diff --git a/dist/mercury.js b/dist/mercury.js index 8d83b50e..65bedada 100644 --- a/dist/mercury.js +++ b/dist/mercury.js @@ -670,6 +670,49 @@ var WiredExtractor = { excerpt: null }; +// Rename CustomExtractor +// to fit your publication +// (e.g., NYTimesExtractor) +var MSNExtractor = { + domain: 'www.msn.com', + title: { + selectors: ['h1'] + }, + + author: { + selectors: ['span.authorname-txt'] + }, + + content: { + selectors: ['div.richtext'], + + // Is there anything in the content you selected that needs transformed + // before it's consumable content? E.g., unusual lazy loaded images + transforms: [], + + // Is there anything that is in the result that shouldn't be? + // The clean selectors will remove anything that matches from + // the result + clean: ['span.caption'] + }, + + date_published: { + selectors: ['span.time'] + }, + + lead_image_url: { + selectors: [] + }, + + dek: { + selectors: [['meta[name="description"]', 'value']] + }, + + next_page_url: null, + + excerpt: null +}; + var Extractors = { 'nymag.com': NYMagExtractor, 'blogspot.com': BloggerExtractor, @@ -678,7 +721,8 @@ var Extractors = { 'www.nytimes.com': NYTimesExtractor, 'www.theatlantic.com': TheAtlanticExtractor, 'www.newyorker.com': NewYorkerExtractor, - 'www.wired.com': WiredExtractor + 'www.wired.com': WiredExtractor, + 'www.msn.com': MSNExtractor }; diff --git a/dist/mercury.js.map b/dist/mercury.js.map index 2bcd35bd..1c7a5508 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/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/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/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 'babel-polyfill';\n\nimport 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 $;\n}\n","export const IS_LINK = new RegExp('https?://', 'i');\nexport const IS_IMAGE = new RegExp('.(png|gif|jpe?g)', 'i');\n\nexport const TAGS_TO_REMOVE = [\n 'script',\n 'style',\n 'form',\n].join(',');\n","import 'babel-polyfill';\n\nimport {\n IS_LINK,\n IS_IMAGE,\n} from './constants';\n\n// Convert all instances of images with potentially\n// lazy loaded images into normal images.\n// Many sites will have img tags with no source, or an image tag with a src\n// attribute that a is a placeholer. We need to be able to properly fill in\n// the src attribute so the images are no longer lazy loaded.\nexport default function convertLazyLoadedImages($) {\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 'babel-polyfill';\n\nimport 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 && $children.get(0).tagName === 'img') {\n return 'figure';\n }\n\n return null;\n },\n },\n },\n\n title: {\n selectors: [\n 'h1.lede-feature-title',\n 'h1.headline-primary',\n 'h1',\n ],\n },\n\n author: {\n selectors: [\n '.by-authors',\n '.lede-feature-author',\n ],\n },\n\n dek: {\n selectors: [\n '.lede-feature-teaser',\n ],\n },\n\n date_published: {\n selectors: [\n ['time.article-timestamp[datetime]', 'datetime'],\n 'time.article-timestamp',\n ],\n },\n};\n","export const BloggerExtractor = {\n domain: 'blogspot.com',\n content: {\n // Blogger is insane and does not load its content\n // initially in the page, but it's all there\n // in noscript\n selectors: [\n '.post-content noscript',\n ],\n\n // Selectors to remove from the extracted content\n clean: [\n ],\n\n // Convert the noscript tag to a div\n transforms: {\n noscript: 'div',\n },\n },\n\n author: {\n selectors: [\n '.post-author-name',\n ],\n },\n\n title: {\n selectors: [\n 'h2.title',\n ],\n },\n\n date_published: {\n selectors: [\n 'span.publishdate',\n ],\n },\n};\n","export const WikipediaExtractor = {\n domain: 'wikipedia.org',\n content: {\n selectors: [\n '#mw-content-text',\n ],\n\n defaultCleaner: false,\n\n // transform top infobox to an image with caption\n transforms: {\n '.infobox img': ($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 '.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","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';\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\n};\n\nexport default Extractors;\n","// Spacer images to be removed\nexport const SPACER_RE = new RegExp('trans|transparent|spacer|blank', 'i');\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', 'score'];\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 $;\n}\n","import 'babel-polyfill';\n\nexport default function convertNodeTo($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 STRIP_OUTPUT_TAGS,\n} from './constants';\n\nexport default function stripJunkTags(article, $, tags = []) {\n if (tags.length === 0) {\n tags = STRIP_OUTPUT_TAGS;\n }\n\n $(tags.join(','), article).remove();\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 $;\n}\n","import 'babel-polyfill';\n\nimport { WHITELIST_ATTRS_RE } from './constants';\n\nfunction removeAllButWhitelist($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.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","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';\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, { $ }) {\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 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 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 // 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 'babel-polyfill';\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';\nimport 'babel-polyfill';\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 'babel-polyfill';\n\nimport { 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 'babel-polyfill';\nimport 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 'babel-polyfill';\nimport 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.$ = $;\n }\n\n const title = this.title(options);\n const date_published = this.date_published(options);\n const author = this.author(options);\n const content = this.content({ ...options, title });\n const lead_image_url = this.lead_image_url({ ...options, content });\n const dek = this.dek({ ...options, content });\n const next_page_url = this.next_page_url(options);\n const excerpt = this.excerpt({ ...options, content });\n const word_count = this.word_count({ ...options, content });\n const direction = this.direction({ title });\n const { url, domain } = this.url_and_domain(options);\n\n return {\n title,\n author,\n date_published: date_published || null,\n dek,\n lead_image_url,\n content,\n next_page_url,\n url,\n domain,\n excerpt,\n word_count,\n direction,\n };\n },\n};\n\nexport default GenericExtractor;\n","import URL from 'url';\n\nimport Extractors from './all';\nimport GenericExtractor from './generic';\n\nexport default function getExtractor(url, parsedUrl) {\n parsedUrl = parsedUrl || URL.parse(url);\n const { hostname } = parsedUrl;\n const baseDomain = hostname.split('.').slice(-2).join('.');\n\n return Extractors[hostname] || Extractors[baseDomain] || GenericExtractor;\n}\n","import 'babel-polyfill';\n\nimport Cleaners from 'cleaners';\nimport { convertNodeTo } from 'utils/dom';\nimport GenericExtractor from './generic';\n\n// Remove elements by an array of selectors\nexport function cleanBySelectors($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 dek = extractResult({ ...opts, type: 'dek', content });\n const excerpt = extractResult({ ...opts, type: 'excerpt', content });\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 'babel-polyfill';\nimport { 