looks like it's a\n// caption. Bonus if so.\nexport function scoreBySibling($img) {\n let score = 0;\n const $sibling = $img.next();\n const sibling = $sibling.get(0);\n\n if (sibling && sibling.tagName === 'figcaption') {\n score += 25;\n }\n\n if (PHOTO_HINTS_RE.test(getSig($sibling))) {\n score += 15;\n }\n\n return score;\n}\n\nexport function scoreByDimensions($img) {\n let score = 0;\n\n const width = parseFloat($img.attr('width'));\n const height = parseFloat($img.attr('height'));\n const src = $img.attr('src');\n\n // Penalty for skinny images\n if (width && width <= 50) {\n score -= 50;\n }\n\n // Penalty for short images\n if (height && height <= 50) {\n score -= 50;\n }\n\n if (width && height && !src.includes('sprite')) {\n const area = width * height;\n if (area < 5000) { // Smaller than 50 x 100\n score -= 100;\n } else {\n score += Math.round(area / 1000);\n }\n }\n\n return score;\n}\n\nexport function scoreByPosition($imgs, index) {\n return ($imgs.length / 2) - index;\n}\n","import { extractFromMeta } from 'utils/dom';\nimport { cleanImage } from 'cleaners';\n\nimport {\n LEAD_IMAGE_URL_META_TAGS,\n LEAD_IMAGE_URL_SELECTORS,\n} from './constants';\n\nimport {\n scoreImageUrl,\n scoreAttr,\n scoreByParents,\n scoreBySibling,\n scoreByDimensions,\n scoreByPosition,\n} from './score-image';\n\n// Given a resource, try to find the lead image URL from within\n// it. Like content and next page extraction, uses a scoring system\n// to determine what the most likely image may be. Short circuits\n// on really probable things like og:image meta tags.\n//\n// Potential signals to still take advantage of:\n// * domain\n// * weird aspect ratio\nconst GenericLeadImageUrlExtractor = {\n extract({ $, content, metaCache }) {\n let cleanUrl;\n\n // Check to see if we have a matching meta tag that we can make use of.\n // Moving this higher because common practice is now to use large\n // images on things like Open Graph or Twitter cards.\n // images usually have for things like Open Graph.\n const imageUrl =\n extractFromMeta(\n $,\n LEAD_IMAGE_URL_META_TAGS,\n metaCache,\n false\n );\n\n if (imageUrl) {\n cleanUrl = cleanImage(imageUrl);\n\n if (cleanUrl) return cleanUrl;\n }\n\n // Next, try to find the \"best\" image via the content.\n // We'd rather not have to fetch each image and check dimensions,\n // so try to do some analysis and determine them instead.\n const imgs = $('img', content).toArray();\n const imgScores = {};\n\n imgs.forEach((img, index) => {\n const $img = $(img);\n const src = $img.attr('src');\n\n if (!src) return;\n\n let score = scoreImageUrl(src);\n score += scoreAttr($img);\n score += scoreByParents($img);\n score += scoreBySibling($img);\n score += scoreByDimensions($img);\n score += scoreByPosition(imgs, index);\n\n imgScores[src] = score;\n });\n\n const [topUrl, topScore] =\n Reflect.ownKeys(imgScores).reduce((acc, key) =>\n imgScores[key] > acc[1] ? [key, imgScores[key]] : acc\n , [null, 0]);\n\n if (topScore > 0) {\n cleanUrl = cleanImage(topUrl);\n\n if (cleanUrl) return cleanUrl;\n }\n\n // If nothing else worked, check to see if there are any really\n // probable nodes in the doc, like <link rel=\"image_src\" />.\n for (const selector of LEAD_IMAGE_URL_SELECTORS) {\n const $node = $(selector).first();\n const src = $node.attr('src');\n if (src) {\n cleanUrl = cleanImage(src);\n if (cleanUrl) return cleanUrl;\n }\n\n const href = $node.attr('href');\n if (href) {\n cleanUrl = cleanImage(href);\n if (cleanUrl) return cleanUrl;\n }\n\n const value = $node.attr('value');\n if (value) {\n cleanUrl = cleanImage(value);\n if (cleanUrl) return cleanUrl;\n }\n }\n\n return null;\n },\n};\n\nexport default GenericLeadImageUrlExtractor;\n\n// def extract(self):\n// \"\"\"\n// # First, try to find the \"best\" image via the content.\n// # We'd rather not have to fetch each image and check dimensions,\n// # so try to do some analysis and determine them instead.\n// content = self.extractor.extract_content(return_type=\"node\")\n// imgs = content.xpath('.//img')\n// img_scores = defaultdict(int)\n// logger.debug('Scoring %d images from content', len(imgs))\n// for (i, img) in enumerate(imgs):\n// img_score = 0\n//\n// if not 'src' in img.attrib:\n// logger.debug('No src attribute found')\n// continue\n//\n// try:\n// parsed_img = urlparse(img.attrib['src'])\n// img_path = parsed_img.path.lower()\n// except ValueError:\n// logger.debug('ValueError getting img path.')\n// continue\n// logger.debug('Image path is %s', img_path)\n//\n// if constants.POSITIVE_LEAD_IMAGE_URL_HINTS_RE.match(img_path):\n// logger.debug('Positive URL hints match. Adding 20.')\n// img_score += 20\n//\n// if constants.NEGATIVE_LEAD_IMAGE_URL_HINTS_RE.match(img_path):\n// logger.debug('Negative URL hints match. Subtracting 20.')\n// img_score -= 20\n//\n// # Gifs are more often structure than photos\n// if img_path.endswith('gif'):\n// logger.debug('gif found. Subtracting 10.')\n// img_score -= 10\n//\n// # JPGs are more often photographs\n// if img_path.endswith('jpg'):\n// logger.debug('jpg found. Adding 10.')\n// img_score += 10\n//\n// # PNGs are neutral.\n//\n// # Alt attribute usually means non-presentational image.\n// if 'alt' in img.attrib and len(img.attrib['alt']) > 5:\n// logger.debug('alt attribute found. Adding 5.')\n// img_score += 5\n//\n// # Look through our parent and grandparent for figure-like\n// # container elements, give a bonus if we find them\n// parents = [img.getparent()]\n// if parents[0] is not None and parents[0].getparent() is not None:\n// parents.append(parents[0].getparent())\n// for p in parents:\n// if p.tag == 'figure':\n// logger.debug('Parent with <figure> tag found. Adding 25.')\n// img_score += 25\n//\n// p_sig = ' '.join([p.get('id', ''), p.get('class', '')])\n// if constants.PHOTO_HINTS_RE.search(p_sig):\n// logger.debug('Photo hints regex match. Adding 15.')\n// img_score += 15\n//\n// # Look at our immediate sibling and see if it looks like it's a\n// # caption. Bonus if so.\n// sibling = img.getnext()\n// if sibling is not None:\n// if sibling.tag == 'figcaption':\n// img_score += 25\n//\n// sib_sig = ' '.join([sibling.get('id', ''),\n// sibling.get('class', '')]).lower()\n// if 'caption' in sib_sig:\n// img_score += 15\n//\n// # Pull out width/height if they were set.\n// img_width = None\n// img_height = None\n// if 'width' in img.attrib:\n// try:\n// img_width = float(img.get('width'))\n// except ValueError:\n// pass\n// if 'height' in img.attrib:\n// try:\n// img_height = float(img.get('height'))\n// except ValueError:\n// pass\n//\n// # Penalty for skinny images\n// if img_width and img_width <= 50:\n// logger.debug('Skinny image found. Subtracting 50.')\n// img_score -= 50\n//\n// # Penalty for short images\n// if img_height and img_height <= 50:\n// # Wide, short images are more common than narrow, tall ones\n// logger.debug('Short image found. Subtracting 25.')\n// img_score -= 25\n//\n// if img_width and img_height and not 'sprite' in img_path:\n// area = img_width * img_height\n//\n// if area < 5000: # Smaller than 50x100\n// logger.debug('Image with small area found. Subtracting 100.')\n// img_score -= 100\n// else:\n// img_score += round(area/1000.0)\n//\n// # If the image is higher on the page than other images,\n// # it gets a bonus. Penalty if lower.\n// logger.debug('Adding page placement bonus of %d.', len(imgs)/2 - i)\n// img_score += len(imgs)/2 - i\n//\n// # Use the raw src here because we munged img_path for case\n// # insensitivity\n// logger.debug('Final score is %d.', img_score)\n// img_scores[img.attrib['src']] += img_score\n//\n// top_score = 0\n// top_url = None\n// for (url, score) in img_scores.items():\n// if score > top_score:\n// top_url = url\n// top_score = score\n//\n// if top_score > 0:\n// logger.debug('Using top score image from content. Score was %d', top_score)\n// return top_url\n//\n//\n// # If nothing else worked, check to see if there are any really\n// # probable nodes in the doc, like <link rel=\"image_src\" />.\n// logger.debug('Trying to find lead image in probable nodes')\n// for selector in constants.LEAD_IMAGE_URL_SELECTORS:\n// nodes = self.resource.extract_by_selector(selector)\n// for node in nodes:\n// clean_value = None\n// if node.attrib.get('src'):\n// clean_value = self.clean(node.attrib['src'])\n//\n// if not clean_value and node.attrib.get('href'):\n// clean_value = self.clean(node.attrib['href'])\n//\n// if not clean_value and node.attrib.get('value'):\n// clean_value = self.clean(node.attrib['value'])\n//\n// if clean_value:\n// logger.debug('Found lead image in probable nodes.')\n// logger.debug('Node was: %s', node)\n// return clean_value\n//\n// return None\n","import difflib from 'difflib';\n\nexport default function scoreSimilarity(score, articleUrl, href) {\n // Do this last and only if we have a real candidate, because it's\n // potentially expensive computationally. Compare the link to this\n // URL using difflib to get the % similarity of these URLs. On a\n // sliding scale, subtract points from this link based on\n // similarity.\n if (score > 0) {\n const similarity = new difflib.SequenceMatcher(null, articleUrl, href).ratio();\n // Subtract .1 from diff_percent when calculating modifier,\n // which means that if it's less than 10% different, we give a\n // bonus instead. Ex:\n // 3% different = +17.5 points\n // 10% different = 0 points\n // 20% different = -25 points\n const diffPercent = 1.0 - similarity;\n const diffModifier = -(250 * (diffPercent - 0.2));\n return score + diffModifier;\n }\n\n return 0;\n}\n","import { IS_DIGIT_RE } from 'utils/text/constants';\n\nexport default function scoreLinkText(linkText, pageNum) {\n // If the link text can be parsed as a number, give it a minor\n // bonus, with a slight bias towards lower numbered pages. This is\n // so that pages that might not have 'next' in their text can still\n // get scored, and sorted properly by score.\n let score = 0;\n\n if (IS_DIGIT_RE.test(linkText.trim())) {\n const linkTextAsNum = parseInt(linkText, 10);\n // If it's the first page, we already got it on the first call.\n // Give it a negative score. Otherwise, up to page 10, give a\n // small bonus.\n if (linkTextAsNum < 2) {\n score = -30;\n } else {\n score = Math.max(0, 10 - linkTextAsNum);\n }\n\n // If it appears that the current page number is greater than\n // this links page number, it's a very bad sign. Give it a big\n // penalty.\n if (pageNum && pageNum >= linkTextAsNum) {\n score -= 50;\n }\n }\n\n return score;\n}\n","export default function scorePageInLink(pageNum, isWp) {\n // page in the link = bonus. Intentionally ignore wordpress because\n // their ?p=123 link style gets caught by this even though it means\n // separate documents entirely.\n if (pageNum && !isWp) {\n return 50;\n }\n\n return 0;\n}\n","export const DIGIT_RE = /\\d/;\n\n// A list of words that, if found in link text or URLs, likely mean that\n// this link is not a next page link.\nexport const EXTRANEOUS_LINK_HINTS = [\n 'print',\n 'archive',\n 'comment',\n 'discuss',\n 'e-mail',\n 'email',\n 'share',\n 'reply',\n 'all',\n 'login',\n 'sign',\n 'single',\n 'adx',\n 'entry-unrelated',\n];\nexport const EXTRANEOUS_LINK_HINTS_RE = new RegExp(EXTRANEOUS_LINK_HINTS.join('|'), 'i');\n\n// Match any link text/classname/id that looks like it could mean the next\n// page. Things like: next, continue, >, >>, » but not >|, »| as those can\n// mean last page.\nexport const NEXT_LINK_TEXT_RE = new RegExp('(next|weiter|continue|>([^|]|$)|»([^|]|$))', 'i');\n\n// Match any link text/classname/id that looks like it is an end link: things\n// like \"first\", \"last\", \"end\", etc.\nexport const CAP_LINK_TEXT_RE = new RegExp('(first|last|end)', 'i');\n\n// Match any link text/classname/id that looks like it means the previous\n// page.\nexport const PREV_LINK_TEXT_RE = new RegExp('(prev|earl|old|new|<|«)', 'i');\n\n// Match any phrase that looks like it could be page, or paging, or pagination\nexport const PAGE_RE = new RegExp('pag(e|ing|inat)', 'i');\n\n","import { EXTRANEOUS_LINK_HINTS_RE } from '../constants';\n\nexport default function scoreExtraneousLinks(href) {\n // If the URL itself contains extraneous values, give a penalty.\n if (EXTRANEOUS_LINK_HINTS_RE.test(href)) {\n return -25;\n }\n\n return 0;\n}\n","import { range } from 'utils';\nimport {\n NEGATIVE_SCORE_RE,\n POSITIVE_SCORE_RE,\n PAGE_RE,\n} from 'utils/dom/constants';\nimport { EXTRANEOUS_LINK_HINTS_RE } from '../constants';\n\nfunction makeSig($link) {\n return `${$link.attr('class') || ''} ${$link.attr('id') || ''}`;\n}\n\nexport default function scoreByParents($link) {\n // If a parent node contains paging-like classname or id, give a\n // bonus. Additionally, if a parent_node contains bad content\n // (like 'sponsor'), give a penalty.\n let $parent = $link.parent();\n let positiveMatch = false;\n let negativeMatch = false;\n let score = 0;\n\n Array.from(range(0, 4)).forEach(() => {\n if ($parent.length === 0) {\n return;\n }\n\n const parentData = makeSig($parent, ' ');\n\n // If we have 'page' or 'paging' in our data, that's a good\n // sign. Add a bonus.\n if (!positiveMatch && PAGE_RE.test(parentData)) {\n positiveMatch = true;\n score += 25;\n }\n\n // If we have 'comment' or something in our data, and\n // we don't have something like 'content' as well, that's\n // a bad sign. Give a penalty.\n if (!negativeMatch && NEGATIVE_SCORE_RE.test(parentData)\n && EXTRANEOUS_LINK_HINTS_RE.test(parentData)) {\n if (!POSITIVE_SCORE_RE.test(parentData)) {\n negativeMatch = true;\n score -= 25;\n }\n }\n\n $parent = $parent.parent();\n });\n\n return score;\n}\n\n","import { PREV_LINK_TEXT_RE } from '../constants';\n\nexport default function scorePrevLink(linkData) {\n // If the link has something like \"previous\", its definitely\n // an old link, skip it.