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","Extractors","SPACER_RE","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","$img","height","parseInt","removeSpacers","cleanImages","$article","stripJunkTags","article","tags","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","split","isGoodSegment","segment","firstSegmentHasLetters","goodSegment","toLowerCase","articleBaseUrl","parsed","protocol","host","path","cleanedSegments","reverse","rawSegment","possibleSegment","fileExt","push","SENTENCE_END_RE","hasSentenceEnd","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","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","excerpt","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;;AClBA,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;;;AC9EA,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;;ACIP;;;;;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;;;ACtBF,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;;;ACRF,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;;ACpEO,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,UADS;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,WADS,EAET,SAFS;GATiB;;WAerB;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;GA1CqB;;kBAsDd,IAtDc;;kBAwDd,IAxDc;;OA0DzB,IA1DyB;;iBA4Df,IA5De;;WA8DrB;CA9DJ;;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;;ACMP,IAAMC,aAAa;eACJpB,cADI;kBAEDG,gBAFC;mBAGAC,kBAHA;iBAIFI,gBAJE;qBAKEK,gBALF;yBAMMI,oBANN;uBAOIC,kBAPJ;mBAQAC;;CARnB,CAYA;;ACrBA;AACA,AAAO,IAAME,YAAY,IAAIzF,MAAJ,CAAW,gCAAX,EAA6C,GAA7C,CAAlB;;;AAGP,AAAO,IAAM0F,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,aAAa1F,IAAb,CAAkB,GAAlB,CAAzB;AACP,AAAO,IAAM+F,kBAAkB,CAAC,KAAD,EAAQ,QAAR,EAAkB,MAAlB,EAA0B,OAA1B,EAAmC,IAAnC,EAAyC,KAAzC,EAAgD,OAAhD,CAAxB;AACP,AAAO,IAAMC,qBAAqB,IAAIjG,MAAJ,QAAgBgG,gBAAgB/F,IAAhB,CAAqB,GAArB,CAAhB,SAA+C,GAA/C,CAA3B;;;AAGP,AAAO,IAAMiG,oBAAoB,CAAC,GAAD,CAA1B;AACP,AAAO,IAAMC,yBAAyBD,kBAAkBL,GAAlB,CAAsB;SAAUO,GAAV;CAAtB,EAA6CnG,IAA7C,CAAkD,GAAlD,CAA/B;;;AAGP,AAAO,IAAMoG,2BAA2B,CAAC,IAAD,EAAO,IAAP,EAAa,OAAb,EAAsB,KAAtB,EAA6B,QAA7B,EAAuC,MAAvC,EAA+CpG,IAA/C,CAAoD,GAApD,CAAjC;;;AAGP,IAAMqG,cAAc,CAAC,IAAD,EAAO,IAAP,EAAa,IAAb,EAAmB,IAAnB,EAAyB,IAAzB,CAApB;AACA,AAAO,IAAMC,kBAAkBD,YAAYrG,IAAZ,CAAiB,GAAjB,CAAxB;;;;;;;;AASP,AAAO,IAAMuG,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,EASjCzG,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM0G,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,IAAI5G,MAAJ,QAAgB2G,uBAAuB1G,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;AAGP,AAYA,AAAO,IAAM4G,cAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,iBAAiB,IAAI9G,MAAJ,CAAW6G,YAAY5G,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAM8G,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,IAAIhH,MAAJ,CAAW+G,qBAAqB9G,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA;;;;AAIA,AAAO,IAAMgH,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,IAAIlH,MAAJ,CAAWiH,qBAAqBhH,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMkH,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,IAAIrH,MAAJ,CAAWoH,sBAAsBnH,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;AAGP,AAAO,IAAMqH,UAAU,IAAItH,MAAJ,CAAW,iBAAX,EAA8B,GAA9B,CAAhB;;AAEP,AAMA,AAIA,AAIA,AAGA,AAGA;;AAEA,AAAO,IAAMuH,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,IAAIxH,MAAJ,QAAgBuH,iBAAiBtH,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAMwH,sBAAsBjB,8BAA8BvG,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAMyH,uBAAuB,IAAI1H,MAAJ,CAAWyH,mBAAX,EAAgC,GAAhC,CAA7B;;AAEP,IAAME,sBAAsBlB,8BAA8BxG,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM2H,uBAAuB,IAAI5H,MAAJ,CAAW2H,mBAAX,EAAgC,GAAhC,CAA7B,CAEP,AAGA,AACA,AACA,AAEA;;AC3Xe,SAASE,uBAAT,CAAiCjG,CAAjC,EAAoC;;;;;;;;;;IAU/C,GAAF,EAAOkG,GAAP,CAAW,GAAX,EAAgB/F,IAAhB,CAAqB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;QAC9BC,QAAQN,EAAEK,IAAF,CAAd;QACM8F,UAAU7F,MAAME,IAAN,CAAW,OAAX,CAAhB;QACM4F,KAAK9F,MAAME,IAAN,CAAW,IAAX,CAAX;QACI,CAAC4F,EAAD,IAAO,CAACD,OAAZ,EAAqB;;QAEfE,cAAgBF,WAAW,EAA3B,WAAiCC,MAAM,EAAvC,CAAN;QACIJ,qBAAqBzG,IAArB,CAA0B8G,UAA1B,CAAJ,EAA2C;;KAA3C,MAEO,IAAIP,qBAAqBvG,IAArB,CAA0B8G,UAA1B,CAAJ,EAA2C;YAC1C3E,MAAN;;GAVJ;;SAcO1B,CAAP;;;AC3BF;;;;;;;;;;AAUA,AAAe,SAASsG,OAAT,CAAiBtG,CAAjB,EAAoB;MAC7BuG,aAAa,KAAjB;IACE,IAAF,EAAQpG,IAAR,CAAa,UAACgB,KAAD,EAAQqF,OAAR,EAAoB;QACzBC,cAAczG,EAAEwG,OAAF,EAAWE,IAAX,GAAkBnI,GAAlB,CAAsB,CAAtB,CAApB;;QAEIkI,eAAeA,YAAY/D,OAAZ,KAAwB,IAA3C,EAAiD;mBAClC,IAAb;QACE8D,OAAF,EAAW9E,MAAX;KAFF,MAGO,IAAI6E,UAAJ,EAAgB;mBACR,KAAb;;mBAEaC,OAAb,EAAsBxG,CAAtB,EAAyB,IAAzB;;GATJ;;SAaOA,CAAP;;;ACzBF;;;;;;;;;;;AAWA,AAAe,SAAS2G,YAAT,CAAsBtG,IAAtB,EAA4BL,CAA5B,EAA2C;MAAZ4G,EAAY,uEAAP,KAAO;;MAClDtG,QAAQN,EAAEK,IAAF,CAAd;;MAEIuG,EAAJ,EAAQ;QACFC,UAAUxG,KAAKyG,WAAnB;QACMC,IAAI/G,EAAE,SAAF,CAAV;;;;WAIO6G,WAAW,EAAEA,QAAQnE,OAAR,IAAmBkD,oBAAoBrG,IAApB,CAAyBsH,QAAQnE,OAAjC,CAArB,CAAlB,EAAmF;UAC3EoE,cAAcD,QAAQC,WAA5B;QACED,OAAF,EAAWG,QAAX,CAAoBD,CAApB;gBACUD,WAAV;;;UAGI1D,WAAN,CAAkB2D,CAAlB;UACMrF,MAAN;WACO1B,CAAP;;;SAGKA,CAAP;;;AC7BF,SAASiH,WAAT,CAAqBjH,CAArB,EAAwB;IACpB,KAAF,EAASG,IAAT,CAAc,UAACgB,KAAD,EAAQ+F,GAAR,EAAgB;QACtBC,OAAOnH,EAAEkH,GAAF,CAAb;QACME,cAAcD,KAAK7E,QAAL,CAAcwC,mBAAd,EAAmCvC,MAAnC,KAA8C,CAAlE;;QAEI6E,WAAJ,EAAiB;oBACDD,IAAd,EAAoBnH,CAApB,EAAuB,GAAvB;;GALJ;;SASOA,CAAP;;;AAGF,SAASqH,YAAT,CAAsBrH,CAAtB,EAAyB;IACrB,MAAF,EAAUG,IAAV,CAAe,UAACgB,KAAD,EAAQmG,IAAR,EAAiB;QACxBC,QAAQvH,EAAEsH,IAAF,CAAd;QACMF,cAAcG,MAAMzE,OAAN,CAAc,QAAd,EAAwBP,MAAxB,KAAmC,CAAvD;QACI6E,WAAJ,EAAiB;oBACDG,KAAd,EAAqBvH,CAArB,EAAwB,GAAxB;;GAJJ;;SAQOA,CAAP;;;;;;;;;;;;;;;AAeF,AAAe,SAASwH,mBAAT,CAA6BxH,CAA7B,EAAgC;MACzCsG,QAAQtG,CAAR,CAAJ;MACIiH,YAAYjH,CAAZ,CAAJ;MACIqH,aAAarH,CAAb,CAAJ;;SAEOA,CAAP;;;AC5Ca,SAASyH,aAAT,CAAuBnH,KAAvB,EAA8BN,CAA9B,EAA4C;MAAXwE,GAAW,uEAAL,GAAK;;MACnDnE,OAAOC,MAAM/B,GAAN,CAAU,CAAV,CAAb;MACI,CAAC8B,IAAL,EAAW;WACFL,CAAP;;;mBAEkBM,MAAM/B,GAAN,CAAU,CAAV,CALqC;;MAKjDyC,OALiD,cAKjDA,OALiD;;MAMnD0G,eAAe,iBAAgB1G,OAAhB,EACQiD,GADR,CACY;WAAU0D,GAAV,SAAiB3G,QAAQ2G,GAAR,CAAjB;GADZ,EAEQtJ,IAFR,CAEa,GAFb,CAArB;;QAIM+E,WAAN,OAAsBoB,GAAtB,SAA6BkD,YAA7B,SAA6CpH,MAAMkB,QAAN,EAA7C,UAAkEgD,GAAlE;SACOxE,CAAP;;;ACXF,SAAS4H,cAAT,CAAwBC,IAAxB,EAA8B7H,CAA9B,EAAiC;MACzB8H,SAASC,SAASF,KAAKrH,IAAL,CAAU,QAAV,CAAT,EAA8B,EAA9B,CAAf;MACM+C,QAAQwE,SAASF,KAAKrH,IAAL,CAAU,OAAV,CAAT,EAA6B,EAA7B,KAAoC,EAAlD;;;;;MAKI,CAACsH,UAAU,EAAX,IAAiB,EAAjB,IAAuBvE,QAAQ,EAAnC,EAAuC;SAChC7B,MAAL;GADF,MAEO,IAAIoG,MAAJ,EAAY;;;;SAIZrH,UAAL,CAAgB,QAAhB;;;SAGKT,CAAP;;;;;AAKF,SAASgI,aAAT,CAAuBH,IAAvB,EAA6B7H,CAA7B,EAAgC;MAC1B6D,UAAUtE,IAAV,CAAesI,KAAKrH,IAAL,CAAU,KAAV,CAAf,CAAJ,EAAsC;SAC/BkB,MAAL;;;SAGK1B,CAAP;;;AAGF,AAAe,SAASiI,WAAT,CAAqBC,QAArB,EAA+BlI,CAA/B,EAAkC;WACtCuB,IAAT,CAAc,KAAd,EAAqBpB,IAArB,CAA0B,UAACgB,KAAD,EAAQJ,GAAR,EAAgB;QAClC8G,OAAO7H,EAAEe,GAAF,CAAb;;mBAEe8G,IAAf,EAAqB7H,CAArB;kBACc6H,IAAd,EAAoB7H,CAApB;GAJF;;SAOOA,CAAP;;;ACnCa,SAASmI,aAAT,CAAuBC,OAAvB,EAAgCpI,CAAhC,EAA8C;MAAXqI,IAAW,uEAAJ,EAAI;;MACvDA,KAAK9F,MAAL,KAAgB,CAApB,EAAuB;WACduB,iBAAP;;;IAGAuE,KAAKhK,IAAL,CAAU,GAAV,CAAF,EAAkB+J,OAAlB,EAA2B1G,MAA3B;;SAEO1B,CAAP;;;ACTF;;;;AAGA,AAAe,SAASsI,UAAT,CAAoBF,OAApB,EAA6BpI,CAA7B,EAAgC;MACvCuI,SAASvI,EAAE,IAAF,EAAQoI,OAAR,CAAf;;MAEIG,OAAOhG,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;;;ACZF,SAASwI,qBAAT,CAA+BN,QAA/B,EAAyC;;WAE9B3G,IAAT,CAAc,GAAd,EAAmBpB,IAAnB,CAAwB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;SAClCW,OAAL,GAAe,iBAAgBX,KAAKW,OAArB,EAA8ByH,MAA9B,CAAqC,UAACC,GAAD,EAAMlI,IAAN,EAAe;UAC7D6D,mBAAmB9E,IAAnB,CAAwBiB,IAAxB,CAAJ,EAAmC;4BACrBkI,GAAZ,sBAAkBlI,IAAlB,EAAyBH,KAAKW,OAAL,CAAaR,IAAb,CAAzB;;;aAGKkI,GAAP;KALa,EAMZ,EANY,CAAf;GADF;;;;;;;;;;AAkBF,AAAe,SAASC,eAAT,CAAyBT,QAAzB,EAAmC;wBAC1BA,QAAtB;;SAEOA,QAAP;;;AC3Ba,SAASU,WAAT,CAAqBV,QAArB,EAA+BlI,CAA/B,EAAkC;WACtCuB,IAAT,CAAc,GAAd,EAAmBpB,IAAnB,CAAwB,UAACgB,KAAD,EAAQ4F,CAAR,EAAc;QAC9B8B,KAAK7I,EAAE+G,CAAF,CAAX;QACI8B,GAAGC,IAAH,GAAUC,IAAV,OAAqB,EAAzB,EAA6BF,GAAGnH,MAAH;GAF/B;;SAKO1B,CAAP;;;ACNF;;;;;;AAMA,AAAO,IAAM4E,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,EASjCzG,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM0G,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,IAAI5G,MAAJ,QAAgB2G,yBAAuB1G,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;;;;AAMP,AAAO,IAAM2K,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,IAAM/D,gBAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,mBAAiB,IAAI9G,MAAJ,CAAW6G,cAAY5G,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAM8G,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,IAAIhH,MAAJ,CAAW+G,uBAAqB9G,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAM4K,sBAAoB,IAAI7K,MAAJ,CAAW,qBAAX,EAAkC,GAAlC,CAA1B;;;;;;AAMP,AAAO,IAAMiH,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,IAAIlH,MAAJ,CAAWiH,uBAAqBhH,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA,AAGA,AAGA;;AAEA,AAAO,IAAMsH,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,IAAIxH,MAAJ,QAAgBuH,mBAAiBtH,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAMwH,wBAAsBjB,gCAA8BvG,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,IAAM0H,wBAAsBlB,gCAA8BxG,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,AAGA,AAAO,IAAM6K,yBAAuB,IAAI9K,MAAJ,CAAW,mBAAX,EAAgC,GAAhC,CAA7B;AACP,AAAO,IAAM+K,uBAAqB,IAAI/K,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAA3B;AACP,AAAO,IAAMgL,aAAW,IAAIhL,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB,CAEP;;AC3SA;AACA,AAAe,SAASiL,SAAT,CAAmBhJ,IAAnB,EAAyB;MAChC8F,UAAU9F,KAAKG,IAAL,CAAU,OAAV,CAAhB;MACM4F,KAAK/F,KAAKG,IAAL,CAAU,IAAV,CAAX;MACI8I,QAAQ,CAAZ;;MAEIlD,EAAJ,EAAQ;;QAEFhB,oBAAkB7F,IAAlB,CAAuB6G,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;QAEEd,oBAAkB/F,IAAlB,CAAuB6G,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;;;MAIAD,OAAJ,EAAa;QACPmD,UAAU,CAAd,EAAiB;;;UAGXlE,oBAAkB7F,IAAlB,CAAuB4G,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;UAEEb,oBAAkB/F,IAAlB,CAAuB4G,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;;;;;;QAOAjB,iBAAe3F,IAAf,CAAoB4G,OAApB,CAAJ,EAAkC;eACvB,EAAT;;;;;;;QAOE8C,oBAAkB1J,IAAlB,CAAuB4G,OAAvB,CAAJ,EAAqC;eAC1B,EAAT;;;;SAIGmD,KAAP;;;ACpDF;;;AAGA,AAAe,SAASC,QAAT,CAAkBjJ,KAAlB,EAAyB;SAC/BkJ,WAAWlJ,MAAME,IAAN,CAAW,OAAX,CAAX,KAAmC,IAA1C;;;ACJF;AACA,AAAe,SAASiJ,WAAT,CAAqBX,IAArB,EAA2B;SACjC,CAACA,KAAKY,KAAL,CAAW,IAAX,KAAoB,EAArB,EAAyBnH,MAAhC;;;ACFF,IAAMoH,QAAQ,IAAIvL,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAd;;AAEA,AAAe,SAASwL,WAAT,CAAqBC,UAArB,EAAgD;MAAfnH,OAAe,uEAAL,GAAK;;MACvDoH,SAASD,aAAa,EAA5B;;MAEIC,SAAS,CAAb,EAAgB;QACVC,oBAAJ;;;;;;;QAOIJ,MAAMpK,IAAN,CAAWmD,OAAX,CAAJ,EAAyB;oBACToH,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,CAAwB9J,IAAxB,EAA8B;MACvCiJ,QAAQ,CAAZ;MACMR,OAAOzI,KAAKyI,IAAL,GAAYC,IAAZ,EAAb;MACMc,aAAaf,KAAKvG,MAAxB;;;MAGIsH,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,CAAkB/J,KAAlB,EAAyBN,CAAzB,EAA4BsJ,KAA5B,EAAmC;QAC1C9I,IAAN,CAAW,OAAX,EAAoB8I,KAApB;SACOhJ,KAAP;;;ACEa,SAASgK,QAAT,CAAkBhK,KAAlB,EAAyBN,CAAzB,EAA4BuK,MAA5B,EAAoC;MAC7C;QACIjB,QAAQkB,eAAelK,KAAf,EAAsBN,CAAtB,IAA2BuK,MAAzC;aACSjK,KAAT,EAAgBN,CAAhB,EAAmBsJ,KAAnB;GAFF,CAGE,OAAOmB,CAAP,EAAU;;;;SAILnK,KAAP;;;ACXF;AACA,AAAe,SAASoK,WAAT,CAAqBrK,IAArB,EAA2BL,CAA3B,EAA8BsJ,KAA9B,EAAqC;MAC5CqB,SAAStK,KAAKsK,MAAL,EAAf;MACIA,MAAJ,EAAY;aACDA,MAAT,EAAiB3K,CAAjB,EAAoBsJ,QAAQ,IAA5B;;;SAGKjJ,IAAP;;;ACFF;;;AAGA,AAAe,SAASmK,cAAT,CAAwBlK,KAAxB,EAA+BN,CAA/B,EAAsD;MAApB4K,WAAoB,uEAAN,IAAM;;MAC/DtB,QAAQC,SAASjJ,KAAT,CAAZ;;MAEIgJ,KAAJ,EAAW;WACFA,KAAP;;;UAGMuB,UAAUvK,KAAV,CAAR;;MAEIsK,WAAJ,EAAiB;aACNvB,UAAU/I,KAAV,CAAT;;;cAGUA,KAAZ,EAAmBN,CAAnB,EAAsBsJ,KAAtB;;SAEOA,KAAP;;;AClBF;;AAEA,AAAe,SAASuB,SAAT,CAAmBvK,KAAnB,EAA0B;mBACnBA,MAAM/B,GAAN,CAAU,CAAV,CADmB;;MAC/BmE,OAD+B,cAC/BA,OAD+B;;;;;;MAMnCwG,uBAAqB3J,IAArB,CAA0BmD,OAA1B,CAAJ,EAAwC;WAC/ByH,eAAe7J,KAAf,CAAP;GADF,MAEO,IAAIoC,YAAY,KAAhB,EAAuB;WACrB,CAAP;GADK,MAEA,IAAIyG,qBAAmB5J,IAAnB,CAAwBmD,OAAxB,CAAJ,EAAsC;WACpC,CAAP;GADK,MAEA,IAAI0G,WAAS7J,IAAT,CAAcmD,OAAd,CAAJ,EAA4B;WAC1B,CAAC,CAAR;GADK,MAEA,IAAIA,YAAY,IAAhB,EAAsB;WACpB,CAAC,CAAR;;;SAGK,CAAP;;;ACjBF,SAAS2E,cAAT,CAAsB/G,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,SAAS8K,UAAT,CAAoBxK,KAApB,EAA2BN,CAA3B,EAA8BsJ,KAA9B,EAAqC;MAC/BhJ,KAAJ,EAAW;mBACIA,KAAb,EAAoBN,CAApB;aACSM,KAAT,EAAgBN,CAAhB,EAAmBsJ,KAAnB;;;;AAIJ,SAASyB,OAAT,CAAiB/K,CAAjB,EAAoB4K,WAApB,EAAiC;IAC7B,QAAF,EAAY1E,GAAZ,CAAgB,SAAhB,EAA2B/F,IAA3B,CAAgC,UAACgB,KAAD,EAAQd,IAAR,EAAiB;;;QAG3CC,QAAQN,EAAEK,IAAF,CAAZ;YACQgK,SAAS/J,KAAT,EAAgBN,CAAhB,EAAmBwK,eAAelK,KAAf,EAAsBN,CAAtB,EAAyB4K,WAAzB,CAAnB,CAAR;;QAEM/H,UAAUvC,MAAMqK,MAAN,EAAhB;QACMK,WAAWH,UAAUvK,KAAV,CAAjB;;eAEWuC,OAAX,EAAoB7C,CAApB,EAAuBgL,QAAvB,EAAiCJ,WAAjC;QACI/H,OAAJ,EAAa;;;iBAGAA,QAAQ8H,MAAR,EAAX,EAA6B3K,CAA7B,EAAgCgL,WAAW,CAA3C,EAA8CJ,WAA9C;;GAbJ;;SAiBO5K,CAAP;;;;;AAKF,AAAe,SAASiL,YAAT,CAAsBjL,CAAtB,EAA6C;MAApB4K,WAAoB,uEAAN,IAAM;;;;4BAGlC3J,OAAxB,CAAgC,gBAAqC;;;QAAnCiK,cAAmC;QAAnBC,aAAmB;;MAC9DD,cAAL,SAAuBC,aAAvB,EAAwChL,IAAxC,CAA6C,UAACgB,KAAD,EAAQd,IAAR,EAAiB;eACnDL,EAAEK,IAAF,EAAQsK,MAAR,CAAeO,cAAf,CAAT,EAAyClL,CAAzC,EAA4C,EAA5C;KADF;GADF;;;;;;;UAWQA,CAAR,EAAW4K,WAAX;UACQ5K,CAAR,EAAW4K,WAAX;;SAEO5K,CAAP;;;ACpEF,IAAMoL,eAAe,SAArB;;AAEA,AAAe,SAASC,eAAT,CAAyBvC,IAAzB,EAA+B;SACrCA,KAAKtF,OAAL,CAAa4H,YAAb,EAA2B,GAA3B,EAAgCrC,IAAhC,EAAP;;;ACHF;;;;;AAKA,AAAe,SAASuC,cAAT,CAAwB9L,GAAxB,EAA6B+L,SAA7B,EAAwC;MAC/CC,UAAUD,UAAUhK,IAAV,CAAe;WAAMkK,GAAGlM,IAAH,CAAQC,GAAR,CAAN;GAAf,CAAhB;MACIgM,OAAJ,EAAa;WACJA,QAAQE,IAAR,CAAalM,GAAb,EAAkB,CAAlB,CAAP;;;SAGK,IAAP;;;ACXF;;;;;;;;;;;;;;;;AAgBA,AAAO,IAAMmM,kBAAkB,IAAIvN,MAAJ,CAAW,0EAAX,EAAuF,GAAvF,CAAxB;;AAEP,AAAO,IAAMwN,eAAe,QAArB;;AAEP,AAAO,IAAMC,cAAc,WAApB;AACP,AAAO,IAAMC,cAAc,WAApB;;ACnBQ,SAASC,cAAT,CAAwBvM,GAAxB,EAA6B;MACpCwM,UAAUxM,IAAIkK,KAAJ,CAAUiC,eAAV,CAAhB;MACI,CAACK,OAAL,EAAc,OAAO,IAAP;;MAERC,UAAUlE,SAASiE,QAAQ,CAAR,CAAT,EAAqB,EAArB,CAAhB;;;;SAIOC,UAAU,GAAV,GAAgBA,OAAhB,GAA0B,IAAjC;;;ACVa,SAASC,YAAT,CAAsB1M,GAAtB,EAA2B;SACjCA,IAAI2M,KAAJ,CAAU,GAAV,EAAe,CAAf,EAAkB3I,OAAlB,CAA0B,KAA1B,EAAiC,EAAjC,CAAP;;;ACOF,SAAS4I,aAAT,CAAuBC,OAAvB,EAAgClL,KAAhC,EAAuCmL,sBAAvC,EAA+D;MACzDC,cAAc,IAAlB;;;;MAIIpL,QAAQ,CAAR,IAAa2K,YAAYvM,IAAZ,CAAiB8M,OAAjB,CAAb,IAA0CA,QAAQ9J,MAAR,GAAiB,CAA/D,EAAkE;kBAClD,IAAd;;;;;MAKEpB,UAAU,CAAV,IAAekL,QAAQG,WAAR,OAA0B,OAA7C,EAAsD;kBACtC,KAAd;;;;;MAKErL,QAAQ,CAAR,IAAakL,QAAQ9J,MAAR,GAAiB,CAA9B,IAAmC,CAAC+J,sBAAxC,EAAgE;kBAChD,KAAd;;;SAGKC,WAAP;;;;;;AAMF,AAAe,SAASE,cAAT,CAAwBjN,GAAxB,EAA6BkN,MAA7B,EAAqC;MAC5CjN,YAAYiN,UAAUhN,IAAIC,KAAJ,CAAUH,GAAV,CAA5B;MACQmN,QAF0C,GAEjBlN,SAFiB,CAE1CkN,QAF0C;MAEhCC,IAFgC,GAEjBnN,SAFiB,CAEhCmN,IAFgC;MAE1BC,IAF0B,GAEjBpN,SAFiB,CAE1BoN,IAF0B;;;MAI9CP,yBAAyB,KAA7B;MACMQ,kBAAkBD,KAAKV,KAAL,CAAW,GAAX,EACvBY,OADuB,GAEvBtE,MAFuB,CAEhB,UAACC,GAAD,EAAMsE,UAAN,EAAkB7L,KAAlB,EAA4B;QAC9BkL,UAAUW,UAAd;;;QAGIX,QAAQnK,QAAR,CAAiB,GAAjB,CAAJ,EAA2B;2BACUmK,QAAQF,KAAR,CAAc,GAAd,CADV;;;;UAClBc,eADkB;UACDC,OADC;;UAErBrB,YAAYtM,IAAZ,CAAiB2N,OAAjB,CAAJ,EAA+B;kBACnBD,eAAV;;;;;;QAMAtB,gBAAgBpM,IAAhB,CAAqB8M,OAArB,KAAiClL,QAAQ,CAA7C,EAAgD;gBACpCkL,QAAQ7I,OAAR,CAAgBmI,eAAhB,EAAiC,EAAjC,CAAV;;;;;;;QAOExK,UAAU,CAAd,EAAiB;+BACUyK,aAAarM,IAAb,CAAkB8M,OAAlB,CAAzB;;;;QAIED,cAAcC,OAAd,EAAuBlL,KAAvB,EAA8BmL,sBAA9B,CAAJ,EAA2D;UACrDa,IAAJ,CAASd,OAAT;;;WAGK3D,GAAP;GAhCsB,EAiCrB,EAjCqB,CAAxB;;SAmCUiE,QAAV,UAAuBC,IAAvB,GAA8BE,gBAAgBC,OAAhB,GAA0B1O,IAA1B,CAA+B,GAA/B,CAA9B;;;AC3EF;;AAEA,IAAM+O,kBAAkB,IAAIhP,MAAJ,CAAW,QAAX,CAAxB;AACA,AAAe,SAASiP,cAAT,CAAwBvE,IAAxB,EAA8B;SACpCsE,gBAAgB7N,IAAhB,CAAqBuJ,IAArB,CAAP;;;ACKF;;;;;AAKA,AAAe,SAASwE,aAAT,CAAuBC,UAAvB,EAAmCC,QAAnC,EAA6CxN,CAA7C,EAAgD;MACzD,CAACuN,WAAW5C,MAAX,GAAoBpI,MAAzB,EAAiC;WACxBgL,UAAP;;;MAGIE,wBAAwBzD,KAAKE,GAAL,CAAS,EAAT,EAAasD,WAAW,IAAxB,CAA9B;MACME,cAAc1N,EAAE,aAAF,CAApB;;aAEW2K,MAAX,GAAoBrI,QAApB,GAA+BnC,IAA/B,CAAoC,UAACgB,KAAD,EAAQ0F,OAAR,EAAoB;QAChD8G,WAAW3N,EAAE6G,OAAF,CAAjB;;QAEI7B,4BAA0BzF,IAA1B,CAA+BsH,QAAQnE,OAAvC,CAAJ,EAAqD;aAC5C,IAAP;;;QAGIkL,eAAerE,SAASoE,QAAT,CAArB;QACIC,YAAJ,EAAkB;UACZD,aAAaJ,UAAjB,EAA6B;oBACfpK,MAAZ,CAAmBwK,QAAnB;OADF,MAEO;YACDE,eAAe,CAAnB;YACMC,UAAUC,YAAYJ,QAAZ,CAAhB;;;;YAIIG,UAAU,IAAd,EAAoB;0BACF,EAAhB;;;;;YAKEA,WAAW,GAAf,EAAoB;0BACF,EAAhB;;;;;YAKEH,SAASnN,IAAT,CAAc,OAAd,MAA2B+M,WAAW/M,IAAX,CAAgB,OAAhB,CAA/B,EAAyD;0BACvCgN,WAAW,GAA3B;;;YAGIQ,WAAWJ,eAAeC,YAAhC;;YAEIG,YAAYP,qBAAhB,EAAuC;iBAC9BC,YAAYvK,MAAZ,CAAmBwK,QAAnB,CAAP;SADF,MAEO,IAAI9G,QAAQnE,OAAR,KAAoB,GAAxB,EAA6B;cAC5BuL,iBAAiBN,SAAS7E,IAAT,EAAvB;cACMoF,uBAAuBrE,WAAWoE,cAAX,CAA7B;;cAEIC,uBAAuB,EAAvB,IAA6BJ,UAAU,IAA3C,EAAiD;mBACxCJ,YAAYvK,MAAZ,CAAmBwK,QAAnB,CAAP;WADF,MAEO,IAAIO,wBAAwB,EAAxB,IAA8BJ,YAAY,CAA1C,IACDT,eAAeY,cAAf,CADH,EACmC;mBACjCP,YAAYvK,MAAZ,CAAmBwK,QAAnB,CAAP;;;;;;WAMD,IAAP;GAnDF;;SAsDOD,WAAP;;;ACxEF;;AAEA,AAAe,SAASS,gBAAT,CAA0BnO,CAA1B,EAA6B;MACtCuN,mBAAJ;MACIC,WAAW,CAAf;;IAEE,SAAF,EAAarN,IAAb,CAAkB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;;QAE7B2E,4BAA0BzF,IAA1B,CAA+Bc,KAAKqC,OAApC,CAAJ,EAAkD;;;;QAI5CpC,QAAQN,EAAEK,IAAF,CAAd;QACMiJ,QAAQC,SAASjJ,KAAT,CAAd;;QAEIgJ,QAAQkE,QAAZ,EAAsB;iBACTlE,KAAX;mBACahJ,KAAb;;GAXJ;;;;MAiBI,CAACiN,UAAL,EAAiB;WACRvN,EAAE,MAAF,KAAaA,EAAE,GAAF,EAAOoO,KAAP,EAApB;;;eAGWd,cAAcC,UAAd,EAA0BC,QAA1B,EAAoCxN,CAApC,CAAb;;SAEOuN,UAAP;;;ACtBF,SAASc,mBAAT,CAA6B/N,KAA7B,EAAoCN,CAApC,EAAuCsO,MAAvC,EAA+C;;;;;MAKzChO,MAAMiO,QAAN,CAAe,qBAAf,CAAJ,EAA2C;;;;MAIrCtM,UAAUoJ,gBAAgB/K,MAAMwI,IAAN,EAAhB,CAAhB;;MAEIW,YAAYxH,OAAZ,IAAuB,EAA3B,EAA+B;QACvBuM,SAASxO,EAAE,GAAF,EAAOM,KAAP,EAAciC,MAA7B;QACMkM,aAAazO,EAAE,OAAF,EAAWM,KAAX,EAAkBiC,MAArC;;;QAGIkM,aAAcD,SAAS,CAA3B,EAA+B;YACvB9M,MAAN;;;;QAIIpC,gBAAgB2C,QAAQM,MAA9B;QACMmM,WAAW1O,EAAE,KAAF,EAASM,KAAT,EAAgBiC,MAAjC;;;;QAIIjD,gBAAgB,EAAhB,IAAsBoP,aAAa,CAAvC,EAA0C;YAClChN,MAAN;;;;QAIIoM,UAAUC,YAAYzN,KAAZ,CAAhB;;;;;QAKIgO,SAAS,EAAT,IAAeR,UAAU,GAAzB,IAAgCxO,gBAAgB,EAApD,EAAwD;YAChDoC,MAAN;;;;;;QAME4M,UAAU,EAAV,IAAgBR,UAAU,GAA9B,EAAmC;;;;UAI3BpL,UAAUpC,MAAM/B,GAAN,CAAU,CAAV,EAAamE,OAA7B;UACMiM,aAAajM,YAAY,IAAZ,IAAoBA,YAAY,IAAnD;UACIiM,UAAJ,EAAgB;YACRC,eAAetO,MAAMuO,IAAN,EAArB;YACID,gBAAgBvD,gBAAgBuD,aAAa9F,IAAb,EAAhB,EAAqCsB,KAArC,CAA2C,CAAC,CAA5C,MAAmD,GAAvE,EAA4E;;;;;YAKxE1I,MAAN;;;;QAIIoN,cAAc9O,EAAE,QAAF,EAAYM,KAAZ,EAAmBiC,MAAvC;;;QAGIuM,cAAc,CAAd,IAAmBxP,gBAAgB,GAAvC,EAA4C;YACpCoC,MAAN;;;;;;;;;;;;;AAaN,AAAe,SAASqN,SAAT,CAAmB7G,QAAnB,EAA6BlI,CAA7B,EAAgC;IAC3CyE,wBAAF,EAA4ByD,QAA5B,EAAsC/H,IAAtC,CAA2C,UAACgB,KAAD,EAAQd,IAAR,EAAiB;QACpDC,QAAQN,EAAEK,IAAF,CAAd;QACIiO,SAAS/E,SAASjJ,KAAT,CAAb;QACI,CAACgO,MAAL,EAAa;eACF9D,eAAelK,KAAf,EAAsBN,CAAtB,CAAT;eACSM,KAAT,EAAgBN,CAAhB,EAAmBsO,MAAnB;;;;QAIEA,SAAS,CAAb,EAAgB;YACR5M,MAAN;KADF,MAEO;;0BAEepB,KAApB,EAA2BN,CAA3B,EAA8BsO,MAA9B;;GAbJ;;SAiBOtO,CAAP;;;ACrGa,SAASgP,YAAT,CAAsB9G,QAAtB,EAAgClI,CAAhC,EAA+C;MAAZiP,KAAY,uEAAJ,EAAI;;IAC1DtK,eAAF,EAAmBuD,QAAnB,EAA6B/H,IAA7B,CAAkC,UAACgB,KAAD,EAAQ+N,MAAR,EAAmB;QAC7CC,UAAUnP,EAAEkP,MAAF,CAAhB;;;;;QAKIlP,EAAEmP,OAAF,EAAWjH,QAAX,EAAqBkH,OAArB,CAA6B,GAA7B,EAAkC7M,MAAlC,KAA6C,CAAjD,EAAoD;aAC3C4M,QAAQzN,MAAR,EAAP;;;;QAIE2J,gBAAgBrL,EAAEkP,MAAF,EAAUpG,IAAV,EAAhB,MAAsCmG,KAA1C,EAAiD;aACxCE,QAAQzN,MAAR,EAAP;;;;;QAKE2H,UAAUrJ,EAAEkP,MAAF,CAAV,IAAuB,CAA3B,EAA8B;aACrBC,QAAQzN,MAAR,EAAP;;;WAGKyN,OAAP;GArBF;;SAwBOnP,CAAP;;;AC5BF;;;AAEA,AAAe,SAASqP,eAAT,CAAyBjH,OAAzB,EAAkCpI,CAAlC,EAAqC;;;;MAI9CyH,cAAczH,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;MACIyH,cAAczH,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;;SAEOA,CAAP;;;ACTF,SAASsP,UAAT,CAAoBtP,CAApB,EAAuBuP,OAAvB,EAAgC/O,IAAhC,EAAsCgP,QAAtC,EAAgD;UACxChP,IAAN,QAAegP,QAAf,EAAyBrP,IAAzB,CAA8B,UAACC,CAAD,EAAIC,IAAJ,EAAa;QACnCb,MAAMa,KAAKW,OAAL,CAAaR,IAAb,CAAZ;QACMiP,cAAc/P,IAAIjB,OAAJ,CAAY8Q,OAAZ,EAAqB/P,GAArB,CAApB;;SAEKwB,OAAL,CAAaR,IAAb,IAAqBiP,WAArB;GAJF;;;AAQF,AAAe,SAASC,iBAAT,CAA2BF,QAA3B,EAAqCxP,CAArC,EAAwCR,GAAxC,EAA6C;GACzD,MAAD,EAAS,KAAT,EAAgByB,OAAhB,CAAwB;WAAQqO,WAAWtP,CAAX,EAAcR,GAAd,EAAmBgB,IAAnB,EAAyBgP,QAAzB,CAAR;GAAxB;;SAEOA,QAAP;;;ACbK,SAAS3F,UAAT,CAAoBf,IAApB,EAA0B;SACxBA,KAAKC,IAAL,GACKvF,OADL,CACa,MADb,EACqB,GADrB,EAEKjB,MAFZ;;;;;;AAQF,AAAO,SAASwL,WAAT,CAAqBzN,KAArB,EAA4B;MAC3BqP,kBAAkB9F,WAAWvJ,MAAMwI,IAAN,EAAX,CAAxB;;MAEM8G,WAAWtP,MAAMiB,IAAN,CAAW,GAAX,EAAgBuH,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,CACb9P,CADa,EAEb+P,SAFa,EAGbC,WAHa,EAKb;MADAjB,SACA,uEADY,IACZ;;MACMkB,aAAaF,UAAUtO,MAAV,CAAiB;WAAQuO,YAAYE,OAAZ,CAAoBC,IAApB,MAA8B,CAAC,CAAvC;GAAjB,CAAnB;;;;;;;;UAEWA,IAHX;;UAIQ/O,OAAO,MAAb;UACMb,QAAQ,OAAd;;UAEM6P,QAAQpQ,YAAUoB,IAAV,UAAmB+O,IAAnB,QAAd;;;;;UAKME,SACJD,MAAMnM,GAAN,CAAU,UAAC9C,KAAD,EAAQd,IAAR;eAAiBL,EAAEK,IAAF,EAAQG,IAAR,CAAaD,KAAb,CAAjB;OAAV,EACM+P,OADN,GAEM7O,MAFN,CAEa;eAAQqH,SAAS,EAAjB;OAFb,CADF;;;;;;UASIuH,OAAO9N,MAAP,KAAkB,CAAtB,EAAyB;YACnBgO,kBAAJ;;;YAGIxB,SAAJ,EAAe;sBACDyB,UAAUH,OAAO,CAAP,CAAV,EAAqBrQ,CAArB,CAAZ;SADF,MAEO;sBACOqQ,OAAO,CAAP,CAAZ;;;;aAGKE;;;;;sCA5BQN,UAAnB,4GAA+B;;;;;;;;;;;;;;;;;;;;;;SAiCxB,IAAP;;;AC3CF,SAASQ,UAAT,CAAoBnQ,KAApB,EAA2BoQ,WAA3B,EAAwC;;;MAGlCpQ,MAAMgC,QAAN,GAAiBC,MAAjB,GAA0BmO,WAA9B,EAA2C;WAClC,KAAP;;;MAGEC,cAAcrQ,KAAd,CAAJ,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;;;;AAMF,AAAe,SAASsQ,oBAAT,CACb5Q,CADa,EAEb6Q,SAFa,EAKb;MAFAH,WAEA,uEAFc,CAEd;MADAI,QACA,uEADW,IACX;;;;;;sCACuBD,SAAvB,4GAAkC;UAAvB3M,QAAuB;;UAC1BkM,QAAQpQ,EAAEkE,QAAF,CAAd;;;;UAIIkM,MAAM7N,MAAN,KAAiB,CAArB,EAAwB;YAChBjC,QAAQN,EAAEoQ,MAAM,CAAN,CAAF,CAAd;;YAEIK,WAAWnQ,KAAX,EAAkBoQ,WAAlB,CAAJ,EAAoC;cAC9BzO,gBAAJ;cACI6O,QAAJ,EAAc;sBACFxQ,MAAMwI,IAAN,EAAV;WADF,MAEO;sBACKxI,MAAMyQ,IAAN,EAAV;;;cAGE9O,OAAJ,EAAa;mBACJA,OAAP;;;;;;;;;;;;;;;;;;;;SAMD,IAAP;;;AChDF;AACA,AAAe,SAASuO,SAAT,CAAmB1H,IAAnB,EAAyB9I,CAAzB,EAA4B;;;MAGnCgR,YAAYhR,aAAW8I,IAAX,cAA0BA,IAA1B,EAAlB;SACOkI,cAAc,EAAd,GAAmBlI,IAAnB,GAA0BkI,SAAjC;;;ACLa,SAASL,aAAT,CAAuBrQ,KAAvB,EAA8B;MACrCwC,UAAUxC,MAAMwC,OAAN,GAAgBwN,OAAhB,EAAhB;MACMW,gBAAgBnO,QAAQvB,IAAR,CAAa,UAACoJ,MAAD,EAAY;QACvCtE,aAAgBsE,OAAO3J,OAAP,CAAekQ,KAA/B,SAAwCvG,OAAO3J,OAAP,CAAeoF,EAA7D;WACOC,WAAWnE,QAAX,CAAoB,SAApB,CAAP;GAFoB,CAAtB;;SAKO+O,kBAAkBE,SAAzB;;;ACPF;;;;AAIA,AAAe,SAASC,gBAAT,CAA0B9Q,KAA1B,EAAiC;SACvCA,MAAMwI,IAAN,GAAaC,IAAb,GAAoBxG,MAApB,IAA8B,GAArC;;;ACHa,SAAS8O,WAAT,CAAqBrR,CAArB,EAAwB;SAC9BA,EAAEuF,cAAF,EAAkBhD,MAAlB,GAA2B,CAAlC;;;ACHF;AACA,AAAO,IAAM+O,kBAAkB,wCAAxB;;;;AAIP,AAAO,IAAMC,eAAe,IAAInT,MAAJ,CAAW,aAAX,EAA0B,GAA1B,CAArB;AACP,AAYA,AASA;AACA,AAAO,IAAMoT,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,OAAOxT,IAAP,CAAY,GAAZ,CAAlB;AACA,IAAM0T,aAAa,qCAAnB;AACA,IAAMC,aAAa,wCAAnB;AACA,AAAO,IAAMC,oBACX,IAAI7T,MAAJ,OAAe2T,UAAf,WAA+BC,UAA/B,wBAA4DF,SAA5D,QAA0E,IAA1E,CADK;;;;;AAMP,AAAO,IAAMI,qBAAqB,gBAA3B;;AAEP,AAAO,IAAMC,oBACX,IAAI/T,MAAJ,CAAW,2BAAX,EAAwC,GAAxC,CADK;;ACxDP;;AAEA,AAAe,SAASgU,WAAT,CAAqBC,MAArB,EAA6B;SACnCA,OAAO7O,OAAP,CAAe8N,eAAf,EAAgC,IAAhC,EAAsCvI,IAAtC,EAAP;;;ACHa,SAASpH,OAAT,CAAe2Q,YAAf,EAA6B;iBAC3BA,aAAavJ,IAAb,EAAf;MACIwJ,SAASC,QAAT,CAAkBF,YAAlB,CAAJ,EAAqC;WAC5BA,YAAP;;;SAGK,IAAP;;;ACJF;;AAEA,AAAe,SAASG,QAAT,CAAkBC,GAAlB,QAA8B;MAAL1S,CAAK,QAALA,CAAK;;;MAEvC0S,IAAInQ,MAAJ,GAAa,IAAb,IAAqBmQ,IAAInQ,MAAJ,GAAa,CAAtC,EAAyC,OAAO,IAAP;;MAEnCoQ,UAAUnC,UAAUkC,GAAV,EAAe1S,CAAf,CAAhB;;;;MAIIuR,aAAahS,IAAb,CAAkBoT,OAAlB,CAAJ,EAAgC,OAAO,IAAP;;SAEzBA,QAAQ5J,IAAR,EAAP;;;ACfF;;;;AAIA,AASA,AAAO,SAAS6J,eAAT,CAAyBC,UAAzB,EAAqC;SACnC,CAACA,WAAWnJ,KAAX,CAAiBuI,iBAAjB,KAAuC,EAAxC,EACW5T,IADX,CACgB,GADhB,EAEWmF,OAFX,CAEmBoO,qBAFnB,EAE0C,GAF1C,EAGWpO,OAHX,CAGmBmO,sBAHnB,EAG2C,UAH3C,EAIWnO,OAJX,CAImBkO,oBAJnB,EAIyC,IAJzC,EAKW3I,IALX,EAAP;;;;;AAUF,AAAe,SAAS+J,kBAAT,CAA4BD,UAA5B,EAAwC;;MAEjDrB,eAAejS,IAAf,CAAoBsT,UAApB,KAAmCpB,gBAAgBlS,IAAhB,CAAqBsT,UAArB,CAAvC,EAAyE;iBAC1D9K,SAAS8K,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;;;AC1BF;;AACA,AAAe,SAASC,gBAAT,CACbhL,OADa,QASb;MANEpI,CAMF,QANEA,CAMF;mCALEqT,kBAKF;MALEA,kBAKF,yCALuB,IAKvB;wBAJEpE,KAIF;MAJEA,KAIF,8BAJU,EAIV;sBAHEzP,GAGF;MAHEA,GAGF,4BAHQ,EAGR;iCAFE8T,cAEF;MAFEA,cAEF,uCAFmB,IAEnB;;;;kBAGgBlL,OAAhB,EAAyBpI,CAAzB;;;;;MAKIsT,cAAJ,EAAoBrL,YAAYG,OAAZ,EAAqBpI,CAArB;;;;gBAINoI,OAAd,EAAuBpI,CAAvB;;;;;aAKWoI,OAAX,EAAoBpI,CAApB;;;eAGaoI,OAAb,EAAsBpI,CAAtB,EAAyBiP,KAAzB;;;oBAGkB7G,OAAlB,EAA2BpI,CAA3B,EAA8BR,GAA9B;;;kBAGgB4I,OAAhB;;;;;;MAMIkL,cAAJ,EAAoBvE,UAAU3G,OAAV,EAAmBpI,CAAnB,EAAsBqT,kBAAtB;;;cAGRjL,OAAZ,EAAqBpI,CAArB;;SAEOoI,OAAP;;;ACtDa,SAASmL,UAAT,CAAoBtE,KAApB,QAAuC;MAAVzP,GAAU,QAAVA,GAAU;MAALQ,CAAK,QAALA,CAAK;;;;MAGhDkS,mBAAmB3S,IAAnB,CAAwB0P,KAAxB,CAAJ,EAAoC;YAC1BuE,kBAAkBvE,KAAlB,EAAyBzP,GAAzB,CAAR;;;;;MAKEyP,MAAM1M,MAAN,GAAe,GAAnB,EAAwB;;QAEhBkR,KAAKzT,EAAE,IAAF,CAAX;QACIyT,GAAGlR,MAAH,KAAc,CAAlB,EAAqB;cACXkR,GAAG3K,IAAH,EAAR;;;;;SAKG0H,UAAUvB,KAAV,EAAiBjP,CAAjB,EAAoB+I,IAApB,EAAP;;;ACdF,SAAS2K,sBAAT,CAAgCC,UAAhC,EAA4C7K,IAA5C,EAAkD;;;;MAI5C6K,WAAWpR,MAAX,IAAqB,CAAzB,EAA4B;;;;;UAIpBqR,aAAaD,WAAWlL,MAAX,CAAkB,UAACC,GAAD,EAAMmL,SAAN,EAAoB;YACnDA,SAAJ,IAAiBnL,IAAImL,SAAJ,IAAiBnL,IAAImL,SAAJ,IAAiB,CAAlC,GAAsC,CAAvD;eACOnL,GAAP;OAFiB,EAGhB,EAHgB,CAAnB;;kCAME,iBAAgBkL,UAAhB,EACQnL,MADR,CACe,UAACC,GAAD,EAAMf,GAAN,EAAc;YAChBe,IAAI,CAAJ,IAASkL,WAAWjM,GAAX,CAAb,EAA8B;iBACrB,CAACA,GAAD,EAAMiM,WAAWjM,GAAX,CAAN,CAAP;;;eAGKe,GAAP;OANT,EAOU,CAAC,CAAD,EAAI,CAAJ,CAPV,CAVwB;;;;UASnBoL,OATmB;UASVC,SATU;;;;;;;UAuBtBA,aAAa,CAAb,IAAkBD,QAAQvR,MAAR,IAAkB,CAAxC,EAA2C;qBAC5BuG,KAAKqD,KAAL,CAAW2H,OAAX,CAAb;;;UAGIE,YAAY,CAACL,WAAW,CAAX,CAAD,EAAgBA,WAAWvJ,KAAX,CAAiB,CAAC,CAAlB,CAAhB,CAAlB;UACM6J,aAAaD,UAAUvL,MAAV,CAAiB,UAACC,GAAD,EAAM9K,GAAN;eAAc8K,IAAInG,MAAJ,GAAa3E,IAAI2E,MAAjB,GAA0BmG,GAA1B,GAAgC9K,GAA9C;OAAjB,EAAoE,EAApE,CAAnB;;UAEIqW,WAAW1R,MAAX,GAAoB,EAAxB,EAA4B;;aACnB0R;;;;;WAGFnL;;;;;;;SAGF,IAAP;;;AAGF,SAASoL,oBAAT,CAA8BP,UAA9B,EAA0CnU,GAA1C,EAA+C;;;;;;;mBAO5BE,IAAIC,KAAJ,CAAUH,GAAV,CAP4B;;MAOrCoN,IAPqC,cAOrCA,IAPqC;;MAQvCuH,cAAcvH,KAAKpJ,OAAL,CAAa2O,iBAAb,EAAgC,EAAhC,CAApB;;MAEMiC,YAAYT,WAAW,CAAX,EAAcnH,WAAd,GAA4BhJ,OAA5B,CAAoC,GAApC,EAAyC,EAAzC,CAAlB;MACM6Q,iBAAiBC,MAAMC,WAAN,CAAkBH,SAAlB,EAA6BD,WAA7B,CAAvB;;MAEIE,iBAAiB,GAAjB,IAAwBD,UAAU7R,MAAV,GAAmB,CAA/C,EAAkD;WACzCoR,WAAWvJ,KAAX,CAAiB,CAAjB,EAAoB/L,IAApB,CAAyB,EAAzB,CAAP;;;MAGImW,UAAUb,WAAWvJ,KAAX,CAAiB,CAAC,CAAlB,EAAqB,CAArB,EAAwBoC,WAAxB,GAAsChJ,OAAtC,CAA8C,GAA9C,EAAmD,EAAnD,CAAhB;MACMiR,eAAeH,MAAMC,WAAN,CAAkBC,OAAlB,EAA2BL,WAA3B,CAArB;;MAEIM,eAAe,GAAf,IAAsBD,QAAQjS,MAAR,IAAkB,CAA5C,EAA+C;WACtCoR,WAAWvJ,KAAX,CAAiB,CAAjB,EAAoB,CAAC,CAArB,EAAwB/L,IAAxB,CAA6B,EAA7B,CAAP;;;SAGK,IAAP;;;;;AAKF,AAAe,SAASmV,iBAAT,CAA2BvE,KAA3B,EAA4C;MAAVzP,GAAU,uEAAJ,EAAI;;;;MAGnDmU,aAAa1E,MAAM9C,KAAN,CAAY+F,kBAAZ,CAAnB;MACIyB,WAAWpR,MAAX,KAAsB,CAA1B,EAA6B;WACpB0M,KAAP;;;MAGEyF,WAAWhB,uBAAuBC,UAAvB,EAAmC1E,KAAnC,CAAf;MACIyF,QAAJ,EAAc,OAAOA,QAAP;;aAEHR,qBAAqBP,UAArB,EAAiCnU,GAAjC,CAAX;MACIkV,QAAJ,EAAc,OAAOA,QAAP;;;;SAIPzF,KAAP;;;AC3FF,IAAM0F,WAAW;UACPvC,WADO;kBAECwC,OAFD;OAGVnC,QAHU;kBAICK,kBAJD;WAKN+B,gBALM;SAMRtB;CANT,CAUA,AAEA,AACA,AACA,AACA,AACA,AACA,AACA;;ACfA;;;;;;;;;;;AAWA,AAAe,SAASuB,eAAT,CAAyB9U,CAAzB,EAA4B+U,IAA5B,EAAkC;;;;;;;MAO3CA,KAAK9O,uBAAT,EAAkC;QAC5BA,wBAAwBjG,CAAxB,CAAJ;;;MAGEwH,oBAAoBxH,CAApB,CAAJ;MACIiL,aAAajL,CAAb,EAAgB+U,KAAKnK,WAArB,CAAJ;MACMoK,gBAAgB7G,iBAAiBnO,CAAjB,CAAtB;;SAEOgV,aAAP;;;AC3BF,IAAMC,0BAA0B;eACjB;6BACc,IADd;iBAEE,IAFF;wBAGS;GAJQ;;;;;;;;;;;;;;;;;;;;;SAAA,yBA0BGF,IA1BH,EA0BS;QAA7B/U,CAA6B,QAA7BA,CAA6B;QAA1B+Q,IAA0B,QAA1BA,IAA0B;QAApB9B,KAAoB,QAApBA,KAAoB;QAAbzP,GAAa,QAAbA,GAAa;;wBACzB,KAAK0V,WAAjB,EAAiCH,IAAjC;;QAEI/U,KAAKmC,QAAQC,IAAR,CAAa2O,IAAb,CAAT;;;;QAII1Q,OAAO,KAAK8U,cAAL,CAAoBnV,CAApB,EAAuBiP,KAAvB,EAA8BzP,GAA9B,EAAmCuV,IAAnC,CAAX;;QAEI3D,iBAAiB/Q,IAAjB,CAAJ,EAA4B;aACnB,KAAK+U,kBAAL,CAAwB/U,IAAxB,EAA8BL,CAA9B,CAAP;;;;;;;;;;wCAKgB,iBAAgB+U,IAAhB,EAAsBtT,MAAtB,CAA6B;eAAKsT,KAAKM,CAAL,MAAY,IAAjB;OAA7B,CAAlB,4GAAuE;YAA5D1N,GAA4D;;aAChEA,GAAL,IAAY,KAAZ;YACIxF,QAAQC,IAAR,CAAa2O,IAAb,CAAJ;;eAEO,KAAKoE,cAAL,CAAoBnV,CAApB,EAAuBiP,KAAvB,EAA8BzP,GAA9B,EAAmCuV,IAAnC,CAAP;;YAEI3D,iBAAiB/Q,IAAjB,CAAJ,EAA4B;;;;;;;;;;;;;;;;;;;WAKvB,KAAK+U,kBAAL,CAAwB/U,IAAxB,EAA8BL,CAA9B,CAAP;GApD4B;;;;gBAAA,0BAwDfA,CAxDe,EAwDZiP,KAxDY,EAwDLzP,GAxDK,EAwDAuV,IAxDA,EAwDM;WAC3BF,iBACGC,gBAAgB9U,CAAhB,EAAmB+U,IAAnB,CADH,EAEL;UAAA;0BAEsBA,KAAK1B,kBAF3B;kBAAA;;KAFK,CAAP;GAzD4B;;;;;;oBAAA,8BAsEXhT,IAtEW,EAsELL,CAtEK,EAsEF;QACtB,CAACK,IAAL,EAAW;aACF,IAAP;;;WAGKgL,gBAAgBrL,EAAE+Q,IAAF,CAAO1Q,IAAP,CAAhB,CAAP;;;;;;;CA3EJ,CAqFA;;AC9FA;;;;;;;AAOA,AAAO,IAAMiV,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;QAArB1V,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbmW,SAAa,QAAbA,SAAa;;;;QAGzB1G,cAAJ;;YAEQa,gBAAgB9P,CAAhB,EAAmBsV,sBAAnB,EAA2CK,SAA3C,CAAR;QACI1G,KAAJ,EAAW,OAAOsE,WAAWtE,KAAX,EAAkB,EAAEzP,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;;YAIH4Q,qBAAqB5Q,CAArB,EAAwBwV,sBAAxB,CAAR;QACIvG,KAAJ,EAAW,OAAOsE,WAAWtE,KAAX,EAAkB,EAAEzP,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;YAGH8P,gBAAgB9P,CAAhB,EAAmBuV,oBAAnB,EAAyCI,SAAzC,CAAR;QACI1G,KAAJ,EAAW,OAAOsE,WAAWtE,KAAX,EAAkB,EAAEzP,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;YAGH4Q,qBAAqB5Q,CAArB,EAAwByV,oBAAxB,CAAR;QACIxG,KAAJ,EAAW,OAAOsE,WAAWtE,KAAX,EAAkB,EAAEzP,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;WAGJ,EAAP;;CAvBJ,CA2BA;;ACxCA;;;;;;AAMA,AAAO,IAAM4V,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;QAAhBjW,CAAgB,QAAhBA,CAAgB;QAAb2V,SAAa,QAAbA,SAAa;;QACpBtD,eAAJ;;;;aAISvC,gBAAgB9P,CAAhB,EAAmB4V,gBAAnB,EAAqCD,SAArC,CAAT;QACItD,UAAUA,OAAO9P,MAAP,GAAgBsT,iBAA9B,EAAiD;aACxCzD,YAAYC,MAAZ,CAAP;;;;aAIOzB,qBAAqB5Q,CAArB,EAAwB8V,gBAAxB,EAA0C,CAA1C,CAAT;QACIzD,UAAUA,OAAO9P,MAAP,GAAgBsT,iBAA9B,EAAiD;aACxCzD,YAAYC,MAAZ,CAAP;;;;;;;;;;wCAK8B2D,mBAAhC,4GAAqD;;;;;YAAzC9R,QAAyC;YAA/BgS,KAA+B;;YAC7C7V,OAAOL,EAAEkE,QAAF,CAAb;YACI7D,KAAKkC,MAAL,KAAgB,CAApB,EAAuB;cACfuG,OAAOzI,KAAKyI,IAAL,EAAb;cACIoN,MAAM3W,IAAN,CAAWuJ,IAAX,CAAJ,EAAsB;mBACbsJ,YAAYtJ,IAAZ,CAAP;;;;;;;;;;;;;;;;;;;WAKC,IAAP;;CA7BJ,CAiCA;;AC9CA;;;;AAIA,AAAO,IAAMqN,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,IAAIlY,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAFoC;;;;AAMpC,IAAIA,MAAJ,CAAW,6BAAX,EAA0C,GAA1C,CANoC;;AAQpC,IAAIA,MAAJ,iBAAyBiY,eAAzB,kBAAuD,GAAvD,CARoC,CAA/B;;ACrCP,IAAME,gCAAgC;SAAA,yBACL;QAArBvW,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbmW,SAAa,QAAbA,SAAa;;QACzBa,sBAAJ;;;;oBAIgB1G,gBAAgB9P,CAAhB,EAAmBmW,wBAAnB,EAA6CR,SAA7C,EAAwD,KAAxD,CAAhB;QACIa,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;;oBAIH5F,qBAAqB5Q,CAArB,EAAwBoW,wBAAxB,CAAhB;QACII,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;oBAGHlL,eAAe9L,GAAf,EAAoB8W,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,IAAIzY,MAAJ,CAAWwY,8BAA8BvY,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAMyY,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,IAAI3Y,MAAJ,CAAW0Y,8BAA8BzY,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAM2Y,SAAS,gBAAf;AACP,AAAO,IAAMC,SAAS,kBAAf;;AC3CP,SAASC,MAAT,CAAgB5W,KAAhB,EAAuB;UACXA,MAAME,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCF,MAAME,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;;AAIF,AAAO,SAAS2W,aAAT,CAAuB3X,GAAvB,EAA4B;QAC3BA,IAAIuJ,IAAJ,EAAN;MACIO,QAAQ,CAAZ;;MAEIuN,iCAAiCtX,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;MAGEuX,iCAAiCxX,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;;;MAKEwX,OAAOzX,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;MAGEyX,OAAO1X,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;;;SAKK8J,KAAP;;;;AAIF,AAAO,SAAS8N,SAAT,CAAmBvP,IAAnB,EAAyB;MAC1BA,KAAKrH,IAAL,CAAU,KAAV,CAAJ,EAAsB;WACb,CAAP;;;SAGK,CAAP;;;;;AAKF,AAAO,SAAS6W,cAAT,CAAwBxP,IAAxB,EAA8B;MAC/ByB,QAAQ,CAAZ;MACMgO,aAAazP,KAAK/E,OAAL,CAAa,QAAb,EAAuBsL,KAAvB,EAAnB;;MAEIkJ,WAAW/U,MAAX,KAAsB,CAA1B,EAA6B;aAClB,EAAT;;;MAGIM,UAAUgF,KAAK8C,MAAL,EAAhB;MACI4M,iBAAJ;MACI1U,QAAQN,MAAR,KAAmB,CAAvB,EAA0B;eACbM,QAAQ8H,MAAR,EAAX;;;GAGD9H,OAAD,EAAU0U,QAAV,EAAoBtW,OAApB,CAA4B,UAACX,KAAD,EAAW;QACjC4E,iBAAe3F,IAAf,CAAoB2X,OAAO5W,KAAP,CAApB,CAAJ,EAAwC;eAC7B,EAAT;;GAFJ;;SAMOgJ,KAAP;;;;;AAKF,AAAO,SAASkO,cAAT,CAAwB3P,IAAxB,EAA8B;MAC/ByB,QAAQ,CAAZ;MACMqE,WAAW9F,KAAKnB,IAAL,EAAjB;MACMG,UAAU8G,SAASpP,GAAT,CAAa,CAAb,CAAhB;;MAEIsI,WAAWA,QAAQnE,OAAR,KAAoB,YAAnC,EAAiD;aACtC,EAAT;;;MAGEwC,iBAAe3F,IAAf,CAAoB2X,OAAOvJ,QAAP,CAApB,CAAJ,EAA2C;aAChC,EAAT;;;SAGKrE,KAAP;;;AAGF,AAAO,SAASmO,iBAAT,CAA2B5P,IAA3B,EAAiC;MAClCyB,QAAQ,CAAZ;;MAEM/F,QAAQiG,WAAW3B,KAAKrH,IAAL,CAAU,OAAV,CAAX,CAAd;MACMsH,SAAS0B,WAAW3B,KAAKrH,IAAL,CAAU,QAAV,CAAX,CAAf;MACM8C,MAAMuE,KAAKrH,IAAL,CAAU,KAAV,CAAZ;;;MAGI+C,SAASA,SAAS,EAAtB,EAA0B;aACf,EAAT;;;;MAIEuE,UAAUA,UAAU,EAAxB,EAA4B;aACjB,EAAT;;;MAGEvE,SAASuE,MAAT,IAAmB,CAACxE,IAAIpB,QAAJ,CAAa,QAAb,CAAxB,EAAgD;QACxCwV,OAAOnU,QAAQuE,MAArB;QACI4P,OAAO,IAAX,EAAiB;;eACN,GAAT;KADF,MAEO;eACI1N,KAAK2N,KAAL,CAAWD,OAAO,IAAlB,CAAT;;;;SAIGpO,KAAP;;;AAGF,AAAO,SAASsO,eAAT,CAAyBC,KAAzB,EAAgC1W,KAAhC,EAAuC;SACpC0W,MAAMtV,MAAN,GAAe,CAAhB,GAAqBpB,KAA5B;;;ACxGF;;;;;;;;AAQA,IAAM2W,+BAA+B;SAAA,yBACA;QAAzB9X,CAAyB,QAAzBA,CAAyB;QAAtBiC,OAAsB,QAAtBA,OAAsB;QAAb0T,SAAa,QAAbA,SAAa;;QAC7BoC,iBAAJ;;;;;;QAMMC,WACJlI,gBACE9P,CADF,EAEE0W,wBAFF,EAGEf,SAHF,EAIE,KAJF,CADF;;QAQIqC,QAAJ,EAAc;iBACDpD,QAAWoD,QAAX,CAAX;;UAEID,QAAJ,EAAc,OAAOA,QAAP;;;;;;QAMVE,OAAOjY,EAAE,KAAF,EAASiC,OAAT,EAAkBqO,OAAlB,EAAb;QACM4H,YAAY,EAAlB;;SAEKjX,OAAL,CAAa,UAACF,GAAD,EAAMI,KAAN,EAAgB;UACrB0G,OAAO7H,EAAEe,GAAF,CAAb;UACMuC,MAAMuE,KAAKrH,IAAL,CAAU,KAAV,CAAZ;;UAEI,CAAC8C,GAAL,EAAU;;UAENgG,QAAQ6N,cAAc7T,GAAd,CAAZ;eACS8T,UAAUvP,IAAV,CAAT;eACSwP,eAAexP,IAAf,CAAT;eACS2P,eAAe3P,IAAf,CAAT;eACS4P,kBAAkB5P,IAAlB,CAAT;eACS+P,gBAAgBK,IAAhB,EAAsB9W,KAAtB,CAAT;;gBAEUmC,GAAV,IAAiBgG,KAAjB;KAbF;;gCAiBE,iBAAgB4O,SAAhB,EAA2BzP,MAA3B,CAAkC,UAACC,GAAD,EAAMf,GAAN;aAChCuQ,UAAUvQ,GAAV,IAAiBe,IAAI,CAAJ,CAAjB,GAA0B,CAACf,GAAD,EAAMuQ,UAAUvQ,GAAV,CAAN,CAA1B,GAAkDe,GADlB;KAAlC,EAEE,CAAC,IAAD,EAAO,CAAP,CAFF,CA5C+B;;;;QA2C1ByP,MA3C0B;QA2ClB3K,QA3CkB;;;QAgD7BA,WAAW,CAAf,EAAkB;iBACLoH,QAAWuD,MAAX,CAAX;;UAEIJ,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;wCAKOpB,wBAAvB,4GAAiD;YAAtCzS,QAAsC;;YACzC5D,QAAQN,EAAEkE,QAAF,EAAYkK,KAAZ,EAAd;YACM9K,MAAMhD,MAAME,IAAN,CAAW,KAAX,CAAZ;YACI8C,GAAJ,EAAS;qBACIsR,QAAWtR,GAAX,CAAX;cACIyU,QAAJ,EAAc,OAAOA,QAAP;;;YAGVK,OAAO9X,MAAME,IAAN,CAAW,MAAX,CAAb;YACI4X,IAAJ,EAAU;qBACGxD,QAAWwD,IAAX,CAAX;cACIL,QAAJ,EAAc,OAAOA,QAAP;;;YAGVxX,QAAQD,MAAME,IAAN,CAAW,OAAX,CAAd;YACID,KAAJ,EAAW;qBACEqU,QAAWrU,KAAX,CAAX;cACIwX,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;;;;;;;;;WAIX,IAAP;;CA9EJ,CAkFA;;AC3Ge,SAASM,eAAT,CAAyB/O,KAAzB,EAAgCgP,UAAhC,EAA4CF,IAA5C,EAAkD;;;;;;MAM3D9O,QAAQ,CAAZ,EAAe;QACPiP,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;WACOrP,QAAQsP,YAAf;;;SAGK,CAAP;;;ACnBa,SAASC,aAAT,CAAuBjJ,QAAvB,EAAiC3D,OAAjC,EAA0C;;;;;MAKnD3C,QAAQ,CAAZ;;MAEIwC,YAAYvM,IAAZ,CAAiBqQ,SAAS7G,IAAT,EAAjB,CAAJ,EAAuC;QAC/B+P,gBAAgB/Q,SAAS6H,QAAT,EAAmB,EAAnB,CAAtB;;;;QAIIkJ,gBAAgB,CAApB,EAAuB;cACb,CAAC,EAAT;KADF,MAEO;cACG9O,KAAKE,GAAL,CAAS,CAAT,EAAY,KAAK4O,aAAjB,CAAR;;;;;;QAME7M,WAAWA,WAAW6M,aAA1B,EAAyC;eAC9B,EAAT;;;;SAIGxP,KAAP;;;AC5Ba,SAASyP,eAAT,CAAyB9M,OAAzB,EAAkC+M,IAAlC,EAAwC;;;;MAIjD/M,WAAW,CAAC+M,IAAhB,EAAsB;WACb,EAAP;;;SAGK,CAAP;;;ACRK,IAAMC,aAAW,IAAjB;;;;AAIP,AAAO,IAAMzT,