\n if (PREV_LINK_TEXT_RE.test(linkData)) {\n return -200;\n }\n\n return 0;\n}\n","import URL from 'url';\n\nimport {\n DIGIT_RE,\n EXTRANEOUS_LINK_HINTS_RE,\n} from '../constants';\n\nexport default function shouldScore(\n href,\n articleUrl,\n baseUrl,\n parsedUrl,\n linkText,\n previousUrls\n) {\n // skip if we've already fetched this url\n if (previousUrls.find(url => href === url) !== undefined) {\n return false;\n }\n\n // If we've already parsed this URL, or the URL matches the base\n // URL, or is empty, skip it.\n if (!href || href === articleUrl || href === baseUrl) {\n return false;\n }\n\n const { hostname } = parsedUrl;\n const { hostname: linkHost } = URL.parse(href);\n\n // Domain mismatch.\n if (linkHost !== hostname) {\n return false;\n }\n\n // If href doesn't contain a digit after removing the base URL,\n // it's certainly not the next page.\n const fragment = href.replace(baseUrl, '');\n if (!DIGIT_RE.test(fragment)) {\n return false;\n }\n\n // This link has extraneous content (like \"comment\") in its link\n // text, so we skip it.\n if (EXTRANEOUS_LINK_HINTS_RE.test(linkText)) {\n return false;\n }\n\n // Next page link text is never long, skip if it is too long.\n if (linkText.length > 25) {\n return false;\n }\n\n return true;\n}\n\n","export default function scoreBaseUrl(href, baseRegex) {\n // If the baseUrl isn't part of this URL, penalize this\n // link. It could still be the link, but the odds are lower.\n // Example:\n // http://www.actionscript.org/resources/articles/745/1/JavaScript-and-VBScript-Injection-in-ActionScript-3/Page1.html\n if (!baseRegex.test(href)) {\n return -25;\n }\n\n return 0;\n}\n","import { NEXT_LINK_TEXT_RE } from '../constants';\n\nexport default function scoreNextLinkText(linkData) {\n // Things like \"next\", \">>\", etc.\n if (NEXT_LINK_TEXT_RE.test(linkData)) {\n return 50;\n }\n\n return 0;\n}\n","import {\n NEXT_LINK_TEXT_RE,\n CAP_LINK_TEXT_RE,\n} from '../constants';\n\nexport default function scoreCapLinks(linkData) {\n // Cap links are links like \"last\", etc.\n if (CAP_LINK_TEXT_RE.test(linkData)) {\n // If we found a link like \"last\", but we've already seen that\n // this link is also \"next\", it's fine. If it's not been\n // previously marked as \"next\", then it's probably bad.\n // Penalize.\n if (NEXT_LINK_TEXT_RE.test(linkData)) {\n return -65;\n }\n }\n\n return 0;\n}\n","import URL from 'url';\n\nimport { isWordpress } from 'utils/dom';\nimport {\n removeAnchor,\n pageNumFromUrl,\n} from 'utils/text';\n\nimport {\n scoreSimilarity,\n scoreLinkText,\n scorePageInLink,\n scoreExtraneousLinks,\n scoreByParents,\n scorePrevLink,\n shouldScore,\n scoreBaseUrl,\n scoreCapLinks,\n scoreNextLinkText,\n} from './utils';\n\nexport function makeBaseRegex(baseUrl) {\n return new RegExp(`^${baseUrl}`, 'i');\n}\n\nfunction makeSig($link, linkText) {\n return `${linkText || $link.text()} ${$link.attr('class') || ''} ${$link.attr('id') || ''}`;\n}\n\nexport default function scoreLinks({\n links,\n articleUrl,\n baseUrl,\n parsedUrl,\n $,\n previousUrls = [],\n}) {\n parsedUrl = parsedUrl || URL.parse(articleUrl);\n const baseRegex = makeBaseRegex(baseUrl);\n const isWp = isWordpress($);\n\n // Loop through all links, looking for hints that they may be next-page\n // links. Things like having \"page\" in their textContent, className or\n // id, or being a child of a node with a page-y className or id.\n //\n // After we do that, assign each page a score, and pick the one that\n // looks most like the next page link, as long as its score is strong\n // enough to have decent confidence.\n const scoredPages = links.reduce((possiblePages, link) => {\n // Remove any anchor data since we don't do a good job\n // standardizing URLs (it's hard), we're going to do\n // some checking with and without a trailing slash\n const href = removeAnchor(link.attribs.href);\n const $link = $(link);\n const linkText = $link.text();\n\n if (!shouldScore(href, articleUrl, baseUrl, parsedUrl, linkText, previousUrls)) {\n return possiblePages;\n }\n\n // ## PASSED THE FIRST-PASS TESTS. Start scoring. ##\n if (!possiblePages[href]) {\n possiblePages[href] = {\n score: 0,\n linkText,\n href,\n };\n } else {\n possiblePages[href].linkText = `${possiblePages[href].linkText}|${linkText}`;\n }\n\n const possiblePage = possiblePages[href];\n const linkData = makeSig($link, linkText);\n const pageNum = pageNumFromUrl(href);\n\n let score = scoreBaseUrl(href, baseRegex);\n score += scoreNextLinkText(linkData);\n score += scoreCapLinks(linkData);\n score += scorePrevLink(linkData);\n score += scoreByParents($link);\n score += scoreExtraneousLinks(href);\n score += scorePageInLink(pageNum, isWp);\n score += scoreLinkText(linkText, pageNum);\n score += scoreSimilarity(score, articleUrl, href);\n\n possiblePage.score = score;\n\n return possiblePages;\n }, {});\n\n return Reflect.ownKeys(scoredPages).length === 0 ? null : scoredPages;\n}\n","import URL from 'url';\n\nimport {\n articleBaseUrl,\n removeAnchor,\n} from 'utils/text';\nimport scoreLinks from './scoring/score-links';\n\n// Looks for and returns next page url\n// for multi-page articles\nconst GenericNextPageUrlExtractor = {\n extract({ $, url, parsedUrl, previousUrls = [] }) {\n parsedUrl = parsedUrl || URL.parse(url);\n\n const articleUrl = removeAnchor(url);\n const baseUrl = articleBaseUrl(url, parsedUrl);\n\n const links = $('a[href]').toArray();\n\n const scoredLinks = scoreLinks({\n links,\n articleUrl,\n baseUrl,\n parsedUrl,\n $,\n previousUrls,\n });\n\n // If no links were scored, return null\n if (!scoredLinks) return null;\n\n // now that we've scored all possible pages,\n // find the biggest one.\n const topPage = Reflect.ownKeys(scoredLinks).reduce((acc, link) => {\n const scoredLink = scoredLinks[link];\n return scoredLink.score > acc.score ? scoredLink : acc;\n }, { score: -100 });\n\n // If the score is less than 50, we're not confident enough to use it,\n // so we fail.\n if (topPage.score >= 50) {\n return topPage.href;\n }\n\n return null;\n },\n};\n\n\nexport default GenericNextPageUrlExtractor;\n","export const CANONICAL_META_SELECTORS = [\n 'og:url',\n];\n","import URL from 'url';\nimport { extractFromMeta } from 'utils/dom';\n\nimport { CANONICAL_META_SELECTORS } from './constants';\n\nfunction parseDomain(url) {\n const parsedUrl = URL.parse(url);\n const { hostname } = parsedUrl;\n return hostname;\n}\n\nfunction result(url) {\n return {\n url,\n domain: parseDomain(url),\n };\n}\n\nconst GenericUrlExtractor = {\n extract({ $, url, metaCache }) {\n const $canonical = $('link[rel=canonical]');\n if ($canonical.length !== 0) {\n const href = $canonical.attr('href');\n if (href) {\n return result(href);\n }\n }\n\n const metaUrl = extractFromMeta($, CANONICAL_META_SELECTORS, metaCache);\n if (metaUrl) {\n return result(metaUrl);\n }\n\n return result(url);\n },\n\n};\n\nexport default GenericUrlExtractor;\n","export const EXCERPT_META_SELECTORS = [\n 'og:description',\n 'twitter:description',\n];\n","import ellipsize from 'ellipsize';\n\nimport {\n extractFromMeta,\n stripTags,\n} from 'utils/dom';\n\nimport { EXCERPT_META_SELECTORS } from './constants';\n\nexport function clean(content, $, maxLength = 200) {\n content = content.replace(/[\\s\\n]+/g, ' ').trim();\n return ellipsize(content, maxLength, { ellipse: '…' });\n}\n\nconst GenericExcerptExtractor = {\n extract({ $, content, metaCache }) {\n const excerpt = extractFromMeta($, EXCERPT_META_SELECTORS, metaCache);\n if (excerpt) {\n return clean(stripTags(excerpt, $));\n }\n // Fall back to excerpting from the extracted content\n const maxLength = 200;\n const shortContent = content.slice(0, maxLength * 5);\n return clean($(shortContent).text(), $, maxLength);\n },\n};\n\nexport default GenericExcerptExtractor;\n","import cheerio from 'cheerio';\n\nimport { normalizeSpaces } from 'utils/text';\n\nconst GenericWordCountExtractor = {\n extract({ content }) {\n const $ = cheerio.load(content);\n\n const text = normalizeSpaces($('div').first().text());\n return text.split(/\\s/).length;\n },\n};\n\nexport default GenericWordCountExtractor;\n","import cheerio from 'cheerio';\nimport stringDirection from 'string-direction';\n\nimport GenericContentExtractor from './content/extractor';\nimport GenericTitleExtractor from './title/extractor';\nimport GenericAuthorExtractor from './author/extractor';\nimport GenericDatePublishedExtractor from './date-published/extractor';\nimport GenericDekExtractor from './dek/extractor';\nimport GenericLeadImageUrlExtractor from './lead-image-url/extractor';\nimport GenericNextPageUrlExtractor from './next-page-url/extractor';\nimport GenericUrlExtractor from './url/extractor';\nimport GenericExcerptExtractor from './excerpt/extractor';\nimport GenericWordCountExtractor from './word-count/extractor';\n\nconst GenericExtractor = {\n // This extractor is the default for all domains\n domain: '*',\n title: GenericTitleExtractor.extract,\n date_published: GenericDatePublishedExtractor.extract,\n author: GenericAuthorExtractor.extract,\n content: GenericContentExtractor.extract.bind(GenericContentExtractor),\n lead_image_url: GenericLeadImageUrlExtractor.extract,\n dek: GenericDekExtractor.extract,\n next_page_url: GenericNextPageUrlExtractor.extract,\n url_and_domain: GenericUrlExtractor.extract,\n excerpt: GenericExcerptExtractor.extract,\n word_count: GenericWordCountExtractor.extract,\n direction: ({ title }) => stringDirection.getDirection(title),\n\n extract(options) {\n const { html } = options;\n\n if (html) {\n const $ = cheerio.load(html);\n options.$ = $;\n }\n\n const title = this.title(options);\n const date_published = this.date_published(options);\n const author = this.author(options);\n const content = this.content({ ...options, title });\n const lead_image_url = this.lead_image_url({ ...options, content });\n const dek = this.dek({ ...options, content });\n const next_page_url = this.next_page_url(options);\n const excerpt = this.excerpt({ ...options, content });\n const word_count = this.word_count({ ...options, content });\n const direction = this.direction({ title });\n const { url, domain } = this.url_and_domain(options);\n\n return {\n title,\n author,\n date_published: date_published || null,\n dek,\n lead_image_url,\n content,\n next_page_url,\n url,\n domain,\n excerpt,\n word_count,\n direction,\n };\n },\n};\n\nexport default GenericExtractor;\n","import URL from 'url';\n\nimport Extractors from './all';\nimport GenericExtractor from './generic';\n\nexport default function getExtractor(url, parsedUrl) {\n parsedUrl = parsedUrl || URL.parse(url);\n const { hostname } = parsedUrl;\n const baseDomain = hostname.split('.').slice(-2).join('.');\n\n return Extractors[hostname] || Extractors[baseDomain] || GenericExtractor;\n}\n","import Cleaners from 'cleaners';\nimport { convertNodeTo } from 'utils/dom';\nimport GenericExtractor from './generic';\n\n// Remove elements by an array of selectors\nexport function cleanBySelectors($content, $, { clean }) {\n if (!clean) return $content;\n\n $(clean.join(','), $content).remove();\n\n return $content;\n}\n\n// Transform matching elements\nexport function transformElements($content, $, { transforms }) {\n if (!transforms) return $content;\n\n Reflect.ownKeys(transforms).forEach((key) => {\n const $matches = $(key, $content);\n const value = transforms[key];\n\n // If value is a string, convert directly\n if (typeof value === 'string') {\n $matches.each((index, node) => {\n convertNodeTo($(node), $, transforms[key]);\n });\n } else if (typeof value === 'function') {\n // If value is function, apply function to node\n $matches.each((index, node) => {\n const result = value($(node), $);\n // If function returns a string, convert node to that value\n if (typeof result === 'string') {\n convertNodeTo($(node), $, result);\n }\n });\n }\n });\n\n return $content;\n}\n\nfunction findMatchingSelector($, selectors) {\n return selectors.find((selector) => {\n if (Array.isArray(selector)) {\n const [s, attr] = selector;\n return $(s).length === 1 && $(s).attr(attr) && $(s).attr(attr).trim() !== '';\n }\n\n return $(selector).length === 1 && $(selector).text().trim() !