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,IAAIrH,MAAJ,CAAWoH,wBAAsBnH,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;;;AAKP,AAAO,IAAM6a,sBAAoB,IAAI9a,MAAJ,CAAW,4CAAX,EAAyD,GAAzD,CAA1B;;;;AAIP,AAAO,IAAM+a,qBAAmB,IAAI/a,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAzB;;;;AAIP,AAAO,IAAMgb,sBAAoB,IAAIhb,MAAJ,CAAW,yBAAX,EAAsC,GAAtC,CAA1B,CAEP;;ACjCe,SAASib,oBAAT,CAA8BjB,IAA9B,EAAoC;;MAE7C3S,2BAAyBlG,IAAzB,CAA8B6Y,IAA9B,CAAJ,EAAyC;WAChC,CAAC,EAAR;;;SAGK,CAAP;;;ACAF,SAASkB,SAAT,CAAiBC,KAAjB,EAAwB;UACZA,MAAM/Y,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuC+Y,MAAM/Y,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;AAGF,AAAe,SAAS6W,gBAAT,CAAwBkC,KAAxB,EAA+B;;;;MAIxC1W,UAAU0W,MAAM5O,MAAN,EAAd;MACI6O,gBAAgB,KAApB;MACIC,gBAAgB,KAApB;MACInQ,QAAQ,CAAZ;;cAEW5L,MAAM,CAAN,EAAS,CAAT,CAAX,EAAwBuD,OAAxB,CAAgC,YAAM;QAChC4B,QAAQN,MAAR,KAAmB,CAAvB,EAA0B;;;;QAIpBmX,aAAaJ,UAAQzW,OAAR,EAAiB,GAAjB,CAAnB;;;;QAII,CAAC2W,aAAD,IAAkB9T,QAAQnG,IAAR,CAAama,UAAb,CAAtB,EAAgD;sBAC9B,IAAhB;eACS,EAAT;;;;;;QAME,CAACD,aAAD,IAAkBnU,kBAAkB/F,IAAlB,CAAuBma,UAAvB,CAAlB,IACEjU,2BAAyBlG,IAAzB,CAA8Bma,UAA9B,CADN,EACiD;UAC3C,CAACtU,kBAAkB7F,IAAlB,CAAuBma,UAAvB,CAAL,EAAyC;wBACvB,IAAhB;iBACS,EAAT;;;;cAIM7W,QAAQ8H,MAAR,EAAV;GAzBF;;SA4BOrB,KAAP;;;AC/Ca,SAASqQ,aAAT,CAAuBC,QAAvB,EAAiC;;;MAG1CR,oBAAkB7Z,IAAlB,CAAuBqa,QAAvB,CAAJ,EAAsC;WAC7B,CAAC,GAAR;;;SAGK,CAAP;;;ACFa,SAASC,WAAT,CACbzB,IADa,EAEbE,UAFa,EAGbwB,OAHa,EAIbra,SAJa,EAKbmQ,QALa,EAMbmK,YANa,EAOb;;MAEIA,aAAaxY,IAAb,CAAkB;WAAO6W,SAAS5Y,GAAhB;GAAlB,MAA2C2R,SAA/C,EAA0D;WACjD,KAAP;;;;;MAKE,CAACiH,IAAD,IAASA,SAASE,UAAlB,IAAgCF,SAAS0B,OAA7C,EAAsD;WAC7C,KAAP;;;MAGMhc,QAZR,GAYqB2B,SAZrB,CAYQ3B,QAZR;;mBAa+B4B,IAAIC,KAAJ,CAAUyY,IAAV,CAb/B;;MAakB4B,QAblB,cAaQlc,QAbR;;;;MAgBIkc,aAAalc,QAAjB,EAA2B;WAClB,KAAP;;;;;MAKImc,WAAW7B,KAAK5U,OAAL,CAAasW,OAAb,EAAsB,EAAtB,CAAjB;MACI,CAACb,WAAS1Z,IAAT,CAAc0a,QAAd,CAAL,EAA8B;WACrB,KAAP;;;;;MAKExU,2BAAyBlG,IAAzB,CAA8BqQ,QAA9B,CAAJ,EAA6C;WACpC,KAAP;;;;MAIEA,SAASrN,MAAT,GAAkB,EAAtB,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;ACpDa,SAAS2X,YAAT,CAAsB9B,IAAtB,EAA4B+B,SAA5B,EAAuC;;;;;MAKhD,CAACA,UAAU5a,IAAV,CAAe6Y,IAAf,CAAL,EAA2B;WAClB,CAAC,EAAR;;;SAGK,CAAP;;;ACPa,SAASgC,iBAAT,CAA2BR,QAA3B,EAAqC;;MAE9CV,oBAAkB3Z,IAAlB,CAAuBqa,QAAvB,CAAJ,EAAsC;WAC7B,EAAP;;;SAGK,CAAP;;;ACHa,SAASS,aAAT,CAAuBT,QAAvB,EAAiC;;MAE1CT,mBAAiB5Z,IAAjB,CAAsBqa,QAAtB,CAAJ,EAAqC;;;;;QAK/BV,oBAAkB3Z,IAAlB,CAAuBqa,QAAvB,CAAJ,EAAsC;aAC7B,CAAC,EAAR;;;;SAIG,CAAP;;;ACKK,SAASU,aAAT,CAAuBR,OAAvB,EAAgC;SAC9B,IAAI1b,MAAJ,OAAe0b,OAAf,EAA0B,GAA1B,CAAP;;;AAGF,SAASR,OAAT,CAAiBC,KAAjB,EAAwB3J,QAAxB,EAAkC;UACtBA,YAAY2J,MAAMzQ,IAAN,EAAtB,WAAsCyQ,MAAM/Y,IAAN,CAAW,OAAX,KAAuB,EAA7D,WAAmE+Y,MAAM/Y,IAAN,CAAW,IAAX,KAAoB,EAAvF;;;AAGF,AAAe,SAAS+Z,UAAT,OAOZ;MANDC,KAMC,QANDA,KAMC;MALDlC,UAKC,QALDA,UAKC;MAJDwB,OAIC,QAJDA,OAIC;MAHDra,SAGC,QAHDA,SAGC;MAFDO,CAEC,QAFDA,CAEC;+BADD+Z,YACC;MADDA,YACC,qCADc,EACd;;cACWta,aAAaC,IAAIC,KAAJ,CAAU2Y,UAAV,CAAzB;MACM6B,YAAYG,cAAcR,OAAd,CAAlB;MACMd,OAAO3H,YAAYrR,CAAZ,CAAb;;;;;;;;;MASMya,cAAcD,MAAM/R,MAAN,CAAa,UAACiS,aAAD,EAAgBC,IAAhB,EAAyB;;;;QAIlDvC,OAAOlM,aAAayO,KAAK3Z,OAAL,CAAaoX,IAA1B,CAAb;QACMmB,QAAQvZ,EAAE2a,IAAF,CAAd;QACM/K,WAAW2J,MAAMzQ,IAAN,EAAjB;;QAEI,CAAC+Q,YAAYzB,IAAZ,EAAkBE,UAAlB,EAA8BwB,OAA9B,EAAuCra,SAAvC,EAAkDmQ,QAAlD,EAA4DmK,YAA5D,CAAL,EAAgF;aACvEW,aAAP;;;;QAIE,CAACA,cAActC,IAAd,CAAL,EAA0B;oBACVA,IAAd,IAAsB;eACb,CADa;0BAAA;;OAAtB;KADF,MAMO;oBACSA,IAAd,EAAoBxI,QAApB,GAAkC8K,cAActC,IAAd,EAAoBxI,QAAtD,SAAkEA,QAAlE;;;QAGIgL,eAAeF,cAActC,IAAd,CAArB;QACMwB,WAAWN,QAAQC,KAAR,EAAe3J,QAAf,CAAjB;QACM3D,UAAUF,eAAeqM,IAAf,CAAhB;;QAEI9O,QAAQ4Q,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,gBAAgB9M,OAAhB,EAAyB+M,IAAzB,CAAT;aACSH,cAAcjJ,QAAd,EAAwB3D,OAAxB,CAAT;aACSoM,gBAAgB/O,KAAhB,EAAuBgP,UAAvB,EAAmCF,IAAnC,CAAT;;iBAEa9O,KAAb,GAAqBA,KAArB;;WAEOoR,aAAP;GAvCkB,EAwCjB,EAxCiB,CAApB;;SA0CO,iBAAgBD,WAAhB,EAA6BlY,MAA7B,KAAwC,CAAxC,GAA4C,IAA5C,GAAmDkY,WAA1D;;;AClFF;;AAEA,IAAMI,8BAA8B;SAAA,yBACgB;QAAxC7a,CAAwC,QAAxCA,CAAwC;QAArCR,GAAqC,QAArCA,GAAqC;QAAhCC,SAAgC,QAAhCA,SAAgC;iCAArBsa,YAAqB;QAArBA,YAAqB,qCAAN,EAAM;;gBACpCta,aAAaC,IAAIC,KAAJ,CAAUH,GAAV,CAAzB;;QAEM8Y,aAAapM,aAAa1M,GAAb,CAAnB;QACMsa,UAAUrN,eAAejN,GAAf,EAAoBC,SAApB,CAAhB;;QAEM+a,QAAQxa,EAAE,SAAF,EAAasQ,OAAb,EAAd;;QAEMwK,cAAcP,WAAW;kBAAA;4BAAA;sBAAA;0BAAA;UAAA;;KAAX,CAApB;;;QAUI,CAACO,WAAL,EAAkB,OAAO,IAAP;;;;QAIZC,UAAU,iBAAgBD,WAAhB,EAA6BrS,MAA7B,CAAoC,UAACC,GAAD,EAAMiS,IAAN,EAAe;UAC3DK,aAAaF,YAAYH,IAAZ,CAAnB;aACOK,WAAW1R,KAAX,GAAmBZ,IAAIY,KAAvB,GAA+B0R,UAA/B,GAA4CtS,GAAnD;KAFc,EAGb,EAAEY,OAAO,CAAC,GAAV,EAHa,CAAhB;;;;QAOIyR,QAAQzR,KAAR,IAAiB,EAArB,EAAyB;aAChByR,QAAQ3C,IAAf;;;WAGK,IAAP;;CAlCJ,CAuCA;;AClDO,IAAM6C,2BAA2B,CACtC,QADsC,CAAjC;;ACKP,SAASC,WAAT,CAAqB1b,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;YAEG0b,YAAY1b,GAAZ;GAFV;;;AAMF,IAAM2b,sBAAsB;SAAA,yBACK;QAArBnb,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAbmW,SAAa,QAAbA,SAAa;;QACvByF,aAAapb,EAAE,qBAAF,CAAnB;QACIob,WAAW7Y,MAAX,KAAsB,CAA1B,EAA6B;UACrB6V,OAAOgD,WAAW5a,IAAX,CAAgB,MAAhB,CAAb;UACI4X,IAAJ,EAAU;eACDrW,OAAOqW,IAAP,CAAP;;;;QAIEiD,UAAUvL,gBAAgB9P,CAAhB,EAAmBib,wBAAnB,EAA6CtF,SAA7C,CAAhB;QACI0F,OAAJ,EAAa;aACJtZ,OAAOsZ,OAAP,CAAP;;;WAGKtZ,OAAOvC,GAAP,CAAP;;CAfJ,CAoBA;;ACtCO,IAAM8b,yBAAyB,CACpC,gBADoC,EAEpC,qBAFoC,CAA/B;;ACSA,SAAS3Z,OAAT,CAAeM,OAAf,EAAwBjC,CAAxB,EAA4C;MAAjBub,SAAiB,uEAAL,GAAK;;YACvCtZ,QAAQuB,OAAR,CAAgB,UAAhB,EAA4B,GAA5B,EAAiCuF,IAAjC,EAAV;SACOyS,UAAUvZ,OAAV,EAAmBsZ,SAAnB,EAA8B,EAAEE,SAAS,UAAX,EAA9B,CAAP;;;AAGF,IAAMC,0BAA0B;SAAA,yBACK;QAAzB1b,CAAyB,QAAzBA,CAAyB;QAAtBiC,OAAsB,QAAtBA,OAAsB;QAAb0T,SAAa,QAAbA,SAAa;;QAC3BgG,UAAU7L,gBAAgB9P,CAAhB,EAAmBsb,sBAAnB,EAA2C3F,SAA3C,CAAhB;QACIgG,OAAJ,EAAa;aACJha,QAAM6O,UAAUmL,OAAV,EAAmB3b,CAAnB,CAAN,CAAP;;;QAGIub,YAAY,GAAlB;QACMK,eAAe3Z,QAAQmI,KAAR,CAAc,CAAd,EAAiBmR,YAAY,CAA7B,CAArB;WACO5Z,QAAM3B,EAAE4b,YAAF,EAAgB9S,IAAhB,EAAN,EAA8B9I,CAA9B,EAAiCub,SAAjC,CAAP;;CATJ,CAaA;;ACvBA,IAAMM,4BAA4B;SAAA,yBACX;QAAX5Z,OAAW,QAAXA,OAAW;;QACbjC,IAAImC,QAAQC,IAAR,CAAaH,OAAb,CAAV;;QAEM6G,OAAOuC,gBAAgBrL,EAAE,KAAF,EAASoO,KAAT,GAAiBtF,IAAjB,EAAhB,CAAb;WACOA,KAAKqD,KAAL,CAAW,IAAX,EAAiB5J,MAAxB;;CALJ,CASA;;ACCA,IAAMuZ,mBAAmB;;UAEf,GAFe;SAGhBpG,sBAAsBqG,OAHN;kBAIPxF,8BAA8BwF,OAJvB;UAKf9F,uBAAuB8F,OALR;WAMd9G,wBAAwB8G,OAAxB,CAAgCC,IAAhC,CAAqC/G,uBAArC,CANc;kBAOP6C,6BAA6BiE,OAPtB;OAQlBtF,oBAAoBsF,OARF;iBASRlB,4BAA4BkB,OATpB;kBAUPZ,oBAAoBY,OAVb;WAWdL,wBAAwBK,OAXV;cAYXF,0BAA0BE,OAZf;aAaZ;QAAG9M,KAAH,QAAGA,KAAH;WAAegN,gBAAgBC,YAAhB,CAA6BjN,KAA7B,CAAf;GAbY;;SAAA,mBAefzQ,OAfe,EAeN;QACPuS,IADO,GACEvS,OADF,CACPuS,IADO;;;QAGXA,IAAJ,EAAU;UACF/Q,IAAImC,QAAQC,IAAR,CAAa2O,IAAb,CAAV;cACQ/Q,CAAR,GAAYA,CAAZ;;;QAGIiP,QAAQ,KAAKA,KAAL,CAAWzQ,OAAX,CAAd;QACM2d,iBAAiB,KAAKA,cAAL,CAAoB3d,OAApB,CAAvB;QACM6T,SAAS,KAAKA,MAAL,CAAY7T,OAAZ,CAAf;QACMyD,UAAU,KAAKA,OAAL,cAAkBzD,OAAlB,IAA2ByQ,YAA3B,IAAhB;QACMmN,iBAAiB,KAAKA,cAAL,cAAyB5d,OAAzB,IAAkCyD,gBAAlC,IAAvB;QACMyQ,MAAM,KAAKA,GAAL,cAAclU,OAAd,IAAuByD,gBAAvB,IAAZ;QACMoa,gBAAgB,KAAKA,aAAL,CAAmB7d,OAAnB,CAAtB;QACMmd,UAAU,KAAKA,OAAL,cAAkBnd,OAAlB,IAA2ByD,gBAA3B,IAAhB;QACMqa,aAAa,KAAKA,UAAL,cAAqB9d,OAArB,IAA8ByD,gBAA9B,IAAnB;QACMsa,YAAY,KAAKA,SAAL,CAAe,EAAEtN,YAAF,EAAf,CAAlB;;0BACwB,KAAKuN,cAAL,CAAoBhe,OAApB,CAlBT;;QAkBPgB,GAlBO,mBAkBPA,GAlBO;QAkBFid,MAlBE,mBAkBFA,MAlBE;;;WAoBR;kBAAA;oBAAA;sBAGWN,kBAAkB,IAH7B;cAAA;oCAAA;sBAAA;kCAAA;cAAA;oBAAA;sBAAA;4BAAA;;KAAP;;CAnCJ,CAoDA;;AC7De,SAASO,YAAT,CAAsBld,GAAtB,EAA2BC,SAA3B,EAAsC;cACvCA,aAAaC,IAAIC,KAAJ,CAAUH,GAAV,CAAzB;mBACqBC,SAF8B;MAE3C3B,QAF2C,cAE3CA,QAF2C;;MAG7C6e,aAAa7e,SAASqO,KAAT,CAAe,GAAf,EAAoB/B,KAApB,CAA0B,CAAC,CAA3B,EAA8B/L,IAA9B,CAAmC,GAAnC,CAAnB;;SAEOuF,WAAW9F,QAAX,KAAwB8F,WAAW+Y,UAAX,CAAxB,IAAkDb,gBAAzD;;;ACJF;AACA,AAAO,SAASc,gBAAT,CAA0BpN,QAA1B,EAAoCxP,CAApC,QAAkD;MAAT2B,KAAS,QAATA,KAAS;;MACnD,CAACA,KAAL,EAAY,OAAO6N,QAAP;;IAEV7N,MAAMtD,IAAN,CAAW,GAAX,CAAF,EAAmBmR,QAAnB,EAA6B9N,MAA7B;;SAEO8N,QAAP;;;;AAIF,AAAO,SAASqN,iBAAT,CAA2BrN,QAA3B,EAAqCxP,CAArC,SAAwD;MAAd8c,UAAc,SAAdA,UAAc;;MACzD,CAACA,UAAL,EAAiB,OAAOtN,QAAP;;mBAEDsN,UAAhB,EAA4B7b,OAA5B,CAAoC,UAAC0G,GAAD,EAAS;QACrCoV,WAAW/c,EAAE2H,GAAF,EAAO6H,QAAP,CAAjB;QACMjP,QAAQuc,WAAWnV,GAAX,CAAd;;;QAGI,OAAOpH,KAAP,KAAiB,QAArB,EAA+B;eACpBJ,IAAT,CAAc,UAACgB,KAAD,EAAQd,IAAR,EAAiB;sBACfL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0B8c,WAAWnV,GAAX,CAA1B;OADF;KADF,MAIO,IAAI,OAAOpH,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;;SAqBOyN,QAAP;;;AAGF,SAASwN,oBAAT,CAA8Bhd,CAA9B,EAAiC6Q,SAAjC,EAA4C;SACnCA,UAAUtP,IAAV,CAAe,UAAC2C,QAAD,EAAc;QAC9B+Y,MAAMC,OAAN,CAAchZ,QAAd,CAAJ,EAA6B;qCACTA,QADS;;UACpBiZ,CADoB;UACjB3c,IADiB;;aAEpBR,EAAEmd,CAAF,EAAK5a,MAAL,KAAgB,CAAhB,IAAqBvC,EAAEmd,CAAF,EAAK3c,IAAL,CAAUA,IAAV,CAArB,IAAwCR,EAAEmd,CAAF,EAAK3c,IAAL,CAAUA,IAAV,EAAgBuI,IAAhB,OAA2B,EAA1E;;;WAGK/I,EAAEkE,QAAF,EAAY3B,MAAZ,KAAuB,CAAvB,IAA4BvC,EAAEkE,QAAF,EAAY4E,IAAZ,GAAmBC,IAAnB,OAA8B,EAAjE;GANK,CAAP;;;AAUF,AAAO,SAASqU,MAAT,CAAgBrI,IAAhB,EAAsB;MACnB/U,CADmB,GAC8B+U,IAD9B,CACnB/U,CADmB;MAChBoB,IADgB,GAC8B2T,IAD9B,CAChB3T,IADgB;MACVic,cADU,GAC8BtI,IAD9B,CACVsI,cADU;0BAC8BtI,IAD9B,CACMuI,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,CASR/J,cATQ;MASRA,cATQ,yCASS,IATT;;;MAWrBiK,mBAAmBP,qBAAqBhd,CAArB,EAAwB6Q,SAAxB,CAAzB;;MAEI,CAAC0M,gBAAL,EAAuB,OAAO,IAAP;;;;;;;;MAQnBD,WAAJ,EAAiB;QACX9N,WAAWxP,EAAEud,gBAAF,CAAf;;;aAGSC,IAAT,CAAcxd,EAAE,aAAF,CAAd;eACWwP,SAAS7E,MAAT,EAAX;;eAEWkS,kBAAkBrN,QAAlB,EAA4BxP,CAA5B,EAA+Bqd,cAA/B,CAAX;eACWT,iBAAiBpN,QAAjB,EAA2BxP,CAA3B,EAA8Bqd,cAA9B,CAAX;;eAEW1I,SAASvT,IAAT,EAAeoO,QAAf,eAA8BuF,IAA9B,IAAoCzB,8BAApC,IAAX;;WAEOtT,EAAE+Q,IAAF,CAAOvB,QAAP,CAAP;;;MAGEzN,eAAJ;;;;MAIIkb,MAAMC,OAAN,CAAcK,gBAAd,CAAJ,EAAqC;2CACVA,gBADU;;QAC5BrZ,QAD4B;QAClB1D,IADkB;;aAE1BR,EAAEkE,QAAF,EAAY1D,IAAZ,CAAiBA,IAAjB,EAAuBuI,IAAvB,EAAT;GAFF,MAGO;aACI/I,EAAEud,gBAAF,EAAoBzU,IAApB,GAA2BC,IAA3B,EAAT;;;;;MAKEuK,cAAJ,EAAoB;WACXqB,SAASvT,IAAT,EAAeW,MAAf,EAAuBgT,IAAvB,CAAP;;;SAGKhT,MAAP;;;AAGF,SAAS0b,aAAT,CAAuB1I,IAAvB,EAA6B;MACnB3T,IADmB,GACkB2T,IADlB,CACnB3T,IADmB;MACbsc,SADa,GACkB3I,IADlB,CACb2I,SADa;uBACkB3I,IADlB,CACF4I,QADE;MACFA,QADE,kCACS,IADT;;;MAGrB5b,SAASqb,oBAAYrI,IAAZ,IAAkBsI,gBAAgBK,UAAUtc,IAAV,CAAlC,IAAf;;;MAGIW,MAAJ,EAAY;WACHA,MAAP;;;;;MAKE4b,QAAJ,EAAc,OAAO7B,iBAAiB1a,IAAjB,EAAuB2T,IAAvB,CAAP;;SAEP,IAAP;;;AAGF,IAAM6I,gBAAgB;SAAA,qBACwB;QAApCF,SAAoC,uEAAxB5B,gBAAwB;QAAN/G,IAAM;gBACFA,IADE;QAClC8I,WADkC,SAClCA,WADkC;QACrBC,cADqB,SACrBA,cADqB;;;QAGtCJ,UAAUjB,MAAV,KAAqB,GAAzB,EAA8B,OAAOiB,UAAU3B,OAAV,CAAkBhH,IAAlB,CAAP;;wBAGzBA,IADL;;;;QAKI8I,WAAJ,EAAiB;UACT5b,WAAUwb,2BACX1I,IADW,IACL3T,MAAM,SADD,EACYkc,aAAa,IADzB,EAC+BrO,OAAO6O;SADtD;aAGO;;OAAP;;QAII7O,QAAQwO,2BAAmB1I,IAAnB,IAAyB3T,MAAM,OAA/B,IAAd;QACM+a,iBAAiBsB,2BAAmB1I,IAAnB,IAAyB3T,MAAM,gBAA/B,IAAvB;QACMiR,SAASoL,2BAAmB1I,IAAnB,IAAyB3T,MAAM,QAA/B,IAAf;QACMib,gBAAgBoB,2BAAmB1I,IAAnB,IAAyB3T,MAAM,eAA/B,IAAtB;QACMa,UAAUwb,2BACX1I,IADW,IACL3T,MAAM,SADD,EACYkc,aAAa,IADzB,EAC+BrO;OAD/C;QAGMmN,iBAAiBqB,2BAAmB1I,IAAnB,IAAyB3T,MAAM,gBAA/B,EAAiDa,gBAAjD,IAAvB;QACMyQ,MAAM+K,2BAAmB1I,IAAnB,IAAyB3T,MAAM,KAA/B,EAAsCa,gBAAtC,IAAZ;QACM0Z,UAAU8B,2BAAmB1I,IAAnB,IAAyB3T,MAAM,SAA/B,EAA0Ca,gBAA1C,IAAhB;QACMqa,aAAamB,2BAAmB1I,IAAnB,IAAyB3T,MAAM,YAA/B,EAA6Ca,gBAA7C,IAAnB;QACMsa,YAAYkB,2BAAmB1I,IAAnB,IAAyB3T,MAAM,WAA/B,EAA4C6N,YAA5C,IAAlB;;gBAEEwO,2BAAmB1I,IAAnB,IAAyB3T,MAAM,gBAA/B,QAAsD,EAAE5B,KAAK,IAAP,EAAaid,QAAQ,IAArB,EA/Bd;;QA8BlCjd,GA9BkC,SA8BlCA,GA9BkC;QA8B7Bid,MA9B6B,SA8B7BA,MA9B6B;;;WAiCnC;kBAAA;sBAAA;oBAAA;oCAAA;oCAAA;cAAA;kCAAA;cAAA;oBAAA;sBAAA;4BAAA;;KAAP;;CAlCJ,CAmDA;;AC5KA;wDAAe;QAEXJ,aAFW,SAEXA,aAFW;QAGXtL,IAHW,SAGXA,IAHW;QAIX/Q,CAJW,SAIXA,CAJW;QAKX2V,SALW,SAKXA,SALW;QAMX5T,MANW,SAMXA,MANW;QAOXgc,SAPW,SAOXA,SAPW;QAQX9O,KARW,SAQXA,KARW;QASXzP,GATW,SASXA,GATW;;;;;;;iBAAA,GAaD,CAbC;wBAAA,GAcQ,CAAC0M,aAAa1M,GAAb,CAAD,CAdR;;;;;;kBAkBN6c,iBAAiB2B,QAAQ,EAlBnB;;;;;qBAmBF,CAAT;;mBACUpc,SAASqc,MAAT,CAAgB5B,aAAhB,CApBC;;;aAAA;;mBAqBJrc,EAAE+Q,IAAF,EAAP;;yBArBW,GAuBW;mBACfsL,aADe;wBAAA;kBAAA;kCAAA;2BAKP,IALO;8BAMJpN,KANI;;aAvBX;0BAAA,GAiCY2O,cAAc7B,OAAd,CAAsBgC,SAAtB,EAAiCG,aAAjC,CAjCZ;;;yBAmCE/Q,IAAb,CAAkBkP,aAAlB;kCAEKta,MADL;sCAGMA,OAAOE,OADX,yCAGa+b,KAHb,uBAIIG,eAAelc,OAJnB;;;4BAQckc,eAAe9B,aAA/B;;;;;sBA9CW,GAiDMP,iBAAiBQ,UAAjB,CAA4B,EAAEra,mBAAiBF,OAAOE,OAAxB,WAAF,EAA5B,CAjDN;0DAmDRF,MAnDQ;2BAoDEic,KApDF;8BAqDKA,KArDL;;;;;;;;;;GAAf;;WAA8BI,eAA9B;;;;SAA8BA,eAA9B;;;ACKA,IAAMC,UAAU;OAAA,iBACF7e,GADE,EACGuR,IADH,EACoB;;;QAAXgE,IAAW,uEAAJ,EAAI;;;;;;;;oCAI5BA,IAJ4B,CAE9BuJ,aAF8B;2BAAA,uCAEd,IAFc;+BAI5BvJ,IAJ4B,CAG9B4I,QAH8B;sBAAA,kCAGnB,IAHmB;uBAAA,GAMdje,IAAIC,KAAJ,CAAUH,GAAV,CANc;;kBAQ3B3B,YAAY4B,SAAZ,CAR2B;;;;;+CASvB1B,OAAO8B,MATgB;;;uBAAA,GAYd6c,aAAald,GAAb,EAAkBC,SAAlB,CAZc;;;;qBAehBmC,SAASqc,MAAT,CAAgBze,GAAhB,EAAqBuR,IAArB,EAA2BtR,SAA3B,CAfgB;;;eAAA;;mBAkB5BO,EAAEb,KAlB0B;;;;;+CAmBvBa,CAnBuB;;;;qBAsBzBA,EAAE+Q,IAAF,EAAP;;;;uBAtBgC,GA0Bd/Q,EAAE,MAAF,EAAUiE,GAAV,CAAc,UAAC7D,CAAD,EAAIC,IAAJ;uBAAaL,EAAEK,IAAF,EAAQG,IAAR,CAAa,MAAb,CAAb;eAAd,EAAiD8P,OAAjD,EA1Bc;oBAAA,GA4BnBsN,cAAc7B,OAAd,CAAsBgC,SAAtB,EAAiC,EAAEve,QAAF,EAAOuR,UAAP,EAAa/Q,IAAb,EAAgB2V,oBAAhB,EAA2BlW,oBAA3B,EAAsCke,kBAAtC,EAAjC,CA5BmB;wBA6BC5b,MA7BD;mBAAA,WA6BxBkN,KA7BwB;2BAAA,WA6BjBoN,aA7BiB;;;;oBAgC5BiC,iBAAiBjC,aAhCW;;;;;;qBAiCf+B,gBACb;oCAAA;4CAAA;0BAAA;oBAAA;oCAAA;8BAAA;4BAAA;;eADa,CAjCe;;;oBAAA;;;;;oCA+CzBrc,MADL;6BAEe,CAFf;gCAGkB;;;;+CAIbA,MArDyB;;;;;;;;;GADpB;;;;;eAAA,yBA2DMvC,GA3DN,EA2DW;;;;;;;;;qBACVoC,SAASqc,MAAT,CAAgBze,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/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/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/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 'babel-polyfill';\n\nimport 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 $;\n}\n","export const IS_LINK = new RegExp('https?://', 'i');\nexport const IS_IMAGE = new RegExp('.(png|gif|jpe?g)', 'i');\n\nexport const TAGS_TO_REMOVE = [\n 'script',\n 'style',\n 'form',\n].join(',');\n","import 'babel-polyfill';\n\nimport {\n IS_LINK,\n IS_IMAGE,\n} from './constants';\n\n// Convert all instances of images with potentially\n// lazy loaded images into normal images.\n// Many sites will have img tags with no source, or an image tag with a src\n// attribute that a is a placeholer. We need to be able to properly fill in\n// the src attribute so the images are no longer lazy loaded.\nexport default function convertLazyLoadedImages($) {\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 'babel-polyfill';\n\nimport 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 && $children.get(0).tagName === 'img') {\n return 'figure';\n }\n\n return null;\n },\n },\n },\n\n title: {\n selectors: [\n 'h1.lede-feature-title',\n 'h1.headline-primary',\n 'h1',\n ],\n },\n\n author: {\n selectors: [\n '.by-authors',\n '.lede-feature-author',\n ],\n },\n\n dek: {\n selectors: [\n '.lede-feature-teaser',\n ],\n },\n\n date_published: {\n selectors: [\n ['time.article-timestamp[datetime]', 'datetime'],\n 'time.article-timestamp',\n ],\n },\n};\n","export const BloggerExtractor = {\n domain: 'blogspot.com',\n content: {\n // Blogger is insane and does not load its content\n // initially in the page, but it's all there\n // in noscript\n selectors: [\n '.post-content noscript',\n ],\n\n // Selectors to remove from the extracted content\n clean: [\n ],\n\n // Convert the noscript tag to a div\n transforms: {\n noscript: 'div',\n },\n },\n\n author: {\n selectors: [\n '.post-author-name',\n ],\n },\n\n title: {\n selectors: [\n 'h2.title',\n ],\n },\n\n date_published: {\n selectors: [\n 'span.publishdate',\n ],\n },\n};\n","export const WikipediaExtractor = {\n domain: 'wikipedia.org',\n content: {\n selectors: [\n '#mw-content-text',\n ],\n\n defaultCleaner: false,\n\n // transform top infobox to an image with caption\n transforms: {\n '.infobox img': ($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 '.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","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';\n\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\n};\n\nexport default Extractors;\n","// Spacer images to be removed\nexport const SPACER_RE = new RegExp('trans|transparent|spacer|blank', 'i');\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', 'score'];\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 $;\n}\n","import 'babel-polyfill';\n\nexport default function convertNodeTo($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 STRIP_OUTPUT_TAGS,\n} from './constants';\n\nexport default function stripJunkTags(article, $, tags = []) {\n if (tags.length === 0) {\n tags = STRIP_OUTPUT_TAGS;\n }\n\n $(tags.join(','), article).remove();\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 $;\n}\n","import 'babel-polyfill';\n\nimport { WHITELIST_ATTRS_RE } from './constants';\n\nfunction removeAllButWhitelist($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.