== '';\n });\n}\n\nexport function select(opts) {\n const { $, type, extractionOpts, extractHtml = false } = opts;\n // Skip if there's not extraction for this type\n if (!extractionOpts) return null;\n\n // If a string is hardcoded for a type (e.g., Wikipedia\n // contributors), return the string\n if (typeof extractionOpts === 'string') return extractionOpts;\n\n const { selectors, defaultCleaner = true } = extractionOpts;\n\n const matchingSelector = findMatchingSelector($, selectors);\n\n if (!matchingSelector) return null;\n\n // Declaring result; will contain either\n // text or html, which will be cleaned\n // by the appropriate cleaner type\n\n // If the selector type requests html as its return type\n // transform and clean the element with provided selectors\n if (extractHtml) {\n let $content = $(matchingSelector);\n\n // Wrap in div so transformation can take place on root element\n $content.wrap($('<div></div>'));\n $content = $content.parent();\n\n $content = transformElements($content, $, extractionOpts);\n $content = cleanBySelectors($content, $, extractionOpts);\n\n $content = Cleaners[type]($content, { ...opts, defaultCleaner });\n\n return $.html($content);\n }\n\n let result;\n\n // if selector is an array (e.g., ['img', 'src']),\n // extract the attr\n if (Array.isArray(matchingSelector)) {\n const [selector, attr] = matchingSelector;\n result = $(selector).attr(attr).trim();\n } else {\n result = $(matchingSelector).text().trim();\n }\n\n // Allow custom extractor to skip default cleaner\n // for this type; defaults to true\n if (defaultCleaner) {\n return Cleaners[type](result, opts);\n }\n\n return result;\n}\n\nfunction extractResult(opts) {\n const { type, extractor, fallback = true } = opts;\n\n const result = select({ ...opts, extractionOpts: extractor[type] });\n\n // If custom parser succeeds, return the result\n if (result) {\n return result;\n }\n\n // If nothing matches the selector, and fallback is enabled,\n // run the Generic extraction\n if (fallback) return GenericExtractor[type](opts);\n\n return null;\n}\n\nconst RootExtractor = {\n extract(extractor = GenericExtractor, opts) {\n const { contentOnly, extractedTitle } = opts;\n // This is the generic extractor. Run its extract method\n if (extractor.domain === '*') return extractor.extract(opts);\n\n opts = {\n ...opts,\n extractor,\n };\n\n if (contentOnly) {\n const content = extractResult({\n ...opts, type: 'content', extractHtml: true, title: extractedTitle,\n });\n return {\n content,\n };\n }\n const title = extractResult({ ...opts, type: 'title' });\n const date_published = extractResult({ ...opts, type: 'date_published' });\n const author = extractResult({ ...opts, type: 'author' });\n const next_page_url = extractResult({ ...opts, type: 'next_page_url' });\n const content = extractResult({\n ...opts, type: 'content', extractHtml: true, title,\n });\n const lead_image_url = extractResult({ ...opts, type: 'lead_image_url', content });\n const excerpt = extractResult({ ...opts, type: 'excerpt', content });\n const dek = extractResult({ ...opts, type: 'dek', content, excerpt });\n const word_count = extractResult({ ...opts, type: 'word_count', content });\n const direction = extractResult({ ...opts, type: 'direction', title });\n const { url, domain } =\n extractResult({ ...opts, type: 'url_and_domain' }) || { url: null, domain: null };\n\n return {\n title,\n content,\n author,\n date_published,\n lead_image_url,\n dek,\n next_page_url,\n url,\n domain,\n excerpt,\n word_count,\n direction,\n };\n },\n};\n\nexport default RootExtractor;\n","import { removeAnchor } from 'utils/text';\nimport RootExtractor from 'extractors/root-extractor';\nimport GenericExtractor from 'extractors/generic';\nimport Resource from 'resource';\n\nexport default async function collectAllPages(\n {\n next_page_url,\n html,\n $,\n metaCache,\n result,\n Extractor,\n title,\n url,\n }\n) {\n // At this point, we've fetched just the first page\n let pages = 1;\n const previousUrls = [removeAnchor(url)];\n\n // If we've gone over 26 pages, something has\n // likely gone wrong.\n while (next_page_url && pages < 26) {\n pages += 1;\n $ = await Resource.create(next_page_url);\n html = $.html();\n\n const extractorOpts = {\n url: next_page_url,\n html,\n $,\n metaCache,\n contentOnly: true,\n extractedTitle: title,\n previousUrls,\n };\n\n const nextPageResult = RootExtractor.extract(Extractor, extractorOpts);\n\n previousUrls.push(next_page_url);\n result = {\n ...result,\n content: `\n ${result.content}\n <hr>\n <h4>Page ${pages}</h4>\n ${nextPageResult.content}\n `,\n };\n\n next_page_url = nextPageResult.next_page_url;\n }\n\n const word_count = GenericExtractor.word_count({ content: `<div>${result.content}</div>` });\n return {\n ...result,\n total_pages: pages,\n pages_rendered: pages,\n word_count,\n };\n}\n","import URL from 'url';\n\nimport Resource from 'resource';\nimport {\n validateUrl,\n Errors,\n} from 'utils';\nimport getExtractor from 'extractors/get-extractor';\nimport RootExtractor from 'extractors/root-extractor';\nimport collectAllPages from 'extractors/collect-all-pages';\n\nconst Mercury = {\n async parse(url, html, opts = {}) {\n const {\n fetchAllPages = true,\n fallback = true,\n } = opts;\n\n const parsedUrl = URL.parse(url);\n\n if (!validateUrl(parsedUrl)) {\n return Errors.badUrl;\n }\n\n const Extractor = getExtractor(url, parsedUrl);\n // console.log(`Using extractor for ${Extractor.domain}`);\n\n const $ = await Resource.create(url, html, parsedUrl);\n\n // If we found an error creating the resource, return that error\n if ($.error) {\n return $;\n }\n\n html = $.html();\n\n // Cached value of every meta name in our document.\n // Used when extracting title/author/date_published/dek\n const metaCache = $('meta').map((_, node) => $(node).attr('name')).toArray();\n\n let result = RootExtractor.extract(Extractor, { url, html, $, metaCache, parsedUrl, fallback });\n const { title, next_page_url } = result;\n\n // Fetch more pages if next_page_url found\n if (fetchAllPages && next_page_url) {\n result = await collectAllPages(\n {\n Extractor,\n next_page_url,\n html,\n $,\n metaCache,\n result,\n title,\n url,\n }\n );\n } else {\n result = {\n ...result,\n total_pages: 1,\n rendered_pages: 1,\n };\n }\n\n return result;\n },\n\n // A convenience method for getting a resource\n // to work with, e.g., for custom extractor generator\n async fetchResource(url) {\n return await Resource.create(url);\n },\n\n};\n\nexport default Mercury;\n"],"names":["range","start","end","validateUrl","hostname","Errors","REQUEST_HEADERS","FETCH_TIMEOUT","BAD_CONTENT_TYPES","BAD_CONTENT_TYPES_RE","RegExp","join","MAX_CONTENT_LENGTH","get","options","resolve","reject","err","response","body","validateResponse","parseNon2xx","statusMessage","statusCode","Error","error","headers","contentType","contentLength","test","url","parsedUrl","URL","parse","encodeURI","badUrl","fetchResource","convertMetaProp","$","from","to","each","_","node","$node","value","attr","removeAttr","normalizeMetaTags","IS_LINK","IS_IMAGE","TAGS_TO_REMOVE","convertLazyLoadedImages","img","attribs","forEach","isComment","index","type","cleanComments","root","find","contents","filter","remove","clean","Resource","preparedResponse","validResponse","result","generateDoc","content","includes","cheerio","load","normalizeWhitespace","children","length","NYMagExtractor","$children","tagName","BloggerExtractor","WikipediaExtractor","$parent","parents","prepend","TwitterExtractor","tweets","$tweetContainer","append","replaceWith","NYTimesExtractor","src","width","replace","TheAtlanticExtractor","NewYorkerExtractor","WiredExtractor","MSNExtractor","YahooExtractor","BuzzfeedExtractor","WikiaExtractor","LittleThingsExtractor","PoliticoExtractor","DeadspinExtractor","youtubeId","split","BroadwayWorldExtractor","ApartmentTherapyExtractor","data","JSON","sources","$img","Extractors","SPACER_RE","KEEP_CLASS","KEEP_SELECTORS","STRIP_OUTPUT_TAGS","REMOVE_ATTRS","REMOVE_ATTR_SELECTORS","map","selector","REMOVE_ATTR_LIST","WHITELIST_ATTRS","WHITELIST_ATTRS_RE","REMOVE_EMPTY_TAGS","REMOVE_EMPTY_SELECTORS","tag","CLEAN_CONDITIONALLY_TAGS","HEADER_TAGS","HEADER_TAG_LIST","UNLIKELY_CANDIDATES_BLACKLIST","UNLIKELY_CANDIDATES_WHITELIST","DIV_TO_P_BLOCK_TAGS","NON_TOP_CANDIDATE_TAGS","NON_TOP_CANDIDATE_TAGS_RE","PHOTO_HINTS","PHOTO_HINTS_RE","POSITIVE_SCORE_HINTS","POSITIVE_SCORE_RE","NEGATIVE_SCORE_HINTS","NEGATIVE_SCORE_RE","IS_WP_SELECTOR","EXTRANEOUS_LINK_HINTS","EXTRANEOUS_LINK_HINTS_RE","PAGE_RE","BLOCK_LEVEL_TAGS","BLOCK_LEVEL_TAGS_RE","candidatesBlacklist","CANDIDATES_BLACKLIST","candidatesWhitelist","CANDIDATES_WHITELIST","stripUnlikelyCandidates","not","classes","id","classAndId","brsToPs","collapsing","element","nextElement","next","paragraphize","br","sibling","nextSibling","p","appendTo","convertDivs","div","$div","convertable","convertSpans","span","$span","convertToParagraphs","convertNodeTo","attribString","key","cleanForHeight","height","parseInt","removeSpacers","cleanImages","$article","markToKeep","article","tags","addClass","stripJunkTags","removeClass","cleanHOnes","$hOnes","removeAllButWhitelist","reduce","acc","cleanAttributes","log","parent","removeEmpty","$p","text","trim","HNEWS_CONTENT_SELECTORS","READABILITY_ASSET","PARAGRAPH_SCORE_TAGS","CHILD_CONTENT_TAGS","BAD_TAGS","getWeight","score","getScore","parseFloat","scoreCommas","match","idkRe","scoreLength","textLength","chunks","lengthBonus","Math","min","max","scoreParagraph","slice","setScore","addScore","amount","getOrInitScore","e","addToParent","weightNodes","scoreNode","addScoreTo","scorePs","rawScore","scoreContent","parentSelector","childSelector","NORMALIZE_RE","normalizeSpaces","extractFromUrl","regexList","matchRe","re","exec","PAGE_IN_HREF_RE","HAS_ALPHA_RE","IS_ALPHA_RE","IS_DIGIT_RE","pageNumFromUrl","matches","pageNum","removeAnchor","isGoodSegment","segment","firstSegmentHasLetters","goodSegment","toLowerCase","articleBaseUrl","parsed","protocol","host","path","cleanedSegments","reverse","rawSegment","possibleSegment","fileExt","push","SENTENCE_END_RE","hasSentenceEnd","excerptContent","words","mergeSiblings","$candidate","topScore","siblingScoreThreshold","wrappingDiv","$sibling","siblingScore","contentBonus","density","linkDensity","newScore","siblingContent","siblingContentLength","findTopCandidate","first","removeUnlessContent","weight","hasClass","pCount","inputCount","imgCount","nodeIsList","previousNode","prev","scriptCount","cleanTags","cleanHeaders","title","header","$header","prevAll","rewriteTopLevel","absolutize","rootUrl","$content","absoluteUrl","makeLinksAbsolute","totalTextLength","linkText","linkLength","extractFromMeta","metaNames","cachedNames","foundNames","indexOf","name","nodes","values","toArray","metaValue","stripTags","isGoodNode","maxChildren","withinComment","extractFromSelectors","selectors","textOnly","html","cleanText","commentParent","class","undefined","nodeIsSufficient","isWordpress","CLEAN_AUTHOR_RE","TEXT_LINK_RE","MS_DATE_STRING","SEC_DATE_STRING","CLEAN_DATE_STRING_RE","TIME_MERIDIAN_SPACE_RE","TIME_MERIDIAN_DOTS_RE","months","allMonths","timestamp1","timestamp2","SPLIT_DATE_STRING","TITLE_SPLITTERS_RE","DOMAIN_ENDINGS_RE","cleanAuthor","author","leadImageUrl","validUrl","isWebUri","cleanDek","dek","excerpt","dekText","cleanDateString","dateString","cleanDatePublished","date","moment","Date","isValid","toISOString","extractCleanNode","cleanConditionally","defaultCleaner","cleanTitle","resolveSplitTitle","h1","extractBreadcrumbTitle","splitTitle","termCounts","titleText","maxTerm","termCount","splitEnds","longestEnd","cleanDomainFromTitle","nakedDomain","startSlug","startSlugRatio","wuzzy","levenshtein","endSlug","endSlugRatio","newTitle","Cleaners","cleanImage","cleanContent","extractBestNode","opts","$topCandidate","GenericContentExtractor","defaultOpts","getContentNode","cleanAndReturnNode","k","STRONG_TITLE_META_TAGS","WEAK_TITLE_META_TAGS","STRONG_TITLE_SELECTORS","WEAK_TITLE_SELECTORS","GenericTitleExtractor","metaCache","AUTHOR_META_TAGS","AUTHOR_MAX_LENGTH","AUTHOR_SELECTORS","bylineRe","BYLINE_SELECTORS_RE","GenericAuthorExtractor","regex","DATE_PUBLISHED_META_TAGS","DATE_PUBLISHED_SELECTORS","abbrevMonthsStr","DATE_PUBLISHED_URL_RES","GenericDatePublishedExtractor","datePublished","GenericDekExtractor","LEAD_IMAGE_URL_META_TAGS","LEAD_IMAGE_URL_SELECTORS","POSITIVE_LEAD_IMAGE_URL_HINTS","POSITIVE_LEAD_IMAGE_URL_HINTS_RE","NEGATIVE_LEAD_IMAGE_URL_HINTS","NEGATIVE_LEAD_IMAGE_URL_HINTS_RE","GIF_RE","JPG_RE","getSig","scoreImageUrl","scoreAttr","scoreByParents","$figParent","$gParent","scoreBySibling","scoreByDimensions","area","round","scoreByPosition","$imgs","GenericLeadImageUrlExtractor","cleanUrl","imageUrl","imgs","imgScores","topUrl","href","scoreSimilarity","articleUrl","similarity","difflib","SequenceMatcher","ratio","diffPercent","diffModifier","scoreLinkText","linkTextAsNum","scorePageInLink","isWp","DIGIT_RE","NEXT_LINK_TEXT_RE","CAP_LINK_TEXT_RE","PREV_LINK_TEXT_RE","scoreExtraneousLinks","makeSig","$link","positiveMatch","negativeMatch","parentData","scorePrevLink","linkData","shouldScore","baseUrl","previousUrls","linkHost","fragment","scoreBaseUrl","baseRegex","scoreNextLinkText","scoreCapLinks","makeBaseRegex","scoreLinks","links","scoredPages","possiblePages","link","possiblePage","GenericNextPageUrlExtractor","scoredLinks","topPage","scoredLink","CANONICAL_META_SELECTORS","parseDomain","GenericUrlExtractor","$canonical","metaUrl","EXCERPT_META_SELECTORS","maxLength","ellipsize","ellipse","GenericExcerptExtractor","shortContent","GenericWordCountExtractor","GenericExtractor","extract","bind","stringDirection","getDirection","date_published","lead_image_url","next_page_url","word_count","direction","url_and_domain","domain","getExtractor","baseDomain","cleanBySelectors","transformElements","transforms","$matches","findMatchingSelector","Array","isArray","s","select","extractionOpts","extractHtml","matchingSelector","wrap","extractResult","extractor","fallback","RootExtractor","contentOnly","extractedTitle","Extractor","pages","create","extractorOpts","nextPageResult","collectAllPages","Mercury","fetchAllPages"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;eAAyBA;;AAAzB,AAAe,SAAUA,KAAV;MAAgBC,KAAhB,uEAAwB,CAAxB;MAA2BC,GAA3B,uEAAiC,CAAjC;;;;;gBACND,SAASC,GADH;;;;;;iBAELD,SAAS,CAFJ;;;;;;;;;;;;;;ACAf;AACA,AAAe,SAASE,WAAT,OAAmC;MAAZC,QAAY,QAAZA,QAAY;;;SAEzC,CAAC,CAACA,QAAT;;;ACHF,IAAMC,SAAS;UACL;WACC,IADD;cAEI;;CAHd,CAOA;;ACPO,IAAMC,kBAAkB;gBACf;CADT;;;AAKP,AAAO,IAAMC,gBAAgB,KAAtB;;;AAGP,IAAMC,oBAAoB,CACxB,YADwB,EAExB,WAFwB,EAGxB,YAHwB,EAIxB,WAJwB,CAA1B;;AAOA,AAAO,IAAMC,uBAAuB,IAAIC,MAAJ,QAAgBF,kBAAkBG,IAAlB,CAAuB,GAAvB,CAAhB,SAAiD,GAAjD,CAA7B;;;;AAKP,AAAO,IAAMC,qBAAqB,OAA3B,CAEP,AAIA,AAKA;;ACpBA,SAASC,GAAT,CAAaC,OAAb,EAAsB;SACb,aAAY,UAACC,OAAD,EAAUC,MAAV,EAAqB;YAC9BF,OAAR,EAAiB,UAACG,GAAD,EAAMC,QAAN,EAAgBC,IAAhB,EAAyB;UACpCF,GAAJ,EAAS;eACAA,GAAP;OADF,MAEO;gBACG,EAAEE,UAAF,EAAQD,kBAAR,EAAR;;KAJJ;GADK,CAAP;;;;;;;;AAgBF,AAAO,SAASE,gBAAT,CAA0BF,QAA1B,EAAyD;MAArBG,WAAqB,uEAAP,KAAO;;;MAE1DH,SAASI,aAAT,KAA2B,IAA/B,EAAqC;QAC/B,CAACJ,SAASK,UAAd,EAA0B;YAClB,IAAIC,KAAJ,sDAC+CN,SAASO,KADxD,CAAN;KADF,MAIO,IAAI,CAACJ,WAAL,EAAkB;YACjB,IAAIG,KAAJ,kDAC2CN,SAASK,UADpD,wEAAN;;;;0BASAL,SAASQ,OAjBiD;MAe5CC,WAf4C,qBAe5D,cAf4D;MAgB1CC,aAhB0C,qBAgB5D,gBAhB4D;;;;MAoB1DnB,qBAAqBoB,IAArB,CAA0BF,WAA1B,CAAJ,EAA4C;UACpC,IAAIH,KAAJ,yCACkCG,WADlC,0BAAN;;;;MAMEC,gBAAgBhB,kBAApB,EAAwC;UAChC,IAAIY,KAAJ,yEACkEZ,kBADlE,OAAN;;;SAKK,IAAP;;;AAGF,AAMA;;;;;;AAMA;yDAAe,iBAA6BkB,GAA7B,EAAkCC,SAAlC;;;;;;;wBACDA,aAAaC,IAAIC,KAAJ,CAAUC,UAAUJ,GAAV,CAAV,CAAzB;;mBADa,GAGG;mBACTC,SADS;oCAEAzB,eAAd,CAFc;uBAGLC,aAHK;;;wBAMJ,IANI;;mBAQT,IARS;;oBAUR,IAVQ;;kCAYM;aAfT;;mBAkBoBM,IAAIC,OAAJ,CAlBpB;;;;oBAAA,SAkBLI,QAlBK;gBAAA,SAkBKC,IAlBL;;;6BAqBMD,QAAjB;6CACO,EAAEC,UAAF,EAAQD,kBAAR,EAtBI;;;;;6CAwBJb,OAAO8B,MAxBH;;;;;;;;GAAf;;WAA8BC,aAA9B;;;;SAA8BA,aAA9B;;;AC5EA,SAASC,eAAT,CAAyBC,CAAzB,EAA4BC,IAA5B,EAAkCC,EAAlC,EAAsC;cAC1BD,IAAV,QAAmBE,IAAnB,CAAwB,UAACC,CAAD,EAAIC,IAAJ,EAAa;QAC7BC,QAAQN,EAAEK,IAAF,CAAd;;QAEME,QAAQD,MAAME,IAAN,CAAWP,IAAX,CAAd;UACMO,IAAN,CAAWN,EAAX,EAAeK,KAAf;UACME,UAAN,CAAiBR,IAAjB;GALF;;SAQOD,CAAP;;;;;;;;;;AAUF,AAAe,SAASU,iBAAT,CAA2BV,CAA3B,EAA8B;MACvCD,gBAAgBC,CAAhB,EAAmB,SAAnB,EAA8B,OAA9B,CAAJ;MACID,gBAAgBC,CAAhB,EAAmB,UAAnB,EAA+B,MAA/B,CAAJ;SACOA,CAAP;;;ACtBK,IAAMW,UAAU,IAAIvC,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAhB;AACP,AAAO,IAAMwC,WAAW,IAAIxC,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB;;AAEP,AAAO,IAAMyC,iBAAiB,CAC5B,QAD4B,EAE5B,OAF4B,EAG5B,MAH4B,EAI5BxC,IAJ4B,CAIvB,GAJuB,CAAvB;;ACEP;;;;;AAKA,AAAe,SAASyC,uBAAT,CAAiCd,CAAjC,EAAoC;IAC/C,KAAF,EAASG,IAAT,CAAc,UAACC,CAAD,EAAIW,GAAJ,EAAY;qBACRA,IAAIC,OAApB,EAA6BC,OAA7B,CAAqC,UAACT,IAAD,EAAU;UACvCD,QAAQQ,IAAIC,OAAJ,CAAYR,IAAZ,CAAd;;UAEIA,SAAS,KAAT,IAAkBG,QAAQpB,IAAR,CAAagB,KAAb,CAAlB,IACAK,SAASrB,IAAT,CAAcgB,KAAd,CADJ,EAC0B;UACtBQ,GAAF,EAAOP,IAAP,CAAY,KAAZ,EAAmBD,KAAnB;;KALJ;GADF;;SAWOP,CAAP;;;ACpBF,SAASkB,SAAT,CAAmBC,KAAnB,EAA0Bd,IAA1B,EAAgC;SACvBA,KAAKe,IAAL,KAAc,SAArB;;;AAGF,SAASC,aAAT,CAAuBrB,CAAvB,EAA0B;IACtBsB,IAAF,GAASC,IAAT,CAAc,GAAd,EACSC,QADT,GAESC,MAFT,CAEgBP,SAFhB,EAGSQ,MAHT;;SAKO1B,CAAP;;;AAGF,AAAe,SAAS2B,KAAT,CAAe3B,CAAf,EAAkB;IAC7Ba,cAAF,EAAkBa,MAAlB;;MAEIL,cAAcrB,CAAd,CAAJ;SACOA,CAAP;;;ACVF,IAAM4B,WAAW;;;;;;;;QAAA,kBAQFpC,GARE,EAQGqC,gBARH,EAQqBpC,SARrB,EAQgC;;;;;;;;;oBAAA;;mBAGzCoC,gBAHyC;;;;;2BAAA,GAIrB;+BACL,IADK;4BAER,GAFQ;yBAGX;kCACS,WADT;oCAEW;;eATqB;;;uBAalC,EAAEhD,MAAMgD,gBAAR,EAA0BjD,UAAUkD,aAApC,EAAT;;;;;;qBAEehC,cAAcN,GAAd,EAAmBC,SAAnB,CAf4B;;;oBAAA;;;mBAkBzCsC,OAAO5C,KAlBkC;;;;;+CAmBpC4C,MAnBoC;;;+CAsBtC,MAAKC,WAAL,CAAiBD,MAAjB,CAtBsC;;;;;;;;;GARhC;aAAA,6BAiC0B;QAArBE,OAAqB,QAA3BpD,IAA2B;QAAZD,QAAY,QAAZA,QAAY;QACfS,WADe,GACCT,SAASQ,OADV,CAC/B,cAD+B;;;;;QAKnC,CAACC,YAAY6C,QAAZ,CAAqB,MAArB,CAAD,IACA,CAAC7C,YAAY6C,QAAZ,CAAqB,MAArB,CADL,EACmC;YAC3B,IAAIhD,KAAJ,CAAU,qCAAV,CAAN;;;QAGEc,IAAImC,QAAQC,IAAR,CAAaH,OAAb,EAAsB,EAAEI,qBAAqB,IAAvB,EAAtB,CAAR;;QAEIrC,EAAEsB,IAAF,GAASgB,QAAT,GAAoBC,MAApB,KAA+B,CAAnC,EAAsC;YAC9B,IAAIrD,KAAJ,CAAU,kCAAV,CAAN;;;QAGEwB,kBAAkBV,CAAlB,CAAJ;QACIc,wBAAwBd,CAAxB,CAAJ;QACI2B,MAAM3B,CAAN,CAAJ;;WAEOA,CAAP;;CArDJ,CAyDA;;AClEO,IAAMwC,iBAAiB;UACpB,WADoB;WAEnB;;eAEI,CACT,qBADS,EAET,cAFS,EAGT,iBAHS,CAFJ;;;WASA,CACL,KADK,EAEL,uBAFK,CATA;;;;;;;;gBAoBK;;UAEN,IAFM;;;gBAKA,kBAAClC,KAAD,EAAW;YACbmC,YAAYnC,MAAMgC,QAAN,EAAlB;YACIG,UAAUF,MAAV,KAAqB,CAArB,IAA0BE,UAAUlE,GAAV,CAAc,CAAd,EAAiBmE,OAAjB,KAA6B,KAA3D,EAAkE;iBACzD,QAAP;;;eAGK,IAAP;;;GAjCsB;;SAsCrB;eACM,CACT,uBADS,EAET,qBAFS,EAGT,IAHS;GAvCe;;UA8CpB;eACK,CACT,aADS,EAET,sBAFS;GA/Ce;;OAqDvB;eACQ,CACT,sBADS;GAtDe;;kBA2DZ;eACH,CACT,CAAC,kCAAD,EAAqC,UAArC,CADS,EAET,wBAFS;;CA5DR;;ACAA,IAAMC,mBAAmB;UACtB,cADsB;WAErB;;;;eAII,CACT,wBADS,CAJJ;;;WASA,EATA;;;gBAaK;gBACA;;GAhBgB;;UAoBtB;eACK,CACT,mBADS;GArBiB;;SA0BvB;eACM,CACT,gBADS;GA3BiB;;kBAgCd;eACH,CACT,kBADS;;CAjCR;;ACAA,IAAMC,qBAAqB;UACxB,eADwB;WAEvB;eACI,CACT,kBADS,CADJ;;oBAKS,KALT;;;gBAQK;sBACM,oBAACtC,KAAD,EAAW;YACnBuC,UAAUvC,MAAMwC,OAAN,CAAc,UAAd,CAAhB;;YAEID,QAAQP,QAAR,CAAiB,KAAjB,EAAwBC,MAAxB,KAAmC,CAAvC,EAA0C;kBAChCQ,OAAR,CAAgBzC,KAAhB;;OALM;0BAQU,YARV;kBASE;KAjBP;;;WAqBA,CACL,iBADK,EAEL,oCAFK,EAGL,MAHK,EAIL,SAJK;;GAvBuB;;UAgCxB,wBAhCwB;;SAkCzB;eACM,CACT,UADS;GAnCmB;;kBAwChB;eACH,CACT,sBADS;;;CAzCR;;ACAA,IAAM0C,mBAAmB;UACtB,aADsB;;WAGrB;gBACK;;;;;+BAKe,2BAAC1C,KAAD,EAAQN,CAAR,EAAc;YAC/BiD,SAAS3C,MAAMiB,IAAN,CAAW,QAAX,CAAf;YACM2B,kBAAkBlD,EAAE,iCAAF,CAAxB;wBACgBmD,MAAhB,CAAuBF,MAAvB;cACMG,WAAN,CAAkBF,eAAlB;OATQ;;;;SAcP;KAfE;;eAkBI,CACT,uBADS,CAlBJ;;oBAsBS,KAtBT;;WAwBA,CACL,qBADK,EAEL,QAFK,EAGL,sBAHK;GA3BqB;;UAkCtB;eACK,CACT,kCADS;GAnCiB;;kBAwCd;eACH,CACT,CAAC,4CAAD,EAA+C,cAA/C,CADS;;;CAzCR;;ACAA,IAAMG,mBAAmB;SACvB;eACM,CACT,aADS,EAET,aAFS;GAFiB;;UAQtB;eACK,CACT,CAAC,qBAAD,EAAwB,OAAxB,CADS,EAET,WAFS,EAGT,SAHS;GATiB;;WAgBrB;eACI,CACT,cADS,EAET,eAFS,CADJ;;oBAMS,KANT;;gBAQK;oBACI,kBAAC/C,KAAD,EAAW;YACnBgD,MAAMhD,MAAME,IAAN,CAAW,KAAX,CAAV;;;;;;;;;;YAUM+C,QAAQ,GAAd;;cAEMD,IAAIE,OAAJ,CAAY,UAAZ,EAAwBD,KAAxB,CAAN;cACM/C,IAAN,CAAW,KAAX,EAAkB8C,GAAlB;;KAvBG;;WA2BA,CACL,KADK,EAEL,qBAFK,EAGL,2BAHK,EAIL,kBAJK,EAKL,mBALK,EAML,QANK,EAOL,kBAPK,EAQL,SARK;GA3CqB;;kBAuDd,IAvDc;;kBAyDd,IAzDc;;OA2DzB,IA3DyB;;iBA6Df,IA7De;;WA+DrB;CA/DJ;;ACAP;;AAEA,AAAO,IAAMG,uBAAuB;UAC1B,qBAD0B;SAE3B;eACM,CACT,QADS;GAHqB;;UAQ1B;eACK,CACT,0DADS;GATqB;;WAczB;eACI,CACT,eADS,CADJ;;;;gBAOK,EAPL;;;;;WAaA;GA3ByB;;kBAgClB,IAhCkB;;kBAkClB,IAlCkB;;OAoC7B,IApC6B;;iBAsCnB,IAtCmB;;WAwCzB;CAxCJ;;ACFP;;;AAGA,AAAO,IAAMC,qBAAqB;UACxB,mBADwB;SAEzB;eACM,CACT,UADS;GAHmB;;UAQxB;eACK,CACT,eADS;GATmB;;WAcvB;eACI,CACT,iBADS,EAET,iBAFS,CADJ;;;;gBAQK,EARL;;;;;WAcA;GA5BuB;;kBAiChB;eACH,CACT,CAAC,qCAAD,EAAwC,OAAxC,CADS;GAlCmB;;kBAuChB;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GAxCmB;;OA6C3B;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GA9CmB;;iBAmDjB,IAnDiB;;WAqDvB;CArDJ;;ACHP;;;AAGA,AAAO,IAAMC,iBAAiB;UACpB,eADoB;SAErB;eACM,CACT,eADS;GAHe;;UASpB;eACK,CACT,iBADS;GAVe;;WAgBnB;eACI,CACT,iBADS,CADJ;;;;gBAQK,EARL;;;;;WAcA,CACL,kBADK;GA9BmB;;kBAoCZ;eACH,CACT,CAAC,gCAAD,EAAmC,OAAnC,CADS;GArCe;;kBA0CZ;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA3Ce;;OAgDvB;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GAjDe;;iBAsDb,IAtDa;;WAwDnB;CAxDJ;;ACHP;;;AAGA,AAAO,IAAMC,eAAe;UAClB,aADkB;SAEnB;eACM,CACT,IADS;GAHa;;UASlB;eACK,CACT,qBADS;GAVa;;WAgBjB;eACI,CACT,cADS,CADJ;;;;gBAQK,EARL;;;;;WAcA,CACL,cADK;GA9BiB;;kBAoCV;eACH,CACT,WADS;GArCa;;kBA0CV;eACH;GA3Ca;;OAgDrB;eACQ,CACT,CAAC,0BAAD,EAA6B,OAA7B,CADS;GAjDa;;iBAsDX,IAtDW;;WAwDjB;CAxDJ;;ACHP;;;AAGA,AAAO,IAAMC,iBAAiB;UACpB,eADoB;SAErB;eACM,CACT,sBADS;GAHe;;UASpB;eACK,CACT,oBADS;GAVe;;WAgBnB;eACI;;qBAAA,CADJ;;;;gBAQK,EARL;;;;;WAcA,CACL,iBADK;GA9BmB;;kBAoCZ;eACH,CACT,CAAC,qBAAD,EAAwB,UAAxB,CADS;GArCe;;kBA0CZ;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA3Ce;;OAgDvB;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GAjDe;;iBAuDb,IAvDa;;WAyDnB;CAzDJ;;ACHP;;;AAGA,AAAO,IAAMC,oBAAoB;UACvB,kBADuB;SAExB;eACM,CACT,qBADS;GAHkB;;UASvB;eACK,CACT,gCADS,EACyB,gBADzB;GAVkB;;WAgBtB;eACI,CACT,gBADS,CADJ;;oBAMS,KANT;;;;gBAUK;UACN;KAXC;;;;;WAiBA,CACL,oBADK,EAEL,uEAFK,EAGL,YAHK;GAjCsB;;kBAwCf;eACH,CACT,gBADS;GAzCkB;;kBA+Cf;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GAhDkB;;OAqD1B;eACQ,CACT,CAAC,0BAAD,EAA6B,OAA7B,CADS;GAtDkB;;iBA2DhB,IA3DgB;;WA6DtB;CA7DJ;;ACHP;;;AAGA,AAAO,IAAMC,iBAAiB;UACpB,kBADoB;SAErB;eACM,CACT,gBADS;GAHe;;UASpB;eACK,CACT,eADS,EACQ,KADR;GAVe;;WAgBnB;eACI,CACT,eADS,EAET,gBAFS,CADJ;;;;gBASK,EATL;;;;;WAeA;GA/BmB;;kBAoCZ;eACH,CACT,CAAC,qCAAD,EAAwC,OAAxC,CADS;GArCe;;kBA0CZ;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA3Ce;;OAgDvB;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GAjDe;;iBAsDb,IAtDa;;WAwDnB;CAxDJ;;ACHP;;;AAGA,AAAO,IAAMC,wBAAwB;UAC3B,sBAD2B;SAE5B;eACM,CACT,eADS;GAHsB;;UAS3B;eACK,CACT,CAAC,qBAAD,EAAwB,OAAxB,CADS;GAVsB;;WAgB1B;eACI;;uBAAA,EAGT,kBAHS,CADJ;;;;gBASK,EATL;;;;;WAeA;GA/B0B;;kBAoCnB;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GArCsB;;iBA0CpB,IA1CoB;;WA4C1B;CA5CJ;;ACHP;;;AAGA,AAAO,IAAMC,oBAAoB;UACvB,kBADuB;SAExB;eACM;;KAER,uBAAD,EAA0B,OAA1B,CAFS;GAHkB;;UASvB;eACK,CACT,oCADS;GAVkB;;WAetB;eACI;;yBAAA,EAGT,gBAHS,EAGS,aAHT,EAIT,aAJS,CADJ;;;;gBAUK,EAVL;;;;;WAgBA,CACL,YADK;GA/BsB;;kBAoCf;eACH,CACT,CAAC,+CAAD,EAAkD,UAAlD,CADS;GArCkB;;kBA2Cf;eACH;;KAER,uBAAD,EAA0B,OAA1B,CAFS;GA5CkB;;OAmD1B;eACQ,CACT,CAAC,0BAAD,EAA6B,OAA7B,CADS;GApDkB;;iBAyDhB,IAzDgB;;WA2DtB;CA3DJ;;ACHA,IAAMC,oBAAoB;UACvB,cADuB;SAExB;eACM,CACT,aADS;GAHkB;;UAQvB;eACK,CACT,SADS;GATkB;;WActB;eACI,CACT,eADS,EAET,gBAFS,CADJ;;;;gBAQK;0DAC0C,8CAAC5D,KAAD,EAAW;YACvD6D,YAAY7D,MAAME,IAAN,CAAW,IAAX,EAAiB4D,KAAjB,CAAuB,UAAvB,EAAmC,CAAnC,CAAlB;cACM5D,IAAN,CAAW,KAAX,qCAAmD2D,SAAnD;;KAXG;;;;;WAkBA;GAhCsB;;kBAoCf;eACH,CACT,CAAC,wBAAD,EAA2B,UAA3B,CADS;GArCkB;;kBA0Cf;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA3CkB;;OAgD1B;eACQ;;;GAjDkB;;iBAsDhB;eACF;;;GAvDkB;;WA4DtB;eACI;;;;CA7DR;;ACAP;;;AAGA,AAAO,IAAME,yBAAyB;UAC5B,uBAD4B;SAE7B;eACM,CACT,kBADS;GAHuB;;UAQ5B;eACK,CACT,uBADS;GATuB;;WAc3B;eACI,CACT,2BADS,CADJ;;;;gBAOK,EAPL;;;;;WAaA;GA3B2B;;kBAgCpB;eACH,CACT,CAAC,8BAAD,EAAiC,OAAjC,CADS;GAjCuB;;kBAsCpB;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GAvCuB;;OA4C/B;eACQ,CACT,CAAC,6BAAD,EAAgC,OAAhC,CADS;GA7CuB;;iBAkDrB;eACF;;;GAnDuB;;WAwD3B;eACI;;;;CAzDR;;ACHP;;;AAGA,AAAO,IAAMC,4BAA4B;UAC/B,0BAD+B;SAEhC;eACM,CACT,aADS;GAH0B;;UAQ/B;eACK,CACT,mBADS;GAT0B;;WAc9B;eACI,CACT,mBADS,CADJ;;;;gBAOK;wDACwC,+CAAChE,KAAD,EAAQN,CAAR,EAAc;YACxDuE,OAAOC,KAAK7E,KAAL,CAAWW,MAAME,IAAN,CAAW,YAAX,CAAX,CAAb;YACQ8C,GAFsD,GAE9CiB,KAAKE,OAAL,CAAa,CAAb,CAF8C,CAEtDnB,GAFsD;;YAGxDoB,OAAO1E,EAAE,SAAF,EAAaQ,IAAb,CAAkB,KAAlB,EAAyB8C,GAAzB,CAAb;cACMF,WAAN,CAAkBsB,IAAlB;;KAZG;;;;;WAmBA;GAjC8B;;kBAsCvB;eACH,CACT,CAAC,kCAAD,EAAqC,UAArC,CADS;GAvC0B;;kBA4CvB;eACH,CACT,CAAC,uBAAD,EAA0B,OAA1B,CADS;GA7C0B;;OAkDlC;eACQ,CACT,CAAC,wBAAD,EAA2B,OAA3B,CADS;GAnD0B;;iBAwDxB;eACF;;;GAzD0B;;WA8D9B;eACI;;;;CA/DR;;ACeP,IAAMC,aAAa;eACJnC,cADI;kBAEDG,gBAFC;mBAGAC,kBAHA;iBAIFI,gBAJE;qBAKEK,gBALF;yBAMMI,oBANN;uBAOIC,kBAPJ;mBAQAC,cARA;iBASFC,YATE;mBAUAC,cAVA;sBAWGC,iBAXH;sBAYGC,cAZH;0BAaOC,qBAbP;sBAcGC,iBAdH;kBAeDC,iBAfC;2BAgBQG,sBAhBR;8BAiBWC;CAjB9B,CAoBA;;ACtCA;AACA,AAAO,IAAMM,YAAY,IAAIxG,MAAJ,CAAW,gCAAX,EAA6C,GAA7C,CAAlB;;;;AAIP,AAAO,IAAMyG,aAAa,qBAAnB;;AAEP,AAAO,IAAMC,iBAAiB,CAC5B,wCAD4B,EAE5B,uCAF4B,EAG5B,qCAH4B,EAI5B,oCAJ4B,CAAvB;;;AAQP,AAAO,IAAMC,oBAAoB,CAC/B,OAD+B,EAE/B,QAF+B,EAG/B,UAH+B,EAI/B,MAJ+B,EAK/B,OAL+B,EAM/B,IAN+B,EAO/B,OAP+B,EAQ/B,QAR+B,EAS/B,QAT+B,CAA1B;;;AAaP,AAAO,IAAMC,eAAe,CAAC,OAAD,EAAU,OAAV,CAArB;AACP,AAAO,IAAMC,wBAAwBD,aAAaE,GAAb,CAAiB;eAAgBC,QAAhB;CAAjB,CAA9B;AACP,AAAO,IAAMC,mBAAmBJ,aAAa3G,IAAb,CAAkB,GAAlB,CAAzB;AACP,AAAO,IAAMgH,kBAAkB,CAAC,KAAD,EAAQ,QAAR,EAAkB,MAAlB,EAA0B,OAA1B,EAAmC,IAAnC,EAAyC,KAAzC,CAAxB;AACP,AAAO,IAAMC,qBAAqB,IAAIlH,MAAJ,QAAgBiH,gBAAgBhH,IAAhB,CAAqB,GAArB,CAAhB,SAA+C,GAA/C,CAA3B;;;AAGP,AAAO,IAAMkH,oBAAoB,CAAC,GAAD,CAA1B;AACP,AAAO,IAAMC,yBAAyBD,kBAAkBL,GAAlB,CAAsB;SAAUO,GAAV;CAAtB,EAA6CpH,IAA7C,CAAkD,GAAlD,CAA/B;;;AAGP,AAAO,IAAMqH,2BAA2B,CAAC,IAAD,EAAO,IAAP,EAAa,OAAb,EAAsB,KAAtB,EAA6B,QAA7B,EAAuC,MAAvC,EAA+CrH,IAA/C,CAAoD,GAApD,CAAjC;;;AAGP,IAAMsH,cAAc,CAAC,IAAD,EAAO,IAAP,EAAa,IAAb,EAAmB,IAAnB,EAAyB,IAAzB,CAApB;AACA,AAAO,IAAMC,kBAAkBD,YAAYtH,IAAZ,CAAiB,GAAjB,CAAxB;;;;;;;;AASP,AAAO,IAAMwH,gCAAgC,CAC3C,UAD2C,EAE3C,OAF2C,EAG3C,QAH2C,EAI3C,SAJ2C,EAK3C,SAL2C,EAM3C,KAN2C,EAO3C,gBAP2C,EAQ3C,OAR2C,EAS3C,SAT2C,EAU3C,cAV2C,EAW3C,QAX2C,EAY3C,iBAZ2C,EAa3C,OAb2C,EAc3C,MAd2C;;AAgB3C,QAhB2C,EAiB3C,QAjB2C,EAkB3C,QAlB2C,EAmB3C,OAnB2C;AAoB3C,MApB2C,EAqB3C,MArB2C,EAsB3C,KAtB2C,EAuB3C,UAvB2C,EAwB3C,OAxB2C,EAyB3C,YAzB2C,EA0B3C,UA1B2C;AA2B3C,2BA3B2C;AA4B3C,OA5B2C,EA6B3C,eA7B2C,EA8B3C,SA9B2C,EA+B3C,QA/B2C,EAgC3C,QAhC2C,EAiC3C,KAjC2C,EAkC3C,OAlC2C,EAmC3C,UAnC2C,EAoC3C,SApC2C,EAqC3C,UArC2C,EAsC3C,SAtC2C,EAuC3C,SAvC2C,EAwC3C,OAxC2C,CAAtC;;;;;;;;;;;;;AAsDP,AAAO,IAAMC,gCAAgC,CAC3C,KAD2C,EAE3C,SAF2C,EAG3C,MAH2C,EAI3C,WAJ2C,EAK3C,QAL2C,EAM3C,SAN2C,EAO3C,qBAP2C,EAQ3C,QAR2C;AAS3C,OAT2C,EAU3C,QAV2C,EAW3C,OAX2C,EAY3C,MAZ2C,EAa3C,MAb2C,EAc3C,OAd2C,EAe3C,QAf2C,CAAtC;;;;;AAqBP,AAAO,IAAMC,sBAAsB,CACjC,GADiC,EAEjC,YAFiC,EAGjC,IAHiC,EAIjC,KAJiC,EAKjC,KALiC,EAMjC,GANiC,EAOjC,KAPiC,EAQjC,OARiC,EASjC1H,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM2H,yBAAyB,CACpC,IADoC,EAEpC,GAFoC,EAGpC,GAHoC,EAIpC,OAJoC,EAKpC,IALoC,EAMpC,MANoC,EAOpC,MAPoC,EAQpC,UARoC,EASpC,OAToC,EAUpC,KAVoC,EAWpC,MAXoC,EAYpC,MAZoC,CAA/B;;AAeP,AAAO,IAAMC,4BACX,IAAI7H,MAAJ,QAAgB4H,uBAAuB3H,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;AAGP,AAYA,AAAO,IAAM6H,cAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,iBAAiB,IAAI/H,MAAJ,CAAW8H,YAAY7H,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAM+H,uBAAuB,CAClC,SADkC,EAElC,gBAFkC,EAGlC,iBAHkC,EAIlC,MAJkC,EAKlC,MALkC,EAMlC,SANkC,EAOlC,qBAPkC,EAQlC,OARkC,EASlC,QATkC,EAUlC,MAVkC,EAWlC,QAXkC,EAYlC,MAZkC,EAalC,YAbkC,EAclC,WAdkC,EAelC,MAfkC,EAgBlC,OAhBkC,EAiBlC,MAjBkC,EAkBlC,UAlBkC;AAmBlC,SAnBkC,CAA7B;;;AAuBP,AAAO,IAAMC,oBAAoB,IAAIjI,MAAJ,CAAWgI,qBAAqB/H,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA;;;;AAIA,AAAO,IAAMiI,uBAAuB,CAClC,OADkC,EAElC,QAFkC,EAGlC,QAHkC,EAIlC,KAJkC,EAKlC,UALkC,EAMlC,QANkC,EAOlC,QAPkC,EAQlC,OARkC,EASlC,MATkC,EAUlC,OAVkC,EAWlC,SAXkC,EAYlC,YAZkC,EAalC,SAbkC,EAclC,MAdkC,EAelC,QAfkC,EAgBlC,OAhBkC,EAiBlC,MAjBkC,EAkBlC,MAlBkC,EAmBlC,SAnBkC,EAoBlC,UApBkC;AAqBlC,MArBkC,EAsBlC,QAtBkC,EAuBlC,UAvBkC,EAwBlC,MAxBkC,EAyBlC,MAzBkC,EA0BlC,MA1BkC,EA2BlC,UA3BkC;AA4BlC,mBA5BkC,EA6BlC,MA7BkC,EA8BlC,WA9BkC,EA+BlC,MA/BkC,EAgClC,UAhCkC,EAiClC,OAjCkC,EAkClC,MAlCkC,EAmClC,OAnCkC,EAoClC,UApCkC;AAqClC,OArCkC,EAsClC,KAtCkC;AAuClC,SAvCkC,EAwClC,SAxCkC,EAyClC,cAzCkC;AA0ClC,QA1CkC,EA2ClC,WA3CkC,EA4ClC,OA5CkC,EA6ClC,UA7CkC,EA8ClC,UA9CkC,EA+ClC,MA/CkC,EAgDlC,SAhDkC,EAiDlC,SAjDkC,EAkDlC,OAlDkC,EAmDlC,KAnDkC,EAoDlC,SApDkC,EAqDlC,MArDkC,EAsDlC,OAtDkC,EAuDlC,QAvDkC,CAA7B;;AA0DP,AAAO,IAAMC,oBAAoB,IAAInI,MAAJ,CAAWkI,qBAAqBjI,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMmI,iBAAiB,wCAAvB;;AAEP,AAGA;;AAEA,AAAO,IAAMC,wBAAwB,CACnC,OADmC,EAEnC,SAFmC,EAGnC,SAHmC,EAInC,SAJmC,EAKnC,QALmC,EAMnC,OANmC,EAOnC,OAPmC,EAQnC,OARmC,EASnC,KATmC,EAUnC,OAVmC,EAWnC,MAXmC,EAYnC,QAZmC,EAanC,KAbmC,EAcnC,iBAdmC,CAA9B;AAgBP,AAAO,IAAMC,2BAA2B,IAAItI,MAAJ,CAAWqI,sBAAsBpI,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;AAGP,AAAO,IAAMsI,UAAU,IAAIvI,MAAJ,CAAW,iBAAX,EAA8B,GAA9B,CAAhB;;AAEP,AAMA,AAIA,AAIA,AAGA,AAGA;;AAEA,AAAO,IAAMwI,mBAAmB,CAC9B,SAD8B,EAE9B,OAF8B,EAG9B,YAH8B,EAI9B,MAJ8B,EAK9B,IAL8B,EAM9B,QAN8B,EAO9B,QAP8B,EAQ9B,SAR8B,EAS9B,KAT8B,EAU9B,UAV8B,EAW9B,IAX8B,EAY9B,KAZ8B,EAa9B,IAb8B,EAc9B,IAd8B,EAe9B,OAf8B,EAgB9B,UAhB8B,EAiB9B,YAjB8B,EAkB9B,QAlB8B,EAmB9B,QAnB8B,EAoB9B,MApB8B,EAqB9B,IArB8B,EAsB9B,IAtB8B,EAuB9B,IAvB8B,EAwB9B,IAxB8B,EAyB9B,IAzB8B,EA0B9B,IA1B8B,EA2B9B,QA3B8B,EA4B9B,QA5B8B,EA6B9B,IA7B8B,EA8B9B,IA9B8B,EA+B9B,KA/B8B,EAgC9B,QAhC8B,EAiC9B,IAjC8B,EAkC9B,QAlC8B,EAmC9B,GAnC8B,EAoC9B,KApC8B,EAqC9B,UArC8B,EAsC9B,SAtC8B,EAuC9B,OAvC8B,EAwC9B,OAxC8B,EAyC9B,UAzC8B,EA0C9B,OA1C8B,EA2C9B,IA3C8B,EA4C9B,OA5C8B,EA6C9B,IA7C8B,EA8C9B,IA9C8B,EA+C9B,OA/C8B,CAAzB;AAiDP,AAAO,IAAMC,sBAAsB,IAAIzI,MAAJ,QAAgBwI,iBAAiBvI,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAMyI,sBAAsBjB,8BAA8BxH,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM0I,uBAAuB,IAAI3I,MAAJ,CAAW0I,mBAAX,EAAgC,GAAhC,CAA7B;;AAEP,IAAME,sBAAsBlB,8BAA8BzH,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM4I,uBAAuB,IAAI7I,MAAJ,CAAW4I,mBAAX,EAAgC,GAAhC,CAA7B,CAEP,AAGA,AACA,AACA,AAEA;;ACtYe,SAASE,uBAAT,CAAiClH,CAAjC,EAAoC;;;;;;;;;;IAU/C,GAAF,EAAOmH,GAAP,CAAW,GAAX,EAAgBhH,IAAhB,CAAqB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;QAC9BC,QAAQN,EAAEK,IAAF,CAAd;QACM+G,UAAU9G,MAAME,IAAN,CAAW,OAAX,CAAhB;QACM6G,KAAK/G,MAAME,IAAN,CAAW,IAAX,CAAX;QACI,CAAC6G,EAAD,IAAO,CAACD,OAAZ,EAAqB;;QAEfE,cAAgBF,WAAW,EAA3B,WAAiCC,MAAM,EAAvC,CAAN;QACIJ,qBAAqB1H,IAArB,CAA0B+H,UAA1B,CAAJ,EAA2C;;KAA3C,MAEO,IAAIP,qBAAqBxH,IAArB,CAA0B+H,UAA1B,CAAJ,EAA2C;YAC1C5F,MAAN;;GAVJ;;SAcO1B,CAAP;;;AC3BF;;;;;;;;;;AAUA,AAAe,SAASuH,OAAT,CAAiBvH,CAAjB,EAAoB;MAC7BwH,aAAa,KAAjB;IACE,IAAF,EAAQrH,IAAR,CAAa,UAACgB,KAAD,EAAQsG,OAAR,EAAoB;QACzBC,cAAc1H,EAAEyH,OAAF,EAAWE,IAAX,GAAkBpJ,GAAlB,CAAsB,CAAtB,CAApB;;QAEImJ,eAAeA,YAAYhF,OAAZ,KAAwB,IAA3C,EAAiD;mBAClC,IAAb;QACE+E,OAAF,EAAW/F,MAAX;KAFF,MAGO,IAAI8F,UAAJ,EAAgB;mBACR,KAAb;;mBAEaC,OAAb,EAAsBzH,CAAtB,EAAyB,IAAzB;;GATJ;;SAaOA,CAAP;;;ACzBF;;;;;;;;;;;AAWA,AAAe,SAAS4H,YAAT,CAAsBvH,IAAtB,EAA4BL,CAA5B,EAA2C;MAAZ6H,EAAY,uEAAP,KAAO;;MAClDvH,QAAQN,EAAEK,IAAF,CAAd;;MAEIwH,EAAJ,EAAQ;QACFC,UAAUzH,KAAK0H,WAAnB;QACMC,IAAIhI,EAAE,SAAF,CAAV;;;;WAIO8H,WAAW,EAAEA,QAAQpF,OAAR,IAAmBmE,oBAAoBtH,IAApB,CAAyBuI,QAAQpF,OAAjC,CAArB,CAAlB,EAAmF;UAC3EqF,cAAcD,QAAQC,WAA5B;QACED,OAAF,EAAWG,QAAX,CAAoBD,CAApB;gBACUD,WAAV;;;UAGI3E,WAAN,CAAkB4E,CAAlB;UACMtG,MAAN;WACO1B,CAAP;;;SAGKA,CAAP;;;AC7BF,SAASkI,WAAT,CAAqBlI,CAArB,EAAwB;IACpB,KAAF,EAASG,IAAT,CAAc,UAACgB,KAAD,EAAQgH,GAAR,EAAgB;QACtBC,OAAOpI,EAAEmI,GAAF,CAAb;QACME,cAAcD,KAAK9F,QAAL,CAAcyD,mBAAd,EAAmCxD,MAAnC,KAA8C,CAAlE;;QAEI8F,WAAJ,EAAiB;oBACDD,IAAd,EAAoBpI,CAApB,EAAuB,GAAvB;;GALJ;;SASOA,CAAP;;;AAGF,SAASsI,YAAT,CAAsBtI,CAAtB,EAAyB;IACrB,MAAF,EAAUG,IAAV,CAAe,UAACgB,KAAD,EAAQoH,IAAR,EAAiB;QACxBC,QAAQxI,EAAEuI,IAAF,CAAd;QACMF,cAAcG,MAAM1F,OAAN,CAAc,QAAd,EAAwBP,MAAxB,KAAmC,CAAvD;QACI8F,WAAJ,EAAiB;oBACDG,KAAd,EAAqBxI,CAArB,EAAwB,GAAxB;;GAJJ;;SAQOA,CAAP;;;;;;;;;;;;;;;AAeF,AAAe,SAASyI,mBAAT,CAA6BzI,CAA7B,EAAgC;MACzCuH,QAAQvH,CAAR,CAAJ;MACIkI,YAAYlI,CAAZ,CAAJ;MACIsI,aAAatI,CAAb,CAAJ;;SAEOA,CAAP;;;AC9Ca,SAAS0I,aAAT,CAAuBpI,KAAvB,EAA8BN,CAA9B,EAA4C;MAAXyF,GAAW,uEAAL,GAAK;;MACnDpF,OAAOC,MAAM/B,GAAN,CAAU,CAAV,CAAb;MACI,CAAC8B,IAAL,EAAW;WACFL,CAAP;;;mBAEkBM,MAAM/B,GAAN,CAAU,CAAV,CALqC;;MAKjDyC,OALiD,cAKjDA,OALiD;;MAMnD2H,eAAe,iBAAgB3H,OAAhB,EACQkE,GADR,CACY;WAAU0D,GAAV,SAAiB5H,QAAQ4H,GAAR,CAAjB;GADZ,EAEQvK,IAFR,CAEa,GAFb,CAArB;;QAIM+E,WAAN,OAAsBqC,GAAtB,SAA6BkD,YAA7B,SAA6CrI,MAAMkB,QAAN,EAA7C,UAAkEiE,GAAlE;SACOzF,CAAP;;;ACTF,SAAS6I,cAAT,CAAwBnE,IAAxB,EAA8B1E,CAA9B,EAAiC;MACzB8I,SAASC,SAASrE,KAAKlE,IAAL,CAAU,QAAV,CAAT,EAA8B,EAA9B,CAAf;MACM+C,QAAQwF,SAASrE,KAAKlE,IAAL,CAAU,OAAV,CAAT,EAA6B,EAA7B,KAAoC,EAAlD;;;;;MAKI,CAACsI,UAAU,EAAX,IAAiB,EAAjB,IAAuBvF,QAAQ,EAAnC,EAAuC;SAChC7B,MAAL;GADF,MAEO,IAAIoH,MAAJ,EAAY;;;;SAIZrI,UAAL,CAAgB,QAAhB;;;SAGKT,CAAP;;;;;AAKF,SAASgJ,aAAT,CAAuBtE,IAAvB,EAA6B1E,CAA7B,EAAgC;MAC1B4E,UAAUrF,IAAV,CAAemF,KAAKlE,IAAL,CAAU,KAAV,CAAf,CAAJ,EAAsC;SAC/BkB,MAAL;;;SAGK1B,CAAP;;;AAGF,AAAe,SAASiJ,WAAT,CAAqBC,QAArB,EAA+BlJ,CAA/B,EAAkC;WACtCuB,IAAT,CAAc,KAAd,EAAqBpB,IAArB,CAA0B,UAACgB,KAAD,EAAQJ,GAAR,EAAgB;QAClC2D,OAAO1E,EAAEe,GAAF,CAAb;;mBAEe2D,IAAf,EAAqB1E,CAArB;kBACc0E,IAAd,EAAoB1E,CAApB;GAJF;;SAOOA,CAAP;;;AClCa,SAASmJ,UAAT,CAAoBC,OAApB,EAA6BpJ,CAA7B,EAA2C;MAAXqJ,IAAW,uEAAJ,EAAI;;MACpDA,KAAK9G,MAAL,KAAgB,CAApB,EAAuB;WACduC,cAAP;;;IAGAuE,KAAKhL,IAAL,CAAU,GAAV,CAAF,EAAkB+K,OAAlB,EAA2BE,QAA3B,CAAoCzE,UAApC;;SAEO7E,CAAP;;;ACPa,SAASuJ,aAAT,CAAuBH,OAAvB,EAAgCpJ,CAAhC,EAA8C;MAAXqJ,IAAW,uEAAJ,EAAI;;MACvDA,KAAK9G,MAAL,KAAgB,CAApB,EAAuB;WACdwC,iBAAP;;;;;IAKAsE,KAAKhL,IAAL,CAAU,GAAV,CAAF,EAAkB+K,OAAlB,EAA2BjC,GAA3B,OAAmCtC,UAAnC,EAAiDnD,MAAjD;;;UAGMmD,UAAN,EAAoBuE,OAApB,EAA6BI,WAA7B,CAAyC3E,UAAzC;;SAEO7E,CAAP;;;ACfF;;;;AAGA,AAAe,SAASyJ,UAAT,CAAoBL,OAApB,EAA6BpJ,CAA7B,EAAgC;MACvC0J,SAAS1J,EAAE,IAAF,EAAQoJ,OAAR,CAAf;;MAEIM,OAAOnH,MAAP,GAAgB,CAApB,EAAuB;WACdpC,IAAP,CAAY,UAACgB,KAAD,EAAQd,IAAR;aAAiBL,EAAEK,IAAF,EAAQqB,MAAR,EAAjB;KAAZ;GADF,MAEO;WACEvB,IAAP,CAAY,UAACgB,KAAD,EAAQd,IAAR,EAAiB;oBACbL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0B,IAA1B;KADF;;;SAKKA,CAAP;;;ACdF,SAAS2J,qBAAT,CAA+BT,QAA/B,EAAyC;;WAE9B3H,IAAT,CAAc,GAAd,EAAmBpB,IAAnB,CAAwB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;SAClCW,OAAL,GAAe,iBAAgBX,KAAKW,OAArB,EAA8B4I,MAA9B,CAAqC,UAACC,GAAD,EAAMrJ,IAAN,EAAe;UAC7D8E,mBAAmB/F,IAAnB,CAAwBiB,IAAxB,CAAJ,EAAmC;4BACrBqJ,GAAZ,sBAAkBrJ,IAAlB,EAAyBH,KAAKW,OAAL,CAAaR,IAAb,CAAzB;;;aAGKqJ,GAAP;KALa,EAMZ,EANY,CAAf;GADF;;SAUOX,QAAP;;;;;;;;;;AAUF,AAAe,SAASY,eAAT,CAAyBZ,QAAzB,EAAmC;;;;UAIxCa,GAAR,CAAY,KAAZ,EAAmBb,SAASc,MAAT,GAAkBzH,MAArC;SACOoH,sBACLT,SAASc,MAAT,GAAkBzH,MAAlB,GACE2G,SAASc,MAAT,EADF,GACsBd,QAFjB,CAAP;;;AC7Ba,SAASe,WAAT,CAAqBf,QAArB,EAA+BlJ,CAA/B,EAAkC;WACtCuB,IAAT,CAAc,GAAd,EAAmBpB,IAAnB,CAAwB,UAACgB,KAAD,EAAQ6G,CAAR,EAAc;QAC9BkC,KAAKlK,EAAEgI,CAAF,CAAX;QACIkC,GAAG3I,IAAH,CAAQ,aAAR,EAAuBgB,MAAvB,KAAkC,CAAlC,IAAuC2H,GAAGC,IAAH,GAAUC,IAAV,OAAqB,EAAhE,EAAoEF,GAAGxI,MAAH;GAFtE;;SAKO1B,CAAP;;;ACNF;;;;;;AAMA,AAAO,IAAM6F,kCAAgC,CAC3C,UAD2C,EAE3C,OAF2C,EAG3C,QAH2C,EAI3C,SAJ2C,EAK3C,SAL2C,EAM3C,KAN2C,EAO3C,gBAP2C,EAQ3C,OAR2C,EAS3C,SAT2C,EAU3C,cAV2C,EAW3C,QAX2C,EAY3C,iBAZ2C,EAa3C,OAb2C,EAc3C,MAd2C,EAe3C,MAf2C,EAgB3C,QAhB2C,EAiB3C,QAjB2C,EAkB3C,QAlB2C,EAmB3C,OAnB2C;AAoB3C,MApB2C,EAqB3C,MArB2C,EAsB3C,KAtB2C,EAuB3C,OAvB2C,EAwB3C,YAxB2C,EAyB3C,UAzB2C;AA0B3C,2BA1B2C;AA2B3C,OA3B2C,EA4B3C,eA5B2C,EA6B3C,SA7B2C,EA8B3C,QA9B2C,EA+B3C,QA/B2C,EAgC3C,KAhC2C,EAiC3C,OAjC2C,EAkC3C,UAlC2C,EAmC3C,SAnC2C,EAoC3C,UApC2C,EAqC3C,SArC2C,EAsC3C,OAtC2C,CAAtC;;;;;;;;;;;;;AAoDP,AAAO,IAAMC,kCAAgC,CAC3C,KAD2C,EAE3C,SAF2C,EAG3C,MAH2C,EAI3C,WAJ2C,EAK3C,QAL2C,EAM3C,SAN2C,EAO3C,qBAP2C,EAQ3C,QAR2C;AAS3C,OAT2C,EAU3C,QAV2C,EAW3C,OAX2C,EAY3C,MAZ2C,EAa3C,MAb2C,EAc3C,OAd2C,EAe3C,QAf2C,CAAtC;;;;;AAqBP,AAAO,IAAMC,wBAAsB,CACjC,GADiC,EAEjC,YAFiC,EAGjC,IAHiC,EAIjC,KAJiC,EAKjC,KALiC,EAMjC,GANiC,EAOjC,KAPiC,EAQjC,OARiC,EASjC1H,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM2H,2BAAyB,CACpC,IADoC,EAEpC,GAFoC,EAGpC,GAHoC,EAIpC,OAJoC,EAKpC,IALoC,EAMpC,MANoC,EAOpC,MAPoC,EAQpC,UARoC,EASpC,OAToC,EAUpC,KAVoC,EAWpC,MAXoC,EAYpC,MAZoC,CAA/B;;AAeP,AAAO,IAAMC,8BACX,IAAI7H,MAAJ,QAAgB4H,yBAAuB3H,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;;;;AAMP,AAAO,IAAMgM,4BAA0B,CACrC,CAAC,SAAD,EAAY,gBAAZ,CADqC,EAErC,CAAC,OAAD,EAAU,gBAAV,CAFqC,EAGrC,CAAC,QAAD,EAAW,gBAAX,CAHqC,EAIrC,CAAC,OAAD,EAAU,WAAV,CAJqC,EAKrC,CAAC,OAAD,EAAU,YAAV,CALqC,EAMrC,CAAC,OAAD,EAAU,YAAV,CANqC,CAAhC;;AASP,AAAO,IAAMnE,gBAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,mBAAiB,IAAI/H,MAAJ,CAAW8H,cAAY7H,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAM+H,yBAAuB,CAClC,SADkC,EAElC,gBAFkC,EAGlC,iBAHkC,EAIlC,MAJkC,EAKlC,MALkC,EAMlC,SANkC,EAOlC,qBAPkC,EAQlC,OARkC,EASlC,QATkC,EAUlC,MAVkC,EAWlC,QAXkC,EAYlC,MAZkC,EAalC,YAbkC,EAclC,WAdkC,EAelC,MAfkC,EAgBlC,OAhBkC,EAiBlC,MAjBkC,EAkBlC,UAlBkC;AAmBlC,SAnBkC,CAA7B;;;AAuBP,AAAO,IAAMC,sBAAoB,IAAIjI,MAAJ,CAAWgI,uBAAqB/H,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMiM,sBAAoB,IAAIlM,MAAJ,CAAW,qBAAX,EAAkC,GAAlC,CAA1B;;;;;;AAMP,AAAO,IAAMkI,yBAAuB,CAClC,OADkC,EAElC,QAFkC,EAGlC,QAHkC,EAIlC,KAJkC,EAKlC,UALkC,EAMlC,QANkC,EAOlC,QAPkC,EAQlC,OARkC,EASlC,MATkC,EAUlC,OAVkC,EAWlC,SAXkC,EAYlC,YAZkC,EAalC,SAbkC,EAclC,MAdkC,EAelC,QAfkC,EAgBlC,OAhBkC,EAiBlC,MAjBkC,EAkBlC,MAlBkC,EAmBlC,SAnBkC,EAoBlC,UApBkC;AAqBlC,MArBkC,EAsBlC,QAtBkC,EAuBlC,UAvBkC,EAwBlC,MAxBkC,EAyBlC,MAzBkC,EA0BlC,MA1BkC,EA2BlC,UA3BkC;AA4BlC,mBA5BkC,EA6BlC,MA7BkC,EA8BlC,WA9BkC,EA+BlC,MA/BkC,EAgClC,UAhCkC,EAiClC,OAjCkC,EAkClC,MAlCkC,EAmClC,OAnCkC,EAoClC,UApCkC;AAqClC,OArCkC,EAsClC,KAtCkC;AAuClC,SAvCkC,EAwClC,SAxCkC,EAyClC,cAzCkC;AA0ClC,QA1CkC,EA2ClC,WA3CkC,EA4ClC,OA5CkC,EA6ClC,UA7CkC,EA8ClC,UA9CkC,EA+ClC,MA/CkC,EAgDlC,SAhDkC,EAiDlC,SAjDkC,EAkDlC,OAlDkC,EAmDlC,KAnDkC,EAoDlC,SApDkC,EAqDlC,MArDkC,EAsDlC,OAtDkC,EAuDlC,QAvDkC,CAA7B;;AA0DP,AAAO,IAAMC,sBAAoB,IAAInI,MAAJ,CAAWkI,uBAAqBjI,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA,AAGA,AAGA;;AAEA,AAAO,IAAMuI,qBAAmB,CAC9B,SAD8B,EAE9B,OAF8B,EAG9B,YAH8B,EAI9B,MAJ8B,EAK9B,IAL8B,EAM9B,QAN8B,EAO9B,QAP8B,EAQ9B,SAR8B,EAS9B,KAT8B,EAU9B,UAV8B,EAW9B,IAX8B,EAY9B,KAZ8B,EAa9B,IAb8B,EAc9B,IAd8B,EAe9B,OAf8B,EAgB9B,UAhB8B,EAiB9B,YAjB8B,EAkB9B,QAlB8B,EAmB9B,QAnB8B,EAoB9B,MApB8B,EAqB9B,IArB8B,EAsB9B,IAtB8B,EAuB9B,IAvB8B,EAwB9B,IAxB8B,EAyB9B,IAzB8B,EA0B9B,IA1B8B,EA2B9B,QA3B8B,EA4B9B,QA5B8B,EA6B9B,IA7B8B,EA8B9B,IA9B8B,EA+B9B,KA/B8B,EAgC9B,QAhC8B,EAiC9B,IAjC8B,EAkC9B,QAlC8B,EAmC9B,GAnC8B,EAoC9B,KApC8B,EAqC9B,UArC8B,EAsC9B,SAtC8B,EAuC9B,OAvC8B,EAwC9B,OAxC8B,EAyC9B,UAzC8B,EA0C9B,OA1C8B,EA2C9B,IA3C8B,EA4C9B,OA5C8B,EA6C9B,IA7C8B,EA8C9B,IA9C8B,EA+C9B,OA/C8B,CAAzB;AAiDP,AAAO,IAAMC,wBAAsB,IAAIzI,MAAJ,QAAgBwI,mBAAiBvI,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAMyI,wBAAsBjB,gCAA8BxH,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,IAAM2I,wBAAsBlB,gCAA8BzH,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,AAGA,AAAO,IAAMkM,yBAAuB,IAAInM,MAAJ,CAAW,mBAAX,EAAgC,GAAhC,CAA7B;AACP,AAAO,IAAMoM,uBAAqB,IAAIpM,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAA3B;AACP,AAAO,IAAMqM,aAAW,IAAIrM,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB,CAEP;;AC3SA;AACA,AAAe,SAASsM,SAAT,CAAmBrK,IAAnB,EAAyB;MAChC+G,UAAU/G,KAAKG,IAAL,CAAU,OAAV,CAAhB;MACM6G,KAAKhH,KAAKG,IAAL,CAAU,IAAV,CAAX;MACImK,QAAQ,CAAZ;;MAEItD,EAAJ,EAAQ;;QAEFhB,oBAAkB9G,IAAlB,CAAuB8H,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;QAEEd,oBAAkBhH,IAAlB,CAAuB8H,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;;;MAIAD,OAAJ,EAAa;QACPuD,UAAU,CAAd,EAAiB;;;UAGXtE,oBAAkB9G,IAAlB,CAAuB6H,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;UAEEb,oBAAkBhH,IAAlB,CAAuB6H,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;;;;;;QAOAjB,iBAAe5G,IAAf,CAAoB6H,OAApB,CAAJ,EAAkC;eACvB,EAAT;;;;;;;QAOEkD,oBAAkB/K,IAAlB,CAAuB6H,OAAvB,CAAJ,EAAqC;eAC1B,EAAT;;;;SAIGuD,KAAP;;;ACpDF;;;AAGA,AAAe,SAASC,QAAT,CAAkBtK,KAAlB,EAAyB;SAC/BuK,WAAWvK,MAAME,IAAN,CAAW,OAAX,CAAX,KAAmC,IAA1C;;;ACJF;AACA,AAAe,SAASsK,WAAT,CAAqBX,IAArB,EAA2B;SACjC,CAACA,KAAKY,KAAL,CAAW,IAAX,KAAoB,EAArB,EAAyBxI,MAAhC;;;ACFF,IAAMyI,QAAQ,IAAI5M,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAd;;AAEA,AAAe,SAAS6M,WAAT,CAAqBC,UAArB,EAAgD;MAAfxI,OAAe,uEAAL,GAAK;;MACvDyI,SAASD,aAAa,EAA5B;;MAEIC,SAAS,CAAb,EAAgB;QACVC,oBAAJ;;;;;;;QAOIJ,MAAMzL,IAAN,CAAWmD,OAAX,CAAJ,EAAyB;oBACTyI,SAAS,CAAvB;KADF,MAEO;oBACSA,SAAS,IAAvB;;;WAGKE,KAAKC,GAAL,CAASD,KAAKE,GAAL,CAASH,WAAT,EAAsB,CAAtB,CAAT,EAAmC,CAAnC,CAAP;;;SAGK,CAAP;;;ACjBF;;AAEA,AAAe,SAASI,cAAT,CAAwBnL,IAAxB,EAA8B;MACvCsK,QAAQ,CAAZ;MACMR,OAAO9J,KAAK8J,IAAL,GAAYC,IAAZ,EAAb;MACMc,aAAaf,KAAK5H,MAAxB;;;MAGI2I,aAAa,EAAjB,EAAqB;WACZ,CAAP;;;;WAIOJ,YAAYX,IAAZ,CAAT;;;;WAISc,YAAYC,UAAZ,CAAT;;;;;;MAMIf,KAAKsB,KAAL,CAAW,CAAC,CAAZ,MAAmB,GAAvB,EAA4B;aACjB,CAAT;;;SAGKd,KAAP;;;AC/Ba,SAASe,QAAT,CAAkBpL,KAAlB,EAAyBN,CAAzB,EAA4B2K,KAA5B,EAAmC;QAC1CnK,IAAN,CAAW,OAAX,EAAoBmK,KAApB;SACOrK,KAAP;;;ACEa,SAASqL,QAAT,CAAkBrL,KAAlB,EAAyBN,CAAzB,EAA4B4L,MAA5B,EAAoC;MAC7C;QACIjB,QAAQkB,eAAevL,KAAf,EAAsBN,CAAtB,IAA2B4L,MAAzC;aACStL,KAAT,EAAgBN,CAAhB,EAAmB2K,KAAnB;GAFF,CAGE,OAAOmB,CAAP,EAAU;;;;SAILxL,KAAP;;;ACXF;AACA,AAAe,SAASyL,WAAT,CAAqB1L,IAArB,EAA2BL,CAA3B,EAA8B2K,KAA9B,EAAqC;MAC5CX,SAAS3J,KAAK2J,MAAL,EAAf;MACIA,MAAJ,EAAY;aACDA,MAAT,EAAiBhK,CAAjB,EAAoB2K,QAAQ,IAA5B;;;SAGKtK,IAAP;;;ACFF;;;AAGA,AAAe,SAASwL,cAAT,CAAwBvL,KAAxB,EAA+BN,CAA/B,EAAsD;MAApBgM,WAAoB,uEAAN,IAAM;;MAC/DrB,QAAQC,SAAStK,KAAT,CAAZ;;MAEIqK,KAAJ,EAAW;WACFA,KAAP;;;UAGMsB,UAAU3L,KAAV,CAAR;;MAEI0L,WAAJ,EAAiB;aACNtB,UAAUpK,KAAV,CAAT;;;cAGUA,KAAZ,EAAmBN,CAAnB,EAAsB2K,KAAtB;;SAEOA,KAAP;;;AClBF;;AAEA,AAAe,SAASsB,SAAT,CAAmB3L,KAAnB,EAA0B;mBACnBA,MAAM/B,GAAN,CAAU,CAAV,CADmB;;MAC/BmE,OAD+B,cAC/BA,OAD+B;;;;;;MAMnC6H,uBAAqBhL,IAArB,CAA0BmD,OAA1B,CAAJ,EAAwC;WAC/B8I,eAAelL,KAAf,CAAP;GADF,MAEO,IAAIoC,YAAY,KAAhB,EAAuB;WACrB,CAAP;GADK,MAEA,IAAI8H,qBAAmBjL,IAAnB,CAAwBmD,OAAxB,CAAJ,EAAsC;WACpC,CAAP;GADK,MAEA,IAAI+H,WAASlL,IAAT,CAAcmD,OAAd,CAAJ,EAA4B;WAC1B,CAAC,CAAR;GADK,MAEA,IAAIA,YAAY,IAAhB,EAAsB;WACpB,CAAC,CAAR;;;SAGK,CAAP;;;ACjBF,SAAS4F,cAAT,CAAsBhI,KAAtB,EAA6BN,CAA7B,EAAgC;MAC1BM,MAAM/B,GAAN,CAAU,CAAV,CAAJ,EAAkB;qBACI+B,MAAM/B,GAAN,CAAU,CAAV,CADJ;;QACRmE,OADQ,cACRA,OADQ;;;QAGZA,YAAY,MAAhB,EAAwB;;oBAERpC,KAAd,EAAqBN,CAArB,EAAwB,KAAxB;;;;;AAKN,SAASkM,UAAT,CAAoB5L,KAApB,EAA2BN,CAA3B,EAA8B2K,KAA9B,EAAqC;MAC/BrK,KAAJ,EAAW;mBACIA,KAAb,EAAoBN,CAApB;aACSM,KAAT,EAAgBN,CAAhB,EAAmB2K,KAAnB;;;;AAIJ,SAASwB,OAAT,CAAiBnM,CAAjB,EAAoBgM,WAApB,EAAiC;IAC7B,QAAF,EAAY7E,GAAZ,CAAgB,SAAhB,EAA2BhH,IAA3B,CAAgC,UAACgB,KAAD,EAAQd,IAAR,EAAiB;;;QAG3CC,QAAQN,EAAEK,IAAF,CAAZ;YACQqL,SAASpL,KAAT,EAAgBN,CAAhB,EAAmB6L,eAAevL,KAAf,EAAsBN,CAAtB,EAAyBgM,WAAzB,CAAnB,CAAR;;QAEMnJ,UAAUvC,MAAM0J,MAAN,EAAhB;QACMoC,WAAWH,UAAU3L,KAAV,CAAjB;;eAEWuC,OAAX,EAAoB7C,CAApB,EAAuBoM,QAAvB,EAAiCJ,WAAjC;QACInJ,OAAJ,EAAa;;;iBAGAA,QAAQmH,MAAR,EAAX,EAA6BhK,CAA7B,EAAgCoM,WAAW,CAA3C,EAA8CJ,WAA9C;;GAbJ;;SAiBOhM,CAAP;;;;;AAKF,AAAe,SAASqM,YAAT,CAAsBrM,CAAtB,EAA6C;MAApBgM,WAAoB,uEAAN,IAAM;;;;4BAGlC/K,OAAxB,CAAgC,gBAAqC;;;QAAnCqL,cAAmC;QAAnBC,aAAmB;;MAC9DD,cAAL,SAAuBC,aAAvB,EAAwCpM,IAAxC,CAA6C,UAACgB,KAAD,EAAQd,IAAR,EAAiB;eACnDL,EAAEK,IAAF,EAAQ2J,MAAR,CAAesC,cAAf,CAAT,EAAyCtM,CAAzC,EAA4C,EAA5C;KADF;GADF;;;;;;;UAWQA,CAAR,EAAWgM,WAAX;UACQhM,CAAR,EAAWgM,WAAX;;SAEOhM,CAAP;;;ACpEF,IAAMwM,eAAe,SAArB;;AAEA,AAAe,SAASC,eAAT,CAAyBtC,IAAzB,EAA+B;SACrCA,KAAK3G,OAAL,CAAagJ,YAAb,EAA2B,GAA3B,EAAgCpC,IAAhC,EAAP;;;ACHF;;;;;AAKA,AAAe,SAASsC,cAAT,CAAwBlN,GAAxB,EAA6BmN,SAA7B,EAAwC;MAC/CC,UAAUD,UAAUpL,IAAV,CAAe;WAAMsL,GAAGtN,IAAH,CAAQC,GAAR,CAAN;GAAf,CAAhB;MACIoN,OAAJ,EAAa;WACJA,QAAQE,IAAR,CAAatN,GAAb,EAAkB,CAAlB,CAAP;;;SAGK,IAAP;;;ACXF;;;;;;;;;;;;;;;;AAgBA,AAAO,IAAMuN,kBAAkB,IAAI3O,MAAJ,CAAW,0EAAX,EAAuF,GAAvF,CAAxB;;AAEP,AAAO,IAAM4O,eAAe,QAArB;;AAEP,AAAO,IAAMC,cAAc,WAApB;AACP,AAAO,IAAMC,cAAc,WAApB;;ACnBQ,SAASC,cAAT,CAAwB3N,GAAxB,EAA6B;MACpC4N,UAAU5N,IAAIuL,KAAJ,CAAUgC,eAAV,CAAhB;MACI,CAACK,OAAL,EAAc,OAAO,IAAP;;MAERC,UAAUtE,SAASqE,QAAQ,CAAR,CAAT,EAAqB,EAArB,CAAhB;;;;SAIOC,UAAU,GAAV,GAAgBA,OAAhB,GAA0B,IAAjC;;;ACVa,SAASC,YAAT,CAAsB9N,GAAtB,EAA2B;SACjCA,IAAI4E,KAAJ,CAAU,GAAV,EAAe,CAAf,EAAkBZ,OAAlB,CAA0B,KAA1B,EAAiC,EAAjC,CAAP;;;ACOF,SAAS+J,aAAT,CAAuBC,OAAvB,EAAgCrM,KAAhC,EAAuCsM,sBAAvC,EAA+D;MACzDC,cAAc,IAAlB;;;;MAIIvM,QAAQ,CAAR,IAAa+L,YAAY3N,IAAZ,CAAiBiO,OAAjB,CAAb,IAA0CA,QAAQjL,MAAR,GAAiB,CAA/D,EAAkE;kBAClD,IAAd;;;;;MAKEpB,UAAU,CAAV,IAAeqM,QAAQG,WAAR,OAA0B,OAA7C,EAAsD;kBACtC,KAAd;;;;;MAKExM,QAAQ,CAAR,IAAaqM,QAAQjL,MAAR,GAAiB,CAA9B,IAAmC,CAACkL,sBAAxC,EAAgE;kBAChD,KAAd;;;SAGKC,WAAP;;;;;;AAMF,AAAe,SAASE,cAAT,CAAwBpO,GAAxB,EAA6BqO,MAA7B,EAAqC;MAC5CpO,YAAYoO,UAAUnO,IAAIC,KAAJ,CAAUH,GAAV,CAA5B;MACQsO,QAF0C,GAEjBrO,SAFiB,CAE1CqO,QAF0C;MAEhCC,IAFgC,GAEjBtO,SAFiB,CAEhCsO,IAFgC;MAE1BC,IAF0B,GAEjBvO,SAFiB,CAE1BuO,IAF0B;;;MAI9CP,yBAAyB,KAA7B;MACMQ,kBAAkBD,KAAK5J,KAAL,CAAW,GAAX,EACvB8J,OADuB,GAEvBtE,MAFuB,CAEhB,UAACC,GAAD,EAAMsE,UAAN,EAAkBhN,KAAlB,EAA4B;QAC9BqM,UAAUW,UAAd;;;QAGIX,QAAQtL,QAAR,CAAiB,GAAjB,CAAJ,EAA2B;2BACUsL,QAAQpJ,KAAR,CAAc,GAAd,CADV;;;;UAClBgK,eADkB;UACDC,OADC;;UAErBpB,YAAY1N,IAAZ,CAAiB8O,OAAjB,CAAJ,EAA+B;kBACnBD,eAAV;;;;;;QAMArB,gBAAgBxN,IAAhB,CAAqBiO,OAArB,KAAiCrM,QAAQ,CAA7C,EAAgD;gBACpCqM,QAAQhK,OAAR,CAAgBuJ,eAAhB,EAAiC,EAAjC,CAAV;;;;;;;QAOE5L,UAAU,CAAd,EAAiB;+BACU6L,aAAazN,IAAb,CAAkBiO,OAAlB,CAAzB;;;;QAIED,cAAcC,OAAd,EAAuBrM,KAAvB,EAA8BsM,sBAA9B,CAAJ,EAA2D;UACrDa,IAAJ,CAASd,OAAT;;;WAGK3D,GAAP;GAhCsB,EAiCrB,EAjCqB,CAAxB;;SAmCUiE,QAAV,UAAuBC,IAAvB,GAA8BE,gBAAgBC,OAAhB,GAA0B7P,IAA1B,CAA+B,GAA/B,CAA9B;;;AC3EF;;AAEA,IAAMkQ,kBAAkB,IAAInQ,MAAJ,CAAW,QAAX,CAAxB;AACA,AAAe,SAASoQ,cAAT,CAAwBrE,IAAxB,EAA8B;SACpCoE,gBAAgBhP,IAAhB,CAAqB4K,IAArB,CAAP;;;ACJa,SAASsE,cAAT,CAAwBxM,OAAxB,EAA6C;kBAAZyM,KAAY,uEAAJ,EAAI;;qBACnDzM,QAAQmI,IAAR,GACQhG,KADR,CACc,KADd,EAEQqH,KAFR,CAEc,CAFd,EAEiBiD,KAFjB,EAGQrQ,IAHR,CAGa,GAHb,CAAP;;;ACQF;;;;;AAKA,AAAe,SAASsQ,aAAT,CAAuBC,UAAvB,EAAmCC,QAAnC,EAA6C7O,CAA7C,EAAgD;MACzD,CAAC4O,WAAW5E,MAAX,GAAoBzH,MAAzB,EAAiC;WACxBqM,UAAP;;;MAGIE,wBAAwBzD,KAAKE,GAAL,CAAS,EAAT,EAAasD,WAAW,IAAxB,CAA9B;MACME,cAAc/O,EAAE,aAAF,CAApB;;aAEWgK,MAAX,GAAoB1H,QAApB,GAA+BnC,IAA/B,CAAoC,UAACgB,KAAD,EAAQ2G,OAAR,EAAoB;QAChDkH,WAAWhP,EAAE8H,OAAF,CAAjB;;QAEI7B,4BAA0B1G,IAA1B,CAA+BuI,QAAQpF,OAAvC,CAAJ,EAAqD;aAC5C,IAAP;;;QAGIuM,eAAerE,SAASoE,QAAT,CAArB;QACIC,YAAJ,EAAkB;UACZD,aAAaJ,UAAjB,EAA6B;oBACfzL,MAAZ,CAAmB6L,QAAnB;OADF,MAEO;YACDE,eAAe,CAAnB;YACMC,UAAUC,YAAYJ,QAAZ,CAAhB;;;;YAIIG,UAAU,IAAd,EAAoB;0BACF,EAAhB;;;;;YAKEA,WAAW,GAAf,EAAoB;0BACF,EAAhB;;;;;YAKEH,SAASxO,IAAT,CAAc,OAAd,MAA2BoO,WAAWpO,IAAX,CAAgB,OAAhB,CAA/B,EAAyD;0BACvCqO,WAAW,GAA3B;;;YAGIQ,WAAWJ,eAAeC,YAAhC;;YAEIG,YAAYP,qBAAhB,EAAuC;iBAC9BC,YAAY5L,MAAZ,CAAmB6L,QAAnB,CAAP;SADF,MAEO,IAAIlH,QAAQpF,OAAR,KAAoB,GAAxB,EAA6B;cAC5B4M,iBAAiBN,SAAS7E,IAAT,EAAvB;cACMoF,uBAAuBrE,WAAWoE,cAAX,CAA7B;;cAEIC,uBAAuB,EAAvB,IAA6BJ,UAAU,IAA3C,EAAiD;mBACxCJ,YAAY5L,MAAZ,CAAmB6L,QAAnB,CAAP;WADF,MAEO,IAAIO,wBAAwB,EAAxB,IAA8BJ,YAAY,CAA1C,IACDX,eAAec,cAAf,CADH,EACmC;mBACjCP,YAAY5L,MAAZ,CAAmB6L,QAAnB,CAAP;;;;;;WAMD,IAAP;GAnDF;;SAsDOD,WAAP;;;ACxEF;;AAEA,AAAe,SAASS,gBAAT,CAA0BxP,CAA1B,EAA6B;MACtC4O,mBAAJ;MACIC,WAAW,CAAf;;IAEE,SAAF,EAAa1O,IAAb,CAAkB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;;QAE7B4F,4BAA0B1G,IAA1B,CAA+Bc,KAAKqC,OAApC,CAAJ,EAAkD;;;;QAI5CpC,QAAQN,EAAEK,IAAF,CAAd;QACMsK,QAAQC,SAAStK,KAAT,CAAd;;QAEIqK,QAAQkE,QAAZ,EAAsB;iBACTlE,KAAX;mBACarK,KAAb;;GAXJ;;;;MAiBI,CAACsO,UAAL,EAAiB;WACR5O,EAAE,MAAF,KAAaA,EAAE,GAAF,EAAOyP,KAAP,EAApB;;;eAGWd,cAAcC,UAAd,EAA0BC,QAA1B,EAAoC7O,CAApC,CAAb;;SAEO4O,UAAP;;;ACtBF,SAASc,mBAAT,CAA6BpP,KAA7B,EAAoCN,CAApC,EAAuC2P,MAAvC,EAA+C;;;;;MAKzCrP,MAAMsP,QAAN,CAAe,qBAAf,CAAJ,EAA2C;;;;MAIrC3N,UAAUwK,gBAAgBnM,MAAM6J,IAAN,EAAhB,CAAhB;;MAEIW,YAAY7I,OAAZ,IAAuB,EAA3B,EAA+B;QACvB4N,SAAS7P,EAAE,GAAF,EAAOM,KAAP,EAAciC,MAA7B;QACMuN,aAAa9P,EAAE,OAAF,EAAWM,KAAX,EAAkBiC,MAArC;;;QAGIuN,aAAcD,SAAS,CAA3B,EAA+B;YACvBnO,MAAN;;;;QAIIpC,gBAAgB2C,QAAQM,MAA9B;QACMwN,WAAW/P,EAAE,KAAF,EAASM,KAAT,EAAgBiC,MAAjC;;;;QAIIjD,gBAAgB,EAAhB,IAAsByQ,aAAa,CAAvC,EAA0C;YAClCrO,MAAN;;;;QAIIyN,UAAUC,YAAY9O,KAAZ,CAAhB;;;;;QAKIqP,SAAS,EAAT,IAAeR,UAAU,GAAzB,IAAgC7P,gBAAgB,EAApD,EAAwD;YAChDoC,MAAN;;;;;;QAMEiO,UAAU,EAAV,IAAgBR,UAAU,GAA9B,EAAmC;;;;UAI3BzM,UAAUpC,MAAM/B,GAAN,CAAU,CAAV,EAAamE,OAA7B;UACMsN,aAAatN,YAAY,IAAZ,IAAoBA,YAAY,IAAnD;UACIsN,UAAJ,EAAgB;YACRC,eAAe3P,MAAM4P,IAAN,EAArB;YACID,gBAAgBxD,gBAAgBwD,aAAa9F,IAAb,EAAhB,EAAqCsB,KAArC,CAA2C,CAAC,CAA5C,MAAmD,GAAvE,EAA4E;;;;;YAKxE/J,MAAN;;;;QAIIyO,cAAcnQ,EAAE,QAAF,EAAYM,KAAZ,EAAmBiC,MAAvC;;;QAGI4N,cAAc,CAAd,IAAmB7Q,gBAAgB,GAAvC,EAA4C;YACpCoC,MAAN;;;;;;;;;;;;;AAaN,AAAe,SAAS0O,SAAT,CAAmBlH,QAAnB,EAA6BlJ,CAA7B,EAAgC;IAC3C0F,wBAAF,EAA4BwD,QAA5B,EAAsC/I,IAAtC,CAA2C,UAACgB,KAAD,EAAQd,IAAR,EAAiB;QACpDC,QAAQN,EAAEK,IAAF,CAAd;QACIsP,SAAS/E,SAAStK,KAAT,CAAb;QACI,CAACqP,MAAL,EAAa;eACF9D,eAAevL,KAAf,EAAsBN,CAAtB,CAAT;eACSM,KAAT,EAAgBN,CAAhB,EAAmB2P,MAAnB;;;;QAIEA,SAAS,CAAb,EAAgB;YACRjO,MAAN;KADF,MAEO;;0BAEepB,KAApB,EAA2BN,CAA3B,EAA8B2P,MAA9B;;GAbJ;;SAiBO3P,CAAP;;;ACrGa,SAASqQ,YAAT,CAAsBnH,QAAtB,EAAgClJ,CAAhC,EAA+C;MAAZsQ,KAAY,uEAAJ,EAAI;;IAC1D1K,eAAF,EAAmBsD,QAAnB,EAA6B/I,IAA7B,CAAkC,UAACgB,KAAD,EAAQoP,MAAR,EAAmB;QAC7CC,UAAUxQ,EAAEuQ,MAAF,CAAhB;;;;;QAKIvQ,EAAEwQ,OAAF,EAAWtH,QAAX,EAAqBuH,OAArB,CAA6B,GAA7B,EAAkClO,MAAlC,KAA6C,CAAjD,EAAoD;aAC3CiO,QAAQ9O,MAAR,EAAP;;;;QAIE+K,gBAAgBzM,EAAEuQ,MAAF,EAAUpG,IAAV,EAAhB,MAAsCmG,KAA1C,EAAiD;aACxCE,QAAQ9O,MAAR,EAAP;;;;;QAKEgJ,UAAU1K,EAAEuQ,MAAF,CAAV,IAAuB,CAA3B,EAA8B;aACrBC,QAAQ9O,MAAR,EAAP;;;WAGK8O,OAAP;GArBF;;SAwBOxQ,CAAP;;;AC5BF;;;AAEA,AAAe,SAAS0Q,eAAT,CAAyBtH,OAAzB,EAAkCpJ,CAAlC,EAAqC;;;;MAI9C0I,cAAc1I,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;MACI0I,cAAc1I,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;;SAEOA,CAAP;;;ACTF,SAAS2Q,UAAT,CAAoB3Q,CAApB,EAAuB4Q,OAAvB,EAAgCpQ,IAAhC,EAAsCqQ,QAAtC,EAAgD;UACxCrQ,IAAN,QAAeqQ,QAAf,EAAyB1Q,IAAzB,CAA8B,UAACC,CAAD,EAAIC,IAAJ,EAAa;QACnCb,MAAMa,KAAKW,OAAL,CAAaR,IAAb,CAAZ;QACMsQ,cAAcpR,IAAIjB,OAAJ,CAAYmS,OAAZ,EAAqBpR,GAArB,CAApB;;SAEKwB,OAAL,CAAaR,IAAb,IAAqBsQ,WAArB;GAJF;;;AAQF,AAAe,SAASC,iBAAT,CAA2BF,QAA3B,EAAqC7Q,CAArC,EAAwCR,GAAxC,EAA6C;GACzD,MAAD,EAAS,KAAT,EAAgByB,OAAhB,CAAwB;WAAQ0P,WAAW3Q,CAAX,EAAcR,GAAd,EAAmBgB,IAAnB,EAAyBqQ,QAAzB,CAAR;GAAxB;;SAEOA,QAAP;;;ACbK,SAAS3F,UAAT,CAAoBf,IAApB,EAA0B;SACxBA,KAAKC,IAAL,GACK5G,OADL,CACa,MADb,EACqB,GADrB,EAEKjB,MAFZ;;;;;;AAQF,AAAO,SAAS6M,WAAT,CAAqB9O,KAArB,EAA4B;MAC3B0Q,kBAAkB9F,WAAW5K,MAAM6J,IAAN,EAAX,CAAxB;;MAEM8G,WAAW3Q,MAAMiB,IAAN,CAAW,GAAX,EAAgB4I,IAAhB,EAAjB;MACM+G,aAAahG,WAAW+F,QAAX,CAAnB;;MAEID,kBAAkB,CAAtB,EAAyB;WAChBE,aAAaF,eAApB;GADF,MAEO,IAAIA,oBAAoB,CAApB,IAAyBE,aAAa,CAA1C,EAA6C;WAC3C,CAAP;;;SAGK,CAAP;;;ACpBF;;;AAEA,AAAe,SAASC,eAAT,CACbnR,CADa,EAEboR,SAFa,EAGbC,WAHa,EAKb;MADAjB,SACA,uEADY,IACZ;;MACMkB,aAAaF,UAAU3P,MAAV,CAAiB;WAAQ4P,YAAYE,OAAZ,CAAoBC,IAApB,MAA8B,CAAC,CAAvC;GAAjB,CAAnB;;;;;;;;UAEWA,IAHX;;UAIQpQ,OAAO,MAAb;UACMb,QAAQ,OAAd;;UAEMkR,QAAQzR,YAAUoB,IAAV,UAAmBoQ,IAAnB,QAAd;;;;;UAKME,SACJD,MAAMvM,GAAN,CAAU,UAAC/D,KAAD,EAAQd,IAAR;eAAiBL,EAAEK,IAAF,EAAQG,IAAR,CAAaD,KAAb,CAAjB;OAAV,EACMoR,OADN,GAEMlQ,MAFN,CAEa;eAAQ0I,SAAS,EAAjB;OAFb,CADF;;;;;;UASIuH,OAAOnP,MAAP,KAAkB,CAAtB,EAAyB;YACnBqP,kBAAJ;;;YAGIxB,SAAJ,EAAe;sBACDyB,UAAUH,OAAO,CAAP,CAAV,EAAqB1R,CAArB,CAAZ;SADF,MAEO;sBACO0R,OAAO,CAAP,CAAZ;;;;aAGKE;;;;;sCA5BQN,UAAnB,4GAA+B;;;;;;;;;;;;;;;;;;;;;;SAiCxB,IAAP;;;AC3CF,SAASQ,UAAT,CAAoBxR,KAApB,EAA2ByR,WAA3B,EAAwC;;;MAGlCzR,MAAMgC,QAAN,GAAiBC,MAAjB,GAA0BwP,WAA9B,EAA2C;WAClC,KAAP;;;MAGEC,cAAc1R,KAAd,CAAJ,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;;;;AAMF,AAAe,SAAS2R,oBAAT,CACbjS,CADa,EAEbkS,SAFa,EAKb;MAFAH,WAEA,uEAFc,CAEd;MADAI,QACA,uEADW,IACX;;;;;;sCACuBD,SAAvB,4GAAkC;UAAvB/M,QAAuB;;UAC1BsM,QAAQzR,EAAEmF,QAAF,CAAd;;;;UAIIsM,MAAMlP,MAAN,KAAiB,CAArB,EAAwB;YAChBjC,QAAQN,EAAEyR,MAAM,CAAN,CAAF,CAAd;;YAEIK,WAAWxR,KAAX,EAAkByR,WAAlB,CAAJ,EAAoC;cAC9B9P,gBAAJ;cACIkQ,QAAJ,EAAc;sBACF7R,MAAM6J,IAAN,EAAV;WADF,MAEO;sBACK7J,MAAM8R,IAAN,EAAV;;;cAGEnQ,OAAJ,EAAa;mBACJA,OAAP;;;;;;;;;;;;;;;;;;;;SAMD,IAAP;;;AChDF;AACA,AAAe,SAAS4P,SAAT,CAAmB1H,IAAnB,EAAyBnK,CAAzB,EAA4B;;;MAGnCqS,YAAYrS,aAAWmK,IAAX,cAA0BA,IAA1B,EAAlB;SACOkI,cAAc,EAAd,GAAmBlI,IAAnB,GAA0BkI,SAAjC;;;ACLa,SAASL,aAAT,CAAuB1R,KAAvB,EAA8B;MACrCwC,UAAUxC,MAAMwC,OAAN,GAAgB6O,OAAhB,EAAhB;MACMW,gBAAgBxP,QAAQvB,IAAR,CAAa,UAACyI,MAAD,EAAY;QACvC1C,aAAgB0C,OAAOhJ,OAAP,CAAeuR,KAA/B,SAAwCvI,OAAOhJ,OAAP,CAAeqG,EAA7D;WACOC,WAAWpF,QAAX,CAAoB,SAApB,CAAP;GAFoB,CAAtB;;SAKOoQ,kBAAkBE,SAAzB;;;ACPF;;;;AAIA,AAAe,SAASC,gBAAT,CAA0BnS,KAA1B,EAAiC;SACvCA,MAAM6J,IAAN,GAAaC,IAAb,GAAoB7H,MAApB,IAA8B,GAArC;;;ACHa,SAASmQ,WAAT,CAAqB1S,CAArB,EAAwB;SAC9BA,EAAEwG,cAAF,EAAkBjE,MAAlB,GAA2B,CAAlC;;;ACHF;AACA,AAAO,IAAMoQ,kBAAkB,wCAAxB;;;;AAIP,AAAO,IAAMC,eAAe,IAAIxU,MAAJ,CAAW,aAAX,EAA0B,GAA1B,CAArB;AACP,AAYA,AASA;AACA,AAAO,IAAMyU,iBAAiB,WAAvB;AACP,AAAO,IAAMC,kBAAkB,WAAxB;AACP,AAAO,IAAMC,uBAAuB,4BAA7B;AACP,AAAO,IAAMC,yBAAyB,oBAA/B;AACP,AAAO,IAAMC,wBAAwB,QAA9B;AACP,IAAMC,SAAS,CACb,KADa,EAEb,KAFa,EAGb,KAHa,EAIb,KAJa,EAKb,KALa,EAMb,KANa,EAOb,KAPa,EAQb,KARa,EASb,KATa,EAUb,KAVa,EAWb,KAXa,EAYb,KAZa,CAAf;AAcA,IAAMC,YAAYD,OAAO7U,IAAP,CAAY,GAAZ,CAAlB;AACA,IAAM+U,aAAa,qCAAnB;AACA,IAAMC,aAAa,wCAAnB;AACA,AAAO,IAAMC,oBACX,IAAIlV,MAAJ,OAAegV,UAAf,WAA+BC,UAA/B,wBAA4DF,SAA5D,QAA0E,IAA1E,CADK;;;;;AAMP,AAAO,IAAMI,qBAAqB,gBAA3B;;AAEP,AAAO,IAAMC,oBACX,IAAIpV,MAAJ,CAAW,2BAAX,EAAwC,GAAxC,CADK;;ACxDP;;AAEA,AAAe,SAASqV,WAAT,CAAqBC,MAArB,EAA6B;SACnCA,OAAOlQ,OAAP,CAAemP,eAAf,EAAgC,IAAhC,EAAsCvI,IAAtC,EAAP;;;ACHa,SAASzI,OAAT,CAAegS,YAAf,EAA6B;iBAC3BA,aAAavJ,IAAb,EAAf;MACIwJ,SAASC,QAAT,CAAkBF,YAAlB,CAAJ,EAAqC;WAC5BA,YAAP;;;SAGK,IAAP;;;ACHF;;AAEA,AAAe,SAASG,QAAT,CAAkBC,GAAlB,QAAuC;MAAd/T,CAAc,QAAdA,CAAc;MAAXgU,OAAW,QAAXA,OAAW;;;MAEhDD,IAAIxR,MAAJ,GAAa,IAAb,IAAqBwR,IAAIxR,MAAJ,GAAa,CAAtC,EAAyC,OAAO,IAAP;;;MAGrCyR,WAAWvF,eAAeuF,OAAf,EAAwB,EAAxB,MAAgCvF,eAAesF,GAAf,EAAoB,EAApB,CAA/C,EAAwE,OAAO,IAAP;;MAElEE,UAAUpC,UAAUkC,GAAV,EAAe/T,CAAf,CAAhB;;;;MAII4S,aAAarT,IAAb,CAAkB0U,OAAlB,CAAJ,EAAgC,OAAO,IAAP;;SAEzBA,QAAQ7J,IAAR,EAAP;;;ACnBF;;;;AAIA,AASA,AAAO,SAAS8J,eAAT,CAAyBC,UAAzB,EAAqC;SACnC,CAACA,WAAWpJ,KAAX,CAAiBuI,iBAAjB,KAAuC,EAAxC,EACWjV,IADX,CACgB,GADhB,EAEWmF,OAFX,CAEmByP,qBAFnB,EAE0C,GAF1C,EAGWzP,OAHX,CAGmBwP,sBAHnB,EAG2C,UAH3C,EAIWxP,OAJX,CAImBuP,oBAJnB,EAIyC,IAJzC,EAKW3I,IALX,EAAP;;;;;AAUF,AAAe,SAASgK,kBAAT,CAA4BD,UAA5B,EAAwC;;MAEjDtB,eAAetT,IAAf,CAAoB4U,UAApB,KAAmCrB,gBAAgBvT,IAAhB,CAAqB4U,UAArB,CAAvC,EAAyE;iBAC1DpL,SAASoL,UAAT,EAAqB,EAArB,CAAb;;;MAGEE,OAAOC,OAAO,IAAIC,IAAJ,CAASJ,UAAT,CAAP,CAAX;;MAEI,CAACE,KAAKG,OAAL,EAAL,EAAqB;iBACNN,gBAAgBC,UAAhB,CAAb;WACOG,OAAO,IAAIC,IAAJ,CAASJ,UAAT,CAAP,CAAP;;;SAGKE,KAAKG,OAAL,KAAiBH,KAAKI,WAAL,EAAjB,GAAsC,IAA7C;;;ACzBF;;AACA,AAAe,SAASC,gBAAT,CACbtL,OADa,QASb;MANEpJ,CAMF,QANEA,CAMF;mCALE2U,kBAKF;MALEA,kBAKF,yCALuB,IAKvB;wBAJErE,KAIF;MAJEA,KAIF,8BAJU,EAIV;sBAHE9Q,GAGF;MAHEA,GAGF,4BAHQ,EAGR;iCAFEoV,cAEF;MAFEA,cAEF,uCAFmB,IAEnB;;;;kBAGgBxL,OAAhB,EAAyBpJ,CAAzB;;;;;MAKI4U,cAAJ,EAAoB3L,YAAYG,OAAZ,EAAqBpJ,CAArB;;;;;aAKToJ,OAAX,EAAoBpJ,CAApB;;;;gBAIcoJ,OAAd,EAAuBpJ,CAAvB;;;;;aAKWoJ,OAAX,EAAoBpJ,CAApB;;;eAGaoJ,OAAb,EAAsBpJ,CAAtB,EAAyBsQ,KAAzB;;;oBAGkBlH,OAAlB,EAA2BpJ,CAA3B,EAA8BR,GAA9B;;;;;;MAMIoV,cAAJ,EAAoBxE,UAAUhH,OAAV,EAAmBpJ,CAAnB,EAAsB2U,kBAAtB;;;cAGRvL,OAAZ,EAAqBpJ,CAArB;;;kBAGgBoJ,OAAhB,EAAyBpJ,CAAzB;;SAEOoJ,OAAP;;;AC5Da,SAASyL,UAAT,CAAoBvE,KAApB,QAAuC;MAAV9Q,GAAU,QAAVA,GAAU;MAALQ,CAAK,QAALA,CAAK;;;;MAGhDuT,mBAAmBhU,IAAnB,CAAwB+Q,KAAxB,CAAJ,EAAoC;YAC1BwE,kBAAkBxE,KAAlB,EAAyB9Q,GAAzB,CAAR;;;;;MAKE8Q,MAAM/N,MAAN,GAAe,GAAnB,EAAwB;;QAEhBwS,KAAK/U,EAAE,IAAF,CAAX;QACI+U,GAAGxS,MAAH,KAAc,CAAlB,EAAqB;cACXwS,GAAG5K,IAAH,EAAR;;;;;SAKG0H,UAAUvB,KAAV,EAAiBtQ,CAAjB,EAAoBoK,IAApB,EAAP;;;ACfF,SAAS4K,sBAAT,CAAgCC,UAAhC,EAA4C9K,IAA5C,EAAkD;;;;MAI5C8K,WAAW1S,MAAX,IAAqB,CAAzB,EAA4B;;;;;UAIpB2S,aAAaD,WAAWrL,MAAX,CAAkB,UAACC,GAAD,EAAMsL,SAAN,EAAoB;YACnDA,SAAJ,IAAiBtL,IAAIsL,SAAJ,IAAiBtL,IAAIsL,SAAJ,IAAiB,CAAlC,GAAsC,CAAvD;eACOtL,GAAP;OAFiB,EAGhB,EAHgB,CAAnB;;kCAME,iBAAgBqL,UAAhB,EACQtL,MADR,CACe,UAACC,GAAD,EAAMjB,GAAN,EAAc;YAChBiB,IAAI,CAAJ,IAASqL,WAAWtM,GAAX,CAAb,EAA8B;iBACrB,CAACA,GAAD,EAAMsM,WAAWtM,GAAX,CAAN,CAAP;;;eAGKiB,GAAP;OANT,EAOU,CAAC,CAAD,EAAI,CAAJ,CAPV,CAVwB;;;;UASnBuL,OATmB;UASVC,SATU;;;;;;;UAuBtBA,aAAa,CAAb,IAAkBD,QAAQ7S,MAAR,IAAkB,CAAxC,EAA2C;qBAC5B4H,KAAK/F,KAAL,CAAWgR,OAAX,CAAb;;;UAGIE,YAAY,CAACL,WAAW,CAAX,CAAD,EAAgBA,WAAWxJ,KAAX,CAAiB,CAAC,CAAlB,CAAhB,CAAlB;UACM8J,aAAaD,UAAU1L,MAAV,CAAiB,UAACC,GAAD,EAAMjM,GAAN;eAAciM,IAAItH,MAAJ,GAAa3E,IAAI2E,MAAjB,GAA0BsH,GAA1B,GAAgCjM,GAA9C;OAAjB,EAAoE,EAApE,CAAnB;;UAEI2X,WAAWhT,MAAX,GAAoB,EAAxB,EAA4B;;aACnBgT;;;;;WAGFpL;;;;;;;SAGF,IAAP;;;AAGF,SAASqL,oBAAT,CAA8BP,UAA9B,EAA0CzV,GAA1C,EAA+C;;;;;;;mBAO5BE,IAAIC,KAAJ,CAAUH,GAAV,CAP4B;;MAOrCuO,IAPqC,cAOrCA,IAPqC;;MAQvC0H,cAAc1H,KAAKvK,OAAL,CAAagQ,iBAAb,EAAgC,EAAhC,CAApB;;MAEMkC,YAAYT,WAAW,CAAX,EAActH,WAAd,GAA4BnK,OAA5B,CAAoC,GAApC,EAAyC,EAAzC,CAAlB;MACMmS,iBAAiBC,MAAMC,WAAN,CAAkBH,SAAlB,EAA6BD,WAA7B,CAAvB;;MAEIE,iBAAiB,GAAjB,IAAwBD,UAAUnT,MAAV,GAAmB,CAA/C,EAAkD;WACzC0S,WAAWxJ,KAAX,CAAiB,CAAjB,EAAoBpN,IAApB,CAAyB,EAAzB,CAAP;;;MAGIyX,UAAUb,WAAWxJ,KAAX,CAAiB,CAAC,CAAlB,EAAqB,CAArB,EAAwBkC,WAAxB,GAAsCnK,OAAtC,CAA8C,GAA9C,EAAmD,EAAnD,CAAhB;MACMuS,eAAeH,MAAMC,WAAN,CAAkBC,OAAlB,EAA2BL,WAA3B,CAArB;;MAEIM,eAAe,GAAf,IAAsBD,QAAQvT,MAAR,IAAkB,CAA5C,EAA+C;WACtC0S,WAAWxJ,KAAX,CAAiB,CAAjB,EAAoB,CAAC,CAArB,EAAwBpN,IAAxB,CAA6B,EAA7B,CAAP;;;SAGK,IAAP;;;;;AAKF,AAAe,SAASyW,iBAAT,CAA2BxE,KAA3B,EAA4C;MAAV9Q,GAAU,uEAAJ,EAAI;;;;MAGnDyV,aAAa3E,MAAMlM,KAAN,CAAYmP,kBAAZ,CAAnB;MACI0B,WAAW1S,MAAX,KAAsB,CAA1B,EAA6B;WACpB+N,KAAP;;;MAGE0F,WAAWhB,uBAAuBC,UAAvB,EAAmC3E,KAAnC,CAAf;MACI0F,QAAJ,EAAc,OAAOA,QAAP;;aAEHR,qBAAqBP,UAArB,EAAiCzV,GAAjC,CAAX;MACIwW,QAAJ,EAAc,OAAOA,QAAP;;;;SAIP1F,KAAP;;;AC1FF,IAAM2F,WAAW;UACPxC,WADO;kBAECyC,OAFD;OAGVpC,QAHU;kBAICM,kBAJD;WAKN+B,gBALM;SAMRtB;CANT,CAUA,AAEA,AACA,AACA,AACA,AACA,AACA,AACA;;ACfA;;;;;;;;;;;AAWA,AAAe,SAASuB,eAAT,CAAyBpW,CAAzB,EAA4BqW,IAA5B,EAAkC;;;;;;;MAO3CA,KAAKnP,uBAAT,EAAkC;QAC5BA,wBAAwBlH,CAAxB,CAAJ;;;MAGEyI,oBAAoBzI,CAApB,CAAJ;MACIqM,aAAarM,CAAb,EAAgBqW,KAAKrK,WAArB,CAAJ;MACMsK,gBAAgB9G,iBAAiBxP,CAAjB,CAAtB;;SAEOsW,aAAP;;;AC5BF,IAAMC,0BAA0B;eACjB;6BACc,IADd;iBAEE,IAFF;wBAGS;GAJQ;;;;;;;;;;;;;;;;;;;;;SAAA,yBA0BGF,IA1BH,EA0BS;QAA7BrW,CAA6B,QAA7BA,CAA6B;QAA1BoS,IAA0B,QAA1BA,IAA0B;QAApB9B,KAAoB,QAApBA,KAAoB;QAAb9Q,GAAa,QAAbA,GAAa;;wBACzB,KAAKgX,WAAjB,EAAiCH,IAAjC;;QAEIrW,KAAKmC,QAAQC,IAAR,CAAagQ,IAAb,CAAT;;;;QAII/R,OAAO,KAAKoW,cAAL,CAAoBzW,CAApB,EAAuBsQ,KAAvB,EAA8B9Q,GAA9B,EAAmC6W,IAAnC,CAAX;;QAEI5D,iBAAiBpS,IAAjB,CAAJ,EAA4B;aACnB,KAAKqW,kBAAL,CAAwBrW,IAAxB,EAA8BL,CAA9B,CAAP;;;;;;;;;;wCAKgB,iBAAgBqW,IAAhB,EAAsB5U,MAAtB,CAA6B;eAAK4U,KAAKM,CAAL,MAAY,IAAjB;OAA7B,CAAlB,4GAAuE;YAA5D/N,GAA4D;;aAChEA,GAAL,IAAY,KAAZ;YACIzG,QAAQC,IAAR,CAAagQ,IAAb,CAAJ;;eAEO,KAAKqE,cAAL,CAAoBzW,CAApB,EAAuBsQ,KAAvB,EAA8B9Q,GAA9B,EAAmC6W,IAAnC,CAAP;;YAEI5D,iBAAiBpS,IAAjB,CAAJ,EAA4B;;;;;;;;;;;;;;;;;;;WAKvB,KAAKqW,kBAAL,CAAwBrW,IAAxB,EAA8BL,CAA9B,CAAP;GApD4B;;;;gBAAA,0BAwDfA,CAxDe,EAwDZsQ,KAxDY,EAwDL9Q,GAxDK,EAwDA6W,IAxDA,EAwDM;WAC3BF,iBACGC,gBAAgBpW,CAAhB,EAAmBqW,IAAnB,CADH,EAEL;UAAA;0BAEsBA,KAAK1B,kBAF3B;kBAAA;;KAFK,CAAP;GAzD4B;;;;;;oBAAA,8BAsEXtU,IAtEW,EAsELL,CAtEK,EAsEF;QACtB,CAACK,IAAL,EAAW;aACF,IAAP;;;WAGKoM,gBAAgBzM,EAAEoS,IAAF,CAAO/R,IAAP,CAAhB,CAAP;;;;;;;CA3EJ,CAqFA;;AC7FA;;;;;;;AAOA,AAAO,IAAMuW,yBAAyB,CACpC,iBADoC,EAEpC,UAFoC,EAGpC,SAHoC,EAIpC,UAJoC,EAKpC,OALoC,CAA/B;;;;AAUP,AAAO,IAAMC,uBAAuB,CAClC,UADkC,CAA7B;;;;;;;;;AAWP,AAAO,IAAMC,yBAAyB,CACpC,sBADoC,EAEpC,kBAFoC,EAGpC,kBAHoC,EAIpC,YAJoC,EAKpC,mBALoC,EAMpC,cANoC,CAA/B;;AASP,AAAO,IAAMC,uBAAuB,CAClC,YADkC,EAElC,cAFkC,EAGlC,cAHkC,EAIlC,aAJkC,EAKlC,aALkC,EAMlC,aANkC,EAOlC,aAPkC,EAQlC,eARkC,EASlC,eATkC,EAUlC,iBAVkC,EAWlC,UAXkC,EAYlC,YAZkC,EAalC,IAbkC,EAclC,iBAdkC,EAelC,OAfkC,CAA7B;;ACxBP,IAAMC,wBAAwB;SAAA,yBACG;QAArBhX,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbyX,SAAa,QAAbA,SAAa;;;;QAGzB3G,cAAJ;;YAEQa,gBAAgBnR,CAAhB,EAAmB4W,sBAAnB,EAA2CK,SAA3C,CAAR;QACI3G,KAAJ,EAAW,OAAOuE,WAAWvE,KAAX,EAAkB,EAAE9Q,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;;YAIHiS,qBAAqBjS,CAArB,EAAwB8W,sBAAxB,CAAR;QACIxG,KAAJ,EAAW,OAAOuE,WAAWvE,KAAX,EAAkB,EAAE9Q,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;YAGHmR,gBAAgBnR,CAAhB,EAAmB6W,oBAAnB,EAAyCI,SAAzC,CAAR;QACI3G,KAAJ,EAAW,OAAOuE,WAAWvE,KAAX,EAAkB,EAAE9Q,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;YAGHiS,qBAAqBjS,CAArB,EAAwB+W,oBAAxB,CAAR;QACIzG,KAAJ,EAAW,OAAOuE,WAAWvE,KAAX,EAAkB,EAAE9Q,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;WAGJ,EAAP;;CAvBJ,CA2BA;;ACxCA;;;;;;AAMA,AAAO,IAAMkX,mBAAmB,CAC9B,KAD8B,EAE9B,OAF8B,EAG9B,WAH8B,EAI9B,eAJ8B,EAK9B,YAL8B,EAM9B,WAN8B,EAO9B,SAP8B,CAAzB;;AAUP,AAAO,IAAMC,oBAAoB,GAA1B;;;;;;;;;AASP,AAAO,IAAMC,mBAAmB,CAC9B,sBAD8B,EAE9B,mBAF8B,EAG9B,oBAH8B,EAI9B,mBAJ8B,EAK9B,oBAL8B,EAM9B,qBAN8B,EAO9B,aAP8B,EAQ9B,iBAR8B,EAS9B,oBAT8B,EAU9B,qBAV8B,EAW9B,eAX8B,EAY9B,YAZ8B,EAa9B,YAb8B,EAc9B,cAd8B,EAe9B,cAf8B,EAgB9B,yBAhB8B,EAiB9B,qBAjB8B,EAkB9B,qBAlB8B,EAmB9B,SAnB8B,EAoB9B,SApB8B,EAqB9B,gBArB8B,EAsB9B,gBAtB8B,EAuB9B,SAvB8B,CAAzB;;;;AA4BP,IAAMC,WAAW,aAAjB;AACA,AAAO,IAAMC,sBAAsB,CACjC,CAAC,SAAD,EAAYD,QAAZ,CADiC,EAEjC,CAAC,SAAD,EAAYA,QAAZ,CAFiC,CAA5B;;ACzCP,IAAME,yBAAyB;SAAA,yBACH;QAAhBvX,CAAgB,QAAhBA,CAAgB;QAAbiX,SAAa,QAAbA,SAAa;;QACpBvD,eAAJ;;;;aAISvC,gBAAgBnR,CAAhB,EAAmBkX,gBAAnB,EAAqCD,SAArC,CAAT;QACIvD,UAAUA,OAAOnR,MAAP,GAAgB4U,iBAA9B,EAAiD;aACxC1D,YAAYC,MAAZ,CAAP;;;;aAIOzB,qBAAqBjS,CAArB,EAAwBoX,gBAAxB,EAA0C,CAA1C,CAAT;QACI1D,UAAUA,OAAOnR,MAAP,GAAgB4U,iBAA9B,EAAiD;aACxC1D,YAAYC,MAAZ,CAAP;;;;;;;;;;wCAK8B4D,mBAAhC,4GAAqD;;;;;YAAzCnS,QAAyC;YAA/BqS,KAA+B;;YAC7CnX,OAAOL,EAAEmF,QAAF,CAAb;YACI9E,KAAKkC,MAAL,KAAgB,CAApB,EAAuB;cACf4H,OAAO9J,KAAK8J,IAAL,EAAb;cACIqN,MAAMjY,IAAN,CAAW4K,IAAX,CAAJ,EAAsB;mBACbsJ,YAAYtJ,IAAZ,CAAP;;;;;;;;;;;;;;;;;;;WAKC,IAAP;;CA7BJ,CAiCA;;AC9CA;;;;AAIA,AAAO,IAAMsN,2BAA2B,CACtC,wBADsC,EAEtC,aAFsC,EAGtC,SAHsC,EAItC,gBAJsC,EAKtC,WALsC,EAMtC,cANsC,EAOtC,UAPsC,EAQtC,UARsC,EAStC,SATsC,EAUtC,eAVsC,EAWtC,UAXsC,EAYtC,cAZsC,EAatC,qBAbsC,EActC,cAdsC,EAetC,SAfsC,EAgBtC,MAhBsC,CAAjC;;;;;AAsBP,AAAO,IAAMC,2BAA2B,CACtC,4BADsC,EAEtC,oBAFsC,EAGtC,0BAHsC,EAItC,kBAJsC,EAKtC,oBALsC,EAMtC,kBANsC,EAOtC,iBAPsC,EAQtC,aARsC,EAStC,eATsC,EAUtC,qBAVsC,EAWtC,mBAXsC,EAYtC,cAZsC,EAatC,aAbsC,EActC,YAdsC,EAetC,kBAfsC,EAgBtC,WAhBsC,EAiBtC,UAjBsC,CAAjC;;;;;AAuBP,IAAMC,kBAAkB,mDAAxB;AACA,AAAO,IAAMC,yBAAyB;;AAEpC,IAAIxZ,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAFoC;;;;AAMpC,IAAIA,MAAJ,CAAW,6BAAX,EAA0C,GAA1C,CANoC;;AAQpC,IAAIA,MAAJ,iBAAyBuZ,eAAzB,kBAAuD,GAAvD,CARoC,CAA/B;;ACrCP,IAAME,gCAAgC;SAAA,yBACL;QAArB7X,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbyX,SAAa,QAAbA,SAAa;;QACzBa,sBAAJ;;;;oBAIgB3G,gBAAgBnR,CAAhB,EAAmByX,wBAAnB,EAA6CR,SAA7C,EAAwD,KAAxD,CAAhB;QACIa,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;;oBAIH7F,qBAAqBjS,CAArB,EAAwB0X,wBAAxB,CAAhB;QACII,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;oBAGHpL,eAAelN,GAAf,EAAoBoY,sBAApB,CAAhB;QACIE,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;WAEZ,IAAP;;CAlBJ,CAsBA;;ACnCA;;;;;;;;;;;;;;;;;AAiBA,IAAMC,sBAAsB;;SAAA,qBAEhB;WACD,IAAP;;CAHJ,CAOA;;ACxBA;;;AAGA,AAAO,IAAMC,2BAA2B,CACtC,UADsC,EAEtC,eAFsC,EAGtC,WAHsC,CAAjC;;AAMP,AAAO,IAAMC,2BAA2B,CACtC,qBADsC,CAAjC;;AAIP,AAAO,IAAMC,gCAAgC,CAC3C,QAD2C,EAE3C,YAF2C,EAG3C,OAH2C,EAI3C,OAJ2C,EAK3C,UAL2C,CAAtC;AAOP,AAAO,IAAMC,mCAAmC,IAAI/Z,MAAJ,CAAW8Z,8BAA8B7Z,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAM+Z,gCAAgC,CAC3C,QAD2C,EAE3C,QAF2C,EAG3C,OAH2C,EAI3C,UAJ2C,EAK3C,UAL2C,EAM3C,MAN2C,EAO3C,IAP2C,EAQ3C,YAR2C,EAS3C,MAT2C,EAU3C,QAV2C,EAW3C,QAX2C,EAY3C,KAZ2C,EAa3C,QAb2C,EAc3C,SAd2C,EAe3C,QAf2C,EAgB3C,SAhB2C,EAiB3C,SAjB2C,EAkB3C,QAlB2C,EAmB3C,OAnB2C,EAoB3C,UApB2C,EAqB3C,SArB2C,EAsB3C,OAtB2C,EAuB3C,OAvB2C,EAwB3C,KAxB2C,EAyB3C,aAzB2C,CAAtC;AA2BP,AAAO,IAAMC,mCAAmC,IAAIja,MAAJ,CAAWga,8BAA8B/Z,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAMia,SAAS,gBAAf;AACP,AAAO,IAAMC,SAAS,kBAAf;;AC3CP,SAASC,MAAT,CAAgBlY,KAAhB,EAAuB;UACXA,MAAME,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCF,MAAME,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;;AAIF,AAAO,SAASiY,aAAT,CAAuBjZ,GAAvB,EAA4B;QAC3BA,IAAI4K,IAAJ,EAAN;MACIO,QAAQ,CAAZ;;MAEIwN,iCAAiC5Y,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;MAGE6Y,iCAAiC9Y,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;;;MAKE8Y,OAAO/Y,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;MAGE+Y,OAAOhZ,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;;;SAKKmL,KAAP;;;;AAIF,AAAO,SAAS+N,SAAT,CAAmBhU,IAAnB,EAAyB;MAC1BA,KAAKlE,IAAL,CAAU,KAAV,CAAJ,EAAsB;WACb,CAAP;;;SAGK,CAAP;;;;;AAKF,AAAO,SAASmY,cAAT,CAAwBjU,IAAxB,EAA8B;MAC/BiG,QAAQ,CAAZ;MACMiO,aAAalU,KAAK5B,OAAL,CAAa,QAAb,EAAuB2M,KAAvB,EAAnB;;MAEImJ,WAAWrW,MAAX,KAAsB,CAA1B,EAA6B;aAClB,EAAT;;;MAGIM,UAAU6B,KAAKsF,MAAL,EAAhB;MACI6O,iBAAJ;MACIhW,QAAQN,MAAR,KAAmB,CAAvB,EAA0B;eACbM,QAAQmH,MAAR,EAAX;;;GAGDnH,OAAD,EAAUgW,QAAV,EAAoB5X,OAApB,CAA4B,UAACX,KAAD,EAAW;QACjC6F,iBAAe5G,IAAf,CAAoBiZ,OAAOlY,KAAP,CAApB,CAAJ,EAAwC;eAC7B,EAAT;;GAFJ;;SAMOqK,KAAP;;;;;AAKF,AAAO,SAASmO,cAAT,CAAwBpU,IAAxB,EAA8B;MAC/BiG,QAAQ,CAAZ;MACMqE,WAAWtK,KAAKiD,IAAL,EAAjB;MACMG,UAAUkH,SAASzQ,GAAT,CAAa,CAAb,CAAhB;;MAEIuJ,WAAWA,QAAQpF,OAAR,KAAoB,YAAnC,EAAiD;aACtC,EAAT;;;MAGEyD,iBAAe5G,IAAf,CAAoBiZ,OAAOxJ,QAAP,CAApB,CAAJ,EAA2C;aAChC,EAAT;;;SAGKrE,KAAP;;;AAGF,AAAO,SAASoO,iBAAT,CAA2BrU,IAA3B,EAAiC;MAClCiG,QAAQ,CAAZ;;MAEMpH,QAAQsH,WAAWnG,KAAKlE,IAAL,CAAU,OAAV,CAAX,CAAd;MACMsI,SAAS+B,WAAWnG,KAAKlE,IAAL,CAAU,QAAV,CAAX,CAAf;MACM8C,MAAMoB,KAAKlE,IAAL,CAAU,KAAV,CAAZ;;;MAGI+C,SAASA,SAAS,EAAtB,EAA0B;aACf,EAAT;;;;MAIEuF,UAAUA,UAAU,EAAxB,EAA4B;aACjB,EAAT;;;MAGEvF,SAASuF,MAAT,IAAmB,CAACxF,IAAIpB,QAAJ,CAAa,QAAb,CAAxB,EAAgD;QACxC8W,OAAOzV,QAAQuF,MAArB;QACIkQ,OAAO,IAAX,EAAiB;;eACN,GAAT;KADF,MAEO;eACI3N,KAAK4N,KAAL,CAAWD,OAAO,IAAlB,CAAT;;;;SAIGrO,KAAP;;;AAGF,AAAO,SAASuO,eAAT,CAAyBC,KAAzB,EAAgChY,KAAhC,EAAuC;SACpCgY,MAAM5W,MAAN,GAAe,CAAhB,GAAqBpB,KAA5B;;;AC1GF;;;;;;;;AAQA,IAAMiY,+BAA+B;SAAA,yBACA;QAAzBpZ,CAAyB,QAAzBA,CAAyB;QAAtBiC,OAAsB,QAAtBA,OAAsB;QAAbgV,SAAa,QAAbA,SAAa;;QAC7BoC,iBAAJ;;;;;;QAMMC,WACJnI,gBACEnR,CADF,EAEEgY,wBAFF,EAGEf,SAHF,EAIE,KAJF,CADF;;QAQIqC,QAAJ,EAAc;iBACDpD,QAAWoD,QAAX,CAAX;;UAEID,QAAJ,EAAc,OAAOA,QAAP;;;;;;QAMVE,OAAOvZ,EAAE,KAAF,EAASiC,OAAT,EAAkB0P,OAAlB,EAAb;QACM6H,YAAY,EAAlB;;SAEKvY,OAAL,CAAa,UAACF,GAAD,EAAMI,KAAN,EAAgB;UACrBuD,OAAO1E,EAAEe,GAAF,CAAb;UACMuC,MAAMoB,KAAKlE,IAAL,CAAU,KAAV,CAAZ;;UAEI,CAAC8C,GAAL,EAAU;;UAENqH,QAAQ8N,cAAcnV,GAAd,CAAZ;eACSoV,UAAUhU,IAAV,CAAT;eACSiU,eAAejU,IAAf,CAAT;eACSoU,eAAepU,IAAf,CAAT;eACSqU,kBAAkBrU,IAAlB,CAAT;eACSwU,gBAAgBK,IAAhB,EAAsBpY,KAAtB,CAAT;;gBAEUmC,GAAV,IAAiBqH,KAAjB;KAbF;;gCAiBE,iBAAgB6O,SAAhB,EAA2B5P,MAA3B,CAAkC,UAACC,GAAD,EAAMjB,GAAN;aAChC4Q,UAAU5Q,GAAV,IAAiBiB,IAAI,CAAJ,CAAjB,GAA0B,CAACjB,GAAD,EAAM4Q,UAAU5Q,GAAV,CAAN,CAA1B,GAAkDiB,GADlB;KAAlC,EAEE,CAAC,IAAD,EAAO,CAAP,CAFF,CA5C+B;;;;QA2C1B4P,MA3C0B;QA2ClB5K,QA3CkB;;;QAgD7BA,WAAW,CAAf,EAAkB;iBACLqH,QAAWuD,MAAX,CAAX;;UAEIJ,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;wCAKOpB,wBAAvB,4GAAiD;YAAtC9S,QAAsC;;YACzC7E,QAAQN,EAAEmF,QAAF,EAAYsK,KAAZ,EAAd;YACMnM,MAAMhD,MAAME,IAAN,CAAW,KAAX,CAAZ;YACI8C,GAAJ,EAAS;qBACI4S,QAAW5S,GAAX,CAAX;cACI+V,QAAJ,EAAc,OAAOA,QAAP;;;YAGVK,OAAOpZ,MAAME,IAAN,CAAW,MAAX,CAAb;YACIkZ,IAAJ,EAAU;qBACGxD,QAAWwD,IAAX,CAAX;cACIL,QAAJ,EAAc,OAAOA,QAAP;;;YAGV9Y,QAAQD,MAAME,IAAN,CAAW,OAAX,CAAd;YACID,KAAJ,EAAW;qBACE2V,QAAW3V,KAAX,CAAX;cACI8Y,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;;;;;;;;;WAIX,IAAP;;CA9EJ,CAkFA;;ACzGe,SAASM,eAAT,CAAyBhP,KAAzB,EAAgCiP,UAAhC,EAA4CF,IAA5C,EAAkD;;;;;;MAM3D/O,QAAQ,CAAZ,EAAe;QACPkP,aAAa,IAAIC,QAAQC,eAAZ,CAA4B,IAA5B,EAAkCH,UAAlC,EAA8CF,IAA9C,EAAoDM,KAApD,EAAnB;;;;;;;QAOMC,cAAc,MAAMJ,UAA1B;QACMK,eAAe,EAAE,OAAOD,cAAc,GAArB,CAAF,CAArB;WACOtP,QAAQuP,YAAf;;;SAGK,CAAP;;;ACnBa,SAASC,aAAT,CAAuBlJ,QAAvB,EAAiC5D,OAAjC,EAA0C;;;;;MAKnD1C,QAAQ,CAAZ;;MAEIuC,YAAY3N,IAAZ,CAAiB0R,SAAS7G,IAAT,EAAjB,CAAJ,EAAuC;QAC/BgQ,gBAAgBrR,SAASkI,QAAT,EAAmB,EAAnB,CAAtB;;;;QAIImJ,gBAAgB,CAApB,EAAuB;cACb,CAAC,EAAT;KADF,MAEO;cACG/O,KAAKE,GAAL,CAAS,CAAT,EAAY,KAAK6O,aAAjB,CAAR;;;;;;QAME/M,WAAWA,WAAW+M,aAA1B,EAAyC;eAC9B,EAAT;;;;SAIGzP,KAAP;;;AC5Ba,SAAS0P,eAAT,CAAyBhN,OAAzB,EAAkCiN,IAAlC,EAAwC;;;;MAIjDjN,WAAW,CAACiN,IAAhB,EAAsB;WACb,EAAP;;;SAGK,CAAP;;;ACRK,IAAMC,aAAW,IAAjB;;;;AAIP,AAAO,IAAM9T,0BAAwB,CACnC,OADmC,EAEnC,SAFmC,EAGnC,SAHmC,EAInC,SAJmC,EAKnC,QALmC,EAMnC,OANmC,EAOnC,OAPmC,EAQnC,OARmC,EASnC,KATmC,EAUnC,OAVmC,EAWnC,MAXmC,EAYnC,QAZmC,EAanC,KAbmC,EAcnC,iBAdmC,CAA9B;AAgBP,AAAO,IAAMC,6BAA2B,IAAItI,MAAJ,CAAWqI,wBAAsBpI,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;;;AAKP,AAAO,IAAMmc,sBAAoB,IAAIpc,MAAJ,CAAW,4CAAX,EAAyD,GAAzD,CAA1B;;;;AAIP,AAAO,IAAMqc,qBAAmB,IAAIrc,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAzB;;;;AAIP,AAAO,IAAMsc,sBAAoB,IAAItc,MAAJ,CAAW,yBAAX,EAAsC,GAAtC,CAA1B,CAEP;;ACjCe,SAASuc,oBAAT,CAA8BjB,IAA9B,EAAoC;;MAE7ChT,2BAAyBnH,IAAzB,CAA8Bma,IAA9B,CAAJ,EAAyC;WAChC,CAAC,EAAR;;;SAGK,CAAP;;;ACAF,SAASkB,SAAT,CAAiBC,KAAjB,EAAwB;UACZA,MAAMra,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCqa,MAAMra,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;AAGF,AAAe,SAASmY,gBAAT,CAAwBkC,KAAxB,EAA+B;;;;MAIxChY,UAAUgY,MAAM7Q,MAAN,EAAd;MACI8Q,gBAAgB,KAApB;MACIC,gBAAgB,KAApB;MACIpQ,QAAQ,CAAZ;;cAEWjN,MAAM,CAAN,EAAS,CAAT,CAAX,EAAwBuD,OAAxB,CAAgC,YAAM;QAChC4B,QAAQN,MAAR,KAAmB,CAAvB,EAA0B;;;;QAIpByY,aAAaJ,UAAQ/X,OAAR,EAAiB,GAAjB,CAAnB;;;;QAII,CAACiY,aAAD,IAAkBnU,QAAQpH,IAAR,CAAayb,UAAb,CAAtB,EAAgD;sBAC9B,IAAhB;eACS,EAAT;;;;;;QAME,CAACD,aAAD,IAAkBxU,kBAAkBhH,IAAlB,CAAuByb,UAAvB,CAAlB,IACEtU,2BAAyBnH,IAAzB,CAA8Byb,UAA9B,CADN,EACiD;UAC3C,CAAC3U,kBAAkB9G,IAAlB,CAAuByb,UAAvB,CAAL,EAAyC;wBACvB,IAAhB;iBACS,EAAT;;;;cAIMnY,QAAQmH,MAAR,EAAV;GAzBF;;SA4BOW,KAAP;;;AC/Ca,SAASsQ,aAAT,CAAuBC,QAAvB,EAAiC;;;MAG1CR,oBAAkBnb,IAAlB,CAAuB2b,QAAvB,CAAJ,EAAsC;WAC7B,CAAC,GAAR;;;SAGK,CAAP;;;ACFa,SAASC,WAAT,CACbzB,IADa,EAEbE,UAFa,EAGbwB,OAHa,EAIb3b,SAJa,EAKbwR,QALa,EAMboK,YANa,EAOb;;MAEIA,aAAa9Z,IAAb,CAAkB;WAAOmY,SAASla,GAAhB;GAAlB,MAA2CgT,SAA/C,EAA0D;WACjD,KAAP;;;;;MAKE,CAACkH,IAAD,IAASA,SAASE,UAAlB,IAAgCF,SAAS0B,OAA7C,EAAsD;WAC7C,KAAP;;;MAGMtd,QAZR,GAYqB2B,SAZrB,CAYQ3B,QAZR;;mBAa+B4B,IAAIC,KAAJ,CAAU+Z,IAAV,CAb/B;;MAakB4B,QAblB,cAaQxd,QAbR;;;;MAgBIwd,aAAaxd,QAAjB,EAA2B;WAClB,KAAP;;;;;MAKIyd,WAAW7B,KAAKlW,OAAL,CAAa4X,OAAb,EAAsB,EAAtB,CAAjB;MACI,CAACb,WAAShb,IAAT,CAAcgc,QAAd,CAAL,EAA8B;WACrB,KAAP;;;;;MAKE7U,2BAAyBnH,IAAzB,CAA8B0R,QAA9B,CAAJ,EAA6C;WACpC,KAAP;;;;MAIEA,SAAS1O,MAAT,GAAkB,EAAtB,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;ACpDa,SAASiZ,YAAT,CAAsB9B,IAAtB,EAA4B+B,SAA5B,EAAuC;;;;;MAKhD,CAACA,UAAUlc,IAAV,CAAema,IAAf,CAAL,EAA2B;WAClB,CAAC,EAAR;;;SAGK,CAAP;;;ACPa,SAASgC,iBAAT,CAA2BR,QAA3B,EAAqC;;MAE9CV,oBAAkBjb,IAAlB,CAAuB2b,QAAvB,CAAJ,EAAsC;WAC7B,EAAP;;;SAGK,CAAP;;;ACHa,SAASS,aAAT,CAAuBT,QAAvB,EAAiC;;MAE1CT,mBAAiBlb,IAAjB,CAAsB2b,QAAtB,CAAJ,EAAqC;;;;;QAK/BV,oBAAkBjb,IAAlB,CAAuB2b,QAAvB,CAAJ,EAAsC;aAC7B,CAAC,EAAR;;;;SAIG,CAAP;;;ACIK,SAASU,aAAT,CAAuBR,OAAvB,EAAgC;SAC9B,IAAIhd,MAAJ,OAAegd,OAAf,EAA0B,GAA1B,CAAP;;;AAGF,SAASR,OAAT,CAAiBC,KAAjB,EAAwB5J,QAAxB,EAAkC;UACtBA,YAAY4J,MAAM1Q,IAAN,EAAtB,WAAsC0Q,MAAMra,IAAN,CAAW,OAAX,KAAuB,EAA7D,WAAmEqa,MAAMra,IAAN,CAAW,IAAX,KAAoB,EAAvF;;;AAGF,AAAe,SAASqb,UAAT,OAOZ;MANDC,KAMC,QANDA,KAMC;MALDlC,UAKC,QALDA,UAKC;MAJDwB,OAIC,QAJDA,OAIC;MAHD3b,SAGC,QAHDA,SAGC;MAFDO,CAEC,QAFDA,CAEC;+BADDqb,YACC;MADDA,YACC,qCADc,EACd;;cACW5b,aAAaC,IAAIC,KAAJ,CAAUia,UAAV,CAAzB;MACM6B,YAAYG,cAAcR,OAAd,CAAlB;MACMd,OAAO5H,YAAY1S,CAAZ,CAAb;;;;;;;;;MASM+b,cAAcD,MAAMlS,MAAN,CAAa,UAACoS,aAAD,EAAgBC,IAAhB,EAAyB;;;;QAIlDvC,OAAOpM,aAAa2O,KAAKjb,OAAL,CAAa0Y,IAA1B,CAAb;QACMmB,QAAQ7a,EAAEic,IAAF,CAAd;QACMhL,WAAW4J,MAAM1Q,IAAN,EAAjB;;QAEI,CAACgR,YAAYzB,IAAZ,EAAkBE,UAAlB,EAA8BwB,OAA9B,EAAuC3b,SAAvC,EAAkDwR,QAAlD,EAA4DoK,YAA5D,CAAL,EAAgF;aACvEW,aAAP;;;;QAIE,CAACA,cAActC,IAAd,CAAL,EAA0B;oBACVA,IAAd,IAAsB;eACb,CADa;0BAAA;;OAAtB;KADF,MAMO;oBACSA,IAAd,EAAoBzI,QAApB,GAAkC+K,cAActC,IAAd,EAAoBzI,QAAtD,SAAkEA,QAAlE;;;QAGIiL,eAAeF,cAActC,IAAd,CAArB;QACMwB,WAAWN,QAAQC,KAAR,EAAe5J,QAAf,CAAjB;QACM5D,UAAUF,eAAeuM,IAAf,CAAhB;;QAEI/O,QAAQ6Q,aAAa9B,IAAb,EAAmB+B,SAAnB,CAAZ;aACSC,kBAAkBR,QAAlB,CAAT;aACSS,cAAcT,QAAd,CAAT;aACSD,cAAcC,QAAd,CAAT;aACSvC,iBAAekC,KAAf,CAAT;aACSF,qBAAqBjB,IAArB,CAAT;aACSW,gBAAgBhN,OAAhB,EAAyBiN,IAAzB,CAAT;aACSH,cAAclJ,QAAd,EAAwB5D,OAAxB,CAAT;aACSsM,gBAAgBhP,KAAhB,EAAuBiP,UAAvB,EAAmCF,IAAnC,CAAT;;iBAEa/O,KAAb,GAAqBA,KAArB;;WAEOqR,aAAP;GAvCkB,EAwCjB,EAxCiB,CAApB;;SA0CO,iBAAgBD,WAAhB,EAA6BxZ,MAA7B,KAAwC,CAAxC,GAA4C,IAA5C,GAAmDwZ,WAA1D;;;AClFF;;AAEA,IAAMI,8BAA8B;SAAA,yBACgB;QAAxCnc,CAAwC,QAAxCA,CAAwC;QAArCR,GAAqC,QAArCA,GAAqC;QAAhCC,SAAgC,QAAhCA,SAAgC;iCAArB4b,YAAqB;QAArBA,YAAqB,qCAAN,EAAM;;gBACpC5b,aAAaC,IAAIC,KAAJ,CAAUH,GAAV,CAAzB;;QAEMoa,aAAatM,aAAa9N,GAAb,CAAnB;QACM4b,UAAUxN,eAAepO,GAAf,EAAoBC,SAApB,CAAhB;;QAEMqc,QAAQ9b,EAAE,SAAF,EAAa2R,OAAb,EAAd;;QAEMyK,cAAcP,WAAW;kBAAA;4BAAA;sBAAA;0BAAA;UAAA;;KAAX,CAApB;;;QAUI,CAACO,WAAL,EAAkB,OAAO,IAAP;;;;QAIZC,UAAU,iBAAgBD,WAAhB,EAA6BxS,MAA7B,CAAoC,UAACC,GAAD,EAAMoS,IAAN,EAAe;UAC3DK,aAAaF,YAAYH,IAAZ,CAAnB;aACOK,WAAW3R,KAAX,GAAmBd,IAAIc,KAAvB,GAA+B2R,UAA/B,GAA4CzS,GAAnD;KAFc,EAGb,EAAEc,OAAO,CAAC,GAAV,EAHa,CAAhB;;;;QAOI0R,QAAQ1R,KAAR,IAAiB,EAArB,EAAyB;aAChB0R,QAAQ3C,IAAf;;;WAGK,IAAP;;CAlCJ,CAuCA;;ACjDO,IAAM6C,2BAA2B,CACtC,QADsC,CAAjC;;ACKP,SAASC,WAAT,CAAqBhd,GAArB,EAA0B;MAClBC,YAAYC,IAAIC,KAAJ,CAAUH,GAAV,CAAlB;MACQ1B,QAFgB,GAEH2B,SAFG,CAEhB3B,QAFgB;;SAGjBA,QAAP;;;AAGF,SAASiE,MAAT,CAAgBvC,GAAhB,EAAqB;SACZ;YAAA;YAEGgd,YAAYhd,GAAZ;GAFV;;;AAMF,IAAMid,sBAAsB;SAAA,yBACK;QAArBzc,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbyX,SAAa,QAAbA,SAAa;;QACvByF,aAAa1c,EAAE,qBAAF,CAAnB;QACI0c,WAAWna,MAAX,KAAsB,CAA1B,EAA6B;UACrBmX,OAAOgD,WAAWlc,IAAX,CAAgB,MAAhB,CAAb;UACIkZ,IAAJ,EAAU;eACD3X,OAAO2X,IAAP,CAAP;;;;QAIEiD,UAAUxL,gBAAgBnR,CAAhB,EAAmBuc,wBAAnB,EAA6CtF,SAA7C,CAAhB;QACI0F,OAAJ,EAAa;aACJ5a,OAAO4a,OAAP,CAAP;;;WAGK5a,OAAOvC,GAAP,CAAP;;CAfJ,CAoBA;;ACtCO,IAAMod,yBAAyB,CACpC,gBADoC,EAEpC,qBAFoC,CAA/B;;ACSA,SAASjb,OAAT,CAAeM,OAAf,EAAwBjC,CAAxB,EAA4C;MAAjB6c,SAAiB,uEAAL,GAAK;;YACvC5a,QAAQuB,OAAR,CAAgB,UAAhB,EAA4B,GAA5B,EAAiC4G,IAAjC,EAAV;SACO0S,UAAU7a,OAAV,EAAmB4a,SAAnB,EAA8B,EAAEE,SAAS,UAAX,EAA9B,CAAP;;;AAGF,IAAMC,0BAA0B;SAAA,yBACK;QAAzBhd,CAAyB,QAAzBA,CAAyB;QAAtBiC,OAAsB,QAAtBA,OAAsB;QAAbgV,SAAa,QAAbA,SAAa;;QAC3BjD,UAAU7C,gBAAgBnR,CAAhB,EAAmB4c,sBAAnB,EAA2C3F,SAA3C,CAAhB;QACIjD,OAAJ,EAAa;aACJrS,QAAMkQ,UAAUmC,OAAV,EAAmBhU,CAAnB,CAAN,CAAP;;;QAGI6c,YAAY,GAAlB;QACMI,eAAehb,QAAQwJ,KAAR,CAAc,CAAd,EAAiBoR,YAAY,CAA7B,CAArB;WACOlb,QAAM3B,EAAEid,YAAF,EAAgB9S,IAAhB,EAAN,EAA8BnK,CAA9B,EAAiC6c,SAAjC,CAAP;;CATJ,CAaA;;ACvBA,IAAMK,4BAA4B;SAAA,yBACX;QAAXjb,OAAW,QAAXA,OAAW;;QACbjC,IAAImC,QAAQC,IAAR,CAAaH,OAAb,CAAV;;QAEMkI,OAAOsC,gBAAgBzM,EAAE,KAAF,EAASyP,KAAT,GAAiBtF,IAAjB,EAAhB,CAAb;WACOA,KAAK/F,KAAL,CAAW,IAAX,EAAiB7B,MAAxB;;CALJ,CASA;;ACCA,IAAM4a,mBAAmB;;UAEf,GAFe;SAGhBnG,sBAAsBoG,OAHN;kBAIPvF,8BAA8BuF,OAJvB;UAKf7F,uBAAuB6F,OALR;WAMd7G,wBAAwB6G,OAAxB,CAAgCC,IAAhC,CAAqC9G,uBAArC,CANc;kBAOP6C,6BAA6BgE,OAPtB;OAQlBrF,oBAAoBqF,OARF;iBASRjB,4BAA4BiB,OATpB;kBAUPX,oBAAoBW,OAVb;WAWdJ,wBAAwBI,OAXV;cAYXF,0BAA0BE,OAZf;aAaZ;QAAG9M,KAAH,QAAGA,KAAH;WAAegN,gBAAgBC,YAAhB,CAA6BjN,KAA7B,CAAf;GAbY;;SAAA,mBAef9R,OAfe,EAeN;QACP4T,IADO,GACE5T,OADF,CACP4T,IADO;;;QAGXA,IAAJ,EAAU;UACFpS,IAAImC,QAAQC,IAAR,CAAagQ,IAAb,CAAV;cACQpS,CAAR,GAAYA,CAAZ;;;QAGIsQ,QAAQ,KAAKA,KAAL,CAAW9R,OAAX,CAAd;QACMgf,iBAAiB,KAAKA,cAAL,CAAoBhf,OAApB,CAAvB;QACMkV,SAAS,KAAKA,MAAL,CAAYlV,OAAZ,CAAf;QACMyD,UAAU,KAAKA,OAAL,cAAkBzD,OAAlB,IAA2B8R,YAA3B,IAAhB;QACMmN,iBAAiB,KAAKA,cAAL,cAAyBjf,OAAzB,IAAkCyD,gBAAlC,IAAvB;QACM8R,MAAM,KAAKA,GAAL,cAAcvV,OAAd,IAAuByD,gBAAvB,IAAZ;QACMyb,gBAAgB,KAAKA,aAAL,CAAmBlf,OAAnB,CAAtB;QACMwV,UAAU,KAAKA,OAAL,cAAkBxV,OAAlB,IAA2ByD,gBAA3B,IAAhB;QACM0b,aAAa,KAAKA,UAAL,cAAqBnf,OAArB,IAA8ByD,gBAA9B,IAAnB;QACM2b,YAAY,KAAKA,SAAL,CAAe,EAAEtN,YAAF,EAAf,CAAlB;;0BACwB,KAAKuN,cAAL,CAAoBrf,OAApB,CAlBT;;QAkBPgB,GAlBO,mBAkBPA,GAlBO;QAkBFse,MAlBE,mBAkBFA,MAlBE;;;WAoBR;kBAAA;oBAAA;sBAGWN,kBAAkB,IAH7B;cAAA;oCAAA;sBAAA;kCAAA;cAAA;oBAAA;sBAAA;4BAAA;;KAAP;;CAnCJ,CAoDA;;AC7De,SAASO,YAAT,CAAsBve,GAAtB,EAA2BC,SAA3B,EAAsC;cACvCA,aAAaC,IAAIC,KAAJ,CAAUH,GAAV,CAAzB;mBACqBC,SAF8B;MAE3C3B,QAF2C,cAE3CA,QAF2C;;MAG7CkgB,aAAalgB,SAASsG,KAAT,CAAe,GAAf,EAAoBqH,KAApB,CAA0B,CAAC,CAA3B,EAA8BpN,IAA9B,CAAmC,GAAnC,CAAnB;;SAEOsG,WAAW7G,QAAX,KAAwB6G,WAAWqZ,UAAX,CAAxB,IAAkDb,gBAAzD;;;ACNF;AACA,AAAO,SAASc,gBAAT,CAA0BpN,QAA1B,EAAoC7Q,CAApC,QAAkD;MAAT2B,KAAS,QAATA,KAAS;;MACnD,CAACA,KAAL,EAAY,OAAOkP,QAAP;;IAEVlP,MAAMtD,IAAN,CAAW,GAAX,CAAF,EAAmBwS,QAAnB,EAA6BnP,MAA7B;;SAEOmP,QAAP;;;;AAIF,AAAO,SAASqN,iBAAT,CAA2BrN,QAA3B,EAAqC7Q,CAArC,SAAwD;MAAdme,UAAc,SAAdA,UAAc;;MACzD,CAACA,UAAL,EAAiB,OAAOtN,QAAP;;mBAEDsN,UAAhB,EAA4Bld,OAA5B,CAAoC,UAAC2H,GAAD,EAAS;QACrCwV,WAAWpe,EAAE4I,GAAF,EAAOiI,QAAP,CAAjB;QACMtQ,QAAQ4d,WAAWvV,GAAX,CAAd;;;QAGI,OAAOrI,KAAP,KAAiB,QAArB,EAA+B;eACpBJ,IAAT,CAAc,UAACgB,KAAD,EAAQd,IAAR,EAAiB;sBACfL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0Bme,WAAWvV,GAAX,CAA1B;OADF;KADF,MAIO,IAAI,OAAOrI,KAAP,KAAiB,UAArB,EAAiC;;eAE7BJ,IAAT,CAAc,UAACgB,KAAD,EAAQd,IAAR,EAAiB;YACvB0B,SAASxB,MAAMP,EAAEK,IAAF,CAAN,EAAeL,CAAf,CAAf;;YAEI,OAAO+B,MAAP,KAAkB,QAAtB,EAAgC;wBAChB/B,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0B+B,MAA1B;;OAJJ;;GAXJ;;SAqBO8O,QAAP;;;AAGF,SAASwN,oBAAT,CAA8Bre,CAA9B,EAAiCkS,SAAjC,EAA4C;SACnCA,UAAU3Q,IAAV,CAAe,UAAC4D,QAAD,EAAc;QAC9BmZ,MAAMC,OAAN,CAAcpZ,QAAd,CAAJ,EAA6B;qCACTA,QADS;;UACpBqZ,CADoB;UACjBhe,IADiB;;aAEpBR,EAAEwe,CAAF,EAAKjc,MAAL,KAAgB,CAAhB,IAAqBvC,EAAEwe,CAAF,EAAKhe,IAAL,CAAUA,IAAV,CAArB,IAAwCR,EAAEwe,CAAF,EAAKhe,IAAL,CAAUA,IAAV,EAAgB4J,IAAhB,OAA2B,EAA1E;;;WAGKpK,EAAEmF,QAAF,EAAY5C,MAAZ,KAAuB,CAAvB,IAA4BvC,EAAEmF,QAAF,EAAYgF,IAAZ,GAAmBC,IAAnB,OAA8B,EAAjE;GANK,CAAP;;;AAUF,AAAO,SAASqU,MAAT,CAAgBpI,IAAhB,EAAsB;MACnBrW,CADmB,GAC8BqW,IAD9B,CACnBrW,CADmB;MAChBoB,IADgB,GAC8BiV,IAD9B,CAChBjV,IADgB;MACVsd,cADU,GAC8BrI,IAD9B,CACVqI,cADU;0BAC8BrI,IAD9B,CACMsI,WADN;MACMA,WADN,qCACoB,KADpB;;;MAGvB,CAACD,cAAL,EAAqB,OAAO,IAAP;;;;MAIjB,OAAOA,cAAP,KAA0B,QAA9B,EAAwC,OAAOA,cAAP;;MAEhCxM,SATmB,GASkBwM,cATlB,CASnBxM,SATmB;8BASkBwM,cATlB,CASR9J,cATQ;MASRA,cATQ,yCASS,IATT;;;MAWrBgK,mBAAmBP,qBAAqBre,CAArB,EAAwBkS,SAAxB,CAAzB;;MAEI,CAAC0M,gBAAL,EAAuB,OAAO,IAAP;;;;;;;;MAQnBD,WAAJ,EAAiB;QACX9N,WAAW7Q,EAAE4e,gBAAF,CAAf;;;aAGSC,IAAT,CAAc7e,EAAE,aAAF,CAAd;eACW6Q,SAAS7G,MAAT,EAAX;;eAEWkU,kBAAkBrN,QAAlB,EAA4B7Q,CAA5B,EAA+B0e,cAA/B,CAAX;eACWT,iBAAiBpN,QAAjB,EAA2B7Q,CAA3B,EAA8B0e,cAA9B,CAAX;;eAEWzI,SAAS7U,IAAT,EAAeyP,QAAf,eAA8BwF,IAA9B,IAAoCzB,8BAApC,IAAX;;WAEO5U,EAAEoS,IAAF,CAAOvB,QAAP,CAAP;;;MAGE9O,eAAJ;;;;MAIIuc,MAAMC,OAAN,CAAcK,gBAAd,CAAJ,EAAqC;2CACVA,gBADU;;QAC5BzZ,QAD4B;QAClB3E,IADkB;;aAE1BR,EAAEmF,QAAF,EAAY3E,IAAZ,CAAiBA,IAAjB,EAAuB4J,IAAvB,EAAT;GAFF,MAGO;aACIpK,EAAE4e,gBAAF,EAAoBzU,IAApB,GAA2BC,IAA3B,EAAT;;;;;MAKEwK,cAAJ,EAAoB;WACXqB,SAAS7U,IAAT,EAAeW,MAAf,EAAuBsU,IAAvB,CAAP;;;SAGKtU,MAAP;;;AAGF,SAAS+c,aAAT,CAAuBzI,IAAvB,EAA6B;MACnBjV,IADmB,GACkBiV,IADlB,CACnBjV,IADmB;MACb2d,SADa,GACkB1I,IADlB,CACb0I,SADa;uBACkB1I,IADlB,CACF2I,QADE;MACFA,QADE,kCACS,IADT;;;MAGrBjd,SAAS0c,oBAAYpI,IAAZ,IAAkBqI,gBAAgBK,UAAU3d,IAAV,CAAlC,IAAf;;;MAGIW,MAAJ,EAAY;WACHA,MAAP;;;;;MAKEid,QAAJ,EAAc,OAAO7B,iBAAiB/b,IAAjB,EAAuBiV,IAAvB,CAAP;;SAEP,IAAP;;;AAGF,IAAM4I,gBAAgB;SAAA,qBACwB;QAApCF,SAAoC,uEAAxB5B,gBAAwB;QAAN9G,IAAM;gBACFA,IADE;QAClC6I,WADkC,SAClCA,WADkC;QACrBC,cADqB,SACrBA,cADqB;;;QAGtCJ,UAAUjB,MAAV,KAAqB,GAAzB,EAA8B,OAAOiB,UAAU3B,OAAV,CAAkB/G,IAAlB,CAAP;;wBAGzBA,IADL;;;;QAKI6I,WAAJ,EAAiB;UACTjd,WAAU6c,2BACXzI,IADW,IACLjV,MAAM,SADD,EACYud,aAAa,IADzB,EAC+BrO,OAAO6O;SADtD;aAGO;;OAAP;;QAII7O,QAAQwO,2BAAmBzI,IAAnB,IAAyBjV,MAAM,OAA/B,IAAd;QACMoc,iBAAiBsB,2BAAmBzI,IAAnB,IAAyBjV,MAAM,gBAA/B,IAAvB;QACMsS,SAASoL,2BAAmBzI,IAAnB,IAAyBjV,MAAM,QAA/B,IAAf;QACMsc,gBAAgBoB,2BAAmBzI,IAAnB,IAAyBjV,MAAM,eAA/B,IAAtB;QACMa,UAAU6c,2BACXzI,IADW,IACLjV,MAAM,SADD,EACYud,aAAa,IADzB,EAC+BrO;OAD/C;QAGMmN,iBAAiBqB,2BAAmBzI,IAAnB,IAAyBjV,MAAM,gBAA/B,EAAiDa,gBAAjD,IAAvB;QACM+R,UAAU8K,2BAAmBzI,IAAnB,IAAyBjV,MAAM,SAA/B,EAA0Ca,gBAA1C,IAAhB;QACM8R,MAAM+K,2BAAmBzI,IAAnB,IAAyBjV,MAAM,KAA/B,EAAsCa,gBAAtC,EAA+C+R,gBAA/C,IAAZ;QACM2J,aAAamB,2BAAmBzI,IAAnB,IAAyBjV,MAAM,YAA/B,EAA6Ca,gBAA7C,IAAnB;QACM2b,YAAYkB,2BAAmBzI,IAAnB,IAAyBjV,MAAM,WAA/B,EAA4CkP,YAA5C,IAAlB;;gBAEEwO,2BAAmBzI,IAAnB,IAAyBjV,MAAM,gBAA/B,QAAsD,EAAE5B,KAAK,IAAP,EAAase,QAAQ,IAArB,EA/Bd;;QA8BlCte,GA9BkC,SA8BlCA,GA9BkC;QA8B7Bse,MA9B6B,SA8B7BA,MA9B6B;;;WAiCnC;kBAAA;sBAAA;oBAAA;oCAAA;oCAAA;cAAA;kCAAA;cAAA;oBAAA;sBAAA;4BAAA;;KAAP;;CAlCJ,CAmDA;;AC3KA;wDAAe;QAEXJ,aAFW,SAEXA,aAFW;QAGXtL,IAHW,SAGXA,IAHW;QAIXpS,CAJW,SAIXA,CAJW;QAKXiX,SALW,SAKXA,SALW;QAMXlV,MANW,SAMXA,MANW;QAOXqd,SAPW,SAOXA,SAPW;QAQX9O,KARW,SAQXA,KARW;QASX9Q,GATW,SASXA,GATW;;;;;;;iBAAA,GAaD,CAbC;wBAAA,GAcQ,CAAC8N,aAAa9N,GAAb,CAAD,CAdR;;;;;;kBAkBNke,iBAAiB2B,QAAQ,EAlBnB;;;;;qBAmBF,CAAT;;mBACUzd,SAAS0d,MAAT,CAAgB5B,aAAhB,CApBC;;;aAAA;;mBAqBJ1d,EAAEoS,IAAF,EAAP;;yBArBW,GAuBW;mBACfsL,aADe;wBAAA;kBAAA;kCAAA;2BAKP,IALO;8BAMJpN,KANI;;aAvBX;0BAAA,GAiCY2O,cAAc7B,OAAd,CAAsBgC,SAAtB,EAAiCG,aAAjC,CAjCZ;;;yBAmCEjR,IAAb,CAAkBoP,aAAlB;kCAEK3b,MADL;sCAGMA,OAAOE,OADX,yCAGaod,KAHb,uBAIIG,eAAevd,OAJnB;;;4BAQcud,eAAe9B,aAA/B;;;;;sBA9CW,GAiDMP,iBAAiBQ,UAAjB,CAA4B,EAAE1b,mBAAiBF,OAAOE,OAAxB,WAAF,EAA5B,CAjDN;0DAmDRF,MAnDQ;2BAoDEsd,KApDF;8BAqDKA,KArDL;;;;;;;;;;GAAf;;WAA8BI,eAA9B;;;;SAA8BA,eAA9B;;;ACMA,IAAMC,UAAU;OAAA,iBACFlgB,GADE,EACG4S,IADH,EACoB;;;QAAXiE,IAAW,uEAAJ,EAAI;;;;;;;;oCAI5BA,IAJ4B,CAE9BsJ,aAF8B;2BAAA,uCAEd,IAFc;+BAI5BtJ,IAJ4B,CAG9B2I,QAH8B;sBAAA,kCAGnB,IAHmB;uBAAA,GAMdtf,IAAIC,KAAJ,CAAUH,GAAV,CANc;;kBAQ3B3B,YAAY4B,SAAZ,CAR2B;;;;;+CASvB1B,OAAO8B,MATgB;;;uBAAA,GAYdke,aAAave,GAAb,EAAkBC,SAAlB,CAZc;;;;qBAehBmC,SAAS0d,MAAT,CAAgB9f,GAAhB,EAAqB4S,IAArB,EAA2B3S,SAA3B,CAfgB;;;eAAA;;mBAkB5BO,EAAEb,KAlB0B;;;;;+CAmBvBa,CAnBuB;;;;qBAsBzBA,EAAEoS,IAAF,EAAP;;;;uBAtBgC,GA0BdpS,EAAE,MAAF,EAAUkF,GAAV,CAAc,UAAC9E,CAAD,EAAIC,IAAJ;uBAAaL,EAAEK,IAAF,EAAQG,IAAR,CAAa,MAAb,CAAb;eAAd,EAAiDmR,OAAjD,EA1Bc;oBAAA,GA4BnBsN,cAAc7B,OAAd,CAAsBgC,SAAtB,EAAiC,EAAE5f,QAAF,EAAO4S,UAAP,EAAapS,IAAb,EAAgBiX,oBAAhB,EAA2BxX,oBAA3B,EAAsCuf,kBAAtC,EAAjC,CA5BmB;wBA6BCjd,MA7BD;mBAAA,WA6BxBuO,KA7BwB;2BAAA,WA6BjBoN,aA7BiB;;;;oBAgC5BiC,iBAAiBjC,aAhCW;;;;;;qBAiCf+B,gBACb;oCAAA;4CAAA;0BAAA;oBAAA;oCAAA;8BAAA;4BAAA;;eADa,CAjCe;;;oBAAA;;;;;oCA+CzB1d,MADL;6BAEe,CAFf;gCAGkB;;;;+CAIbA,MArDyB;;;;;;;;;GADpB;;;;;eAAA,yBA2DMvC,GA3DN,EA2DW;;;;;;;;;qBACVoC,SAAS0d,MAAT,CAAgB9f,GAAhB,CADU;;;;;;;;;;;;;CA3D3B,CAiEA;;"}
\ No newline at end of file
diff --git a/src/cleaners/content.js b/src/cleaners/content.js
index ca2423d0..021721a4 100644
--- a/src/cleaners/content.js
+++ b/src/cleaners/content.js
@@ -51,9 +51,6 @@ export default function extractCleanNode(
// Make links absolute
makeLinksAbsolute(article, $, url);
- // Remove unnecessary attributes
- cleanAttributes(article);
-
// We used to clean UL's and OL's here, but it was leading to
// too many in-article lists being removed. Consider a better
// way to detect menus particularly and remove them.
@@ -63,5 +60,8 @@ export default function extractCleanNode(
// Remove empty paragraph nodes
removeEmpty(article, $);
+ // Remove unnecessary attributes
+ cleanAttributes(article, $);
+
return article;
}
diff --git a/src/mercury.test.js b/src/mercury.test.js
index 9312fd90..6824e60c 100644
--- a/src/mercury.test.js
+++ b/src/mercury.test.js
@@ -18,7 +18,16 @@ describe('Mercury', () => {
});
it('does the whole thing', async function() {
- const result = await Mercury.parse('http://theconcourse.deadspin.com/phyllis-schlafly-finally-croaks-1786219220');
+ const result =
+ await Mercury.parse('http://deadspin.com/remember-when-donald-trump-got-booed-for-butchering-ta-1788216229');
+
+ assert.equal(typeof result, 'object');
+ assert.equal(result.content.indexOf('score="') === -1, true);
+ });
+
+ it('does blogger', async function() {
+ const result =
+ await Mercury.parse('https://googleblog.blogspot.com/2016/08/onhub-turns-one-today.html');
assert.equal(typeof result, 'object');
});
diff --git a/src/utils/dom/clean-attributes.js b/src/utils/dom/clean-attributes.js
index f8803c91..30b79542 100644
--- a/src/utils/dom/clean-attributes.js
+++ b/src/utils/dom/clean-attributes.js
@@ -1,7 +1,6 @@
import { WHITELIST_ATTRS_RE } from './constants';
function removeAllButWhitelist($article) {
- // $('*', article).each((index, node) => {
$article.find('*').each((index, node) => {
node.attribs = Reflect.ownKeys(node.attribs).reduce((acc, attr) => {
if (WHITELIST_ATTRS_RE.test(attr)) {
@@ -11,6 +10,8 @@ function removeAllButWhitelist($article) {
return acc;
}, {});
});
+
+ return $article;
}
// function removeAttrs(article, $) {
@@ -21,7 +22,11 @@ function removeAllButWhitelist($article) {
// Remove attributes like style or align
export default function cleanAttributes($article) {
- removeAllButWhitelist($article);
-
- return $article;
+ // Grabbing the parent because at this point
+ // $article will be wrapped in a div which will
+ // have a score set on it.
+ return removeAllButWhitelist(
+ $article.parent().length ?
+ $article.parent() : $article
+ );
}