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","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';\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, { $ }) {\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 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 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 // 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 // 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 'babel-polyfill';\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';\nimport 'babel-polyfill';\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 'babel-polyfill';\n\nimport { 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 'babel-polyfill';\nimport 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 'babel-polyfill';\nimport 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.$ = $;\n }\n\n const title = this.title(options);\n const date_published = this.date_published(options);\n const author = this.author(options);\n const content = this.content({ ...options, title });\n const lead_image_url = this.lead_image_url({ ...options, content });\n const dek = this.dek({ ...options, content });\n const next_page_url = this.next_page_url(options);\n const excerpt = this.excerpt({ ...options, content });\n const word_count = this.word_count({ ...options, content });\n const direction = this.direction({ title });\n const { url, domain } = this.url_and_domain(options);\n\n return {\n title,\n author,\n date_published: date_published || null,\n dek,\n lead_image_url,\n content,\n next_page_url,\n url,\n domain,\n excerpt,\n word_count,\n direction,\n };\n },\n};\n\nexport default GenericExtractor;\n","import URL from 'url';\n\nimport Extractors from './all';\nimport GenericExtractor from './generic';\n\nexport default function getExtractor(url, parsedUrl) {\n parsedUrl = parsedUrl || URL.parse(url);\n const { hostname } = parsedUrl;\n const baseDomain = hostname.split('.').slice(-2).join('.');\n\n return Extractors[hostname] || Extractors[baseDomain] || GenericExtractor;\n}\n","import 'babel-polyfill';\n\nimport Cleaners from 'cleaners';\nimport { convertNodeTo } from 'utils/dom';\nimport GenericExtractor from './generic';\n\n// Remove elements by an array of selectors\nexport function cleanBySelectors($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 dek = extractResult({ ...opts, type: 'dek', content });\n const excerpt = extractResult({ ...opts, type: 'excerpt', content });\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 'babel-polyfill';\nimport { 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","Extractors","SPACER_RE","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","$img","height","parseInt","removeSpacers","cleanImages","$article","stripJunkTags","article","tags","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","split","isGoodSegment","segment","firstSegmentHasLetters","goodSegment","toLowerCase","articleBaseUrl","parsed","protocol","host","path","cleanedSegments","reverse","rawSegment","possibleSegment","fileExt","push","SENTENCE_END_RE","hasSentenceEnd","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","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","excerpt","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;;AClBA,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;;;AC9EA,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;;ACIP;;;;;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;;;ACtBF,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;;;ACRF,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;;ACpEO,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,UADS;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,WADS,EAET,SAFS;GATiB;;WAerB;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;GA1CqB;;kBAsDd,IAtDc;;kBAwDd,IAxDc;;OA0DzB,IA1DyB;;iBA4Df,IA5De;;WA8DrB;CA9DJ;;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;;ACQP,IAAMC,aAAa;eACJrB,cADI;kBAEDG,gBAFC;mBAGAC,kBAHA;iBAIFI,gBAJE;qBAKEK,gBALF;yBAMMI,oBANN;uBAOIC,kBAPJ;mBAQAC,cARA;iBASFC;;CATjB,CAaA;;ACxBA;AACA,AAAO,IAAME,YAAY,IAAI1F,MAAJ,CAAW,gCAAX,EAA6C,GAA7C,CAAlB;;;AAGP,AAAO,IAAM2F,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,aAAa3F,IAAb,CAAkB,GAAlB,CAAzB;AACP,AAAO,IAAMgG,kBAAkB,CAAC,KAAD,EAAQ,QAAR,EAAkB,MAAlB,EAA0B,OAA1B,EAAmC,IAAnC,EAAyC,KAAzC,EAAgD,OAAhD,CAAxB;AACP,AAAO,IAAMC,qBAAqB,IAAIlG,MAAJ,QAAgBiG,gBAAgBhG,IAAhB,CAAqB,GAArB,CAAhB,SAA+C,GAA/C,CAA3B;;;AAGP,AAAO,IAAMkG,oBAAoB,CAAC,GAAD,CAA1B;AACP,AAAO,IAAMC,yBAAyBD,kBAAkBL,GAAlB,CAAsB;SAAUO,GAAV;CAAtB,EAA6CpG,IAA7C,CAAkD,GAAlD,CAA/B;;;AAGP,AAAO,IAAMqG,2BAA2B,CAAC,IAAD,EAAO,IAAP,EAAa,OAAb,EAAsB,KAAtB,EAA6B,QAA7B,EAAuC,MAAvC,EAA+CrG,IAA/C,CAAoD,GAApD,CAAjC;;;AAGP,IAAMsG,cAAc,CAAC,IAAD,EAAO,IAAP,EAAa,IAAb,EAAmB,IAAnB,EAAyB,IAAzB,CAApB;AACA,AAAO,IAAMC,kBAAkBD,YAAYtG,IAAZ,CAAiB,GAAjB,CAAxB;;;;;;;;AASP,AAAO,IAAMwG,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,EASjC1G,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM2G,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,IAAI7G,MAAJ,QAAgB4G,uBAAuB3G,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;AAGP,AAYA,AAAO,IAAM6G,cAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,iBAAiB,IAAI/G,MAAJ,CAAW8G,YAAY7G,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAM+G,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,IAAIjH,MAAJ,CAAWgH,qBAAqB/G,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA;;;;AAIA,AAAO,IAAMiH,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,IAAInH,MAAJ,CAAWkH,qBAAqBjH,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMmH,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,IAAItH,MAAJ,CAAWqH,sBAAsBpH,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;AAGP,AAAO,IAAMsH,UAAU,IAAIvH,MAAJ,CAAW,iBAAX,EAA8B,GAA9B,CAAhB;;AAEP,AAMA,AAIA,AAIA,AAGA,AAGA;;AAEA,AAAO,IAAMwH,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,IAAIzH,MAAJ,QAAgBwH,iBAAiBvH,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAMyH,sBAAsBjB,8BAA8BxG,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM0H,uBAAuB,IAAI3H,MAAJ,CAAW0H,mBAAX,EAAgC,GAAhC,CAA7B;;AAEP,IAAME,sBAAsBlB,8BAA8BzG,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM4H,uBAAuB,IAAI7H,MAAJ,CAAW4H,mBAAX,EAAgC,GAAhC,CAA7B,CAEP,AAGA,AACA,AACA,AAEA;;AC3Xe,SAASE,uBAAT,CAAiClG,CAAjC,EAAoC;;;;;;;;;;IAU/C,GAAF,EAAOmG,GAAP,CAAW,GAAX,EAAgBhG,IAAhB,CAAqB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;QAC9BC,QAAQN,EAAEK,IAAF,CAAd;QACM+F,UAAU9F,MAAME,IAAN,CAAW,OAAX,CAAhB;QACM6F,KAAK/F,MAAME,IAAN,CAAW,IAAX,CAAX;QACI,CAAC6F,EAAD,IAAO,CAACD,OAAZ,EAAqB;;QAEfE,cAAgBF,WAAW,EAA3B,WAAiCC,MAAM,EAAvC,CAAN;QACIJ,qBAAqB1G,IAArB,CAA0B+G,UAA1B,CAAJ,EAA2C;;KAA3C,MAEO,IAAIP,qBAAqBxG,IAArB,CAA0B+G,UAA1B,CAAJ,EAA2C;YAC1C5E,MAAN;;GAVJ;;SAcO1B,CAAP;;;AC3BF;;;;;;;;;;AAUA,AAAe,SAASuG,OAAT,CAAiBvG,CAAjB,EAAoB;MAC7BwG,aAAa,KAAjB;IACE,IAAF,EAAQrG,IAAR,CAAa,UAACgB,KAAD,EAAQsF,OAAR,EAAoB;QACzBC,cAAc1G,EAAEyG,OAAF,EAAWE,IAAX,GAAkBpI,GAAlB,CAAsB,CAAtB,CAApB;;QAEImI,eAAeA,YAAYhE,OAAZ,KAAwB,IAA3C,EAAiD;mBAClC,IAAb;QACE+D,OAAF,EAAW/E,MAAX;KAFF,MAGO,IAAI8E,UAAJ,EAAgB;mBACR,KAAb;;mBAEaC,OAAb,EAAsBzG,CAAtB,EAAyB,IAAzB;;GATJ;;SAaOA,CAAP;;;ACzBF;;;;;;;;;;;AAWA,AAAe,SAAS4G,YAAT,CAAsBvG,IAAtB,EAA4BL,CAA5B,EAA2C;MAAZ6G,EAAY,uEAAP,KAAO;;MAClDvG,QAAQN,EAAEK,IAAF,CAAd;;MAEIwG,EAAJ,EAAQ;QACFC,UAAUzG,KAAK0G,WAAnB;QACMC,IAAIhH,EAAE,SAAF,CAAV;;;;WAIO8G,WAAW,EAAEA,QAAQpE,OAAR,IAAmBmD,oBAAoBtG,IAApB,CAAyBuH,QAAQpE,OAAjC,CAArB,CAAlB,EAAmF;UAC3EqE,cAAcD,QAAQC,WAA5B;QACED,OAAF,EAAWG,QAAX,CAAoBD,CAApB;gBACUD,WAAV;;;UAGI3D,WAAN,CAAkB4D,CAAlB;UACMtF,MAAN;WACO1B,CAAP;;;SAGKA,CAAP;;;AC7BF,SAASkH,WAAT,CAAqBlH,CAArB,EAAwB;IACpB,KAAF,EAASG,IAAT,CAAc,UAACgB,KAAD,EAAQgG,GAAR,EAAgB;QACtBC,OAAOpH,EAAEmH,GAAF,CAAb;QACME,cAAcD,KAAK9E,QAAL,CAAcyC,mBAAd,EAAmCxC,MAAnC,KAA8C,CAAlE;;QAEI8E,WAAJ,EAAiB;oBACDD,IAAd,EAAoBpH,CAApB,EAAuB,GAAvB;;GALJ;;SASOA,CAAP;;;AAGF,SAASsH,YAAT,CAAsBtH,CAAtB,EAAyB;IACrB,MAAF,EAAUG,IAAV,CAAe,UAACgB,KAAD,EAAQoG,IAAR,EAAiB;QACxBC,QAAQxH,EAAEuH,IAAF,CAAd;QACMF,cAAcG,MAAM1E,OAAN,CAAc,QAAd,EAAwBP,MAAxB,KAAmC,CAAvD;QACI8E,WAAJ,EAAiB;oBACDG,KAAd,EAAqBxH,CAArB,EAAwB,GAAxB;;GAJJ;;SAQOA,CAAP;;;;;;;;;;;;;;;AAeF,AAAe,SAASyH,mBAAT,CAA6BzH,CAA7B,EAAgC;MACzCuG,QAAQvG,CAAR,CAAJ;MACIkH,YAAYlH,CAAZ,CAAJ;MACIsH,aAAatH,CAAb,CAAJ;;SAEOA,CAAP;;;AC5Ca,SAAS0H,aAAT,CAAuBpH,KAAvB,EAA8BN,CAA9B,EAA4C;MAAXyE,GAAW,uEAAL,GAAK;;MACnDpE,OAAOC,MAAM/B,GAAN,CAAU,CAAV,CAAb;MACI,CAAC8B,IAAL,EAAW;WACFL,CAAP;;;mBAEkBM,MAAM/B,GAAN,CAAU,CAAV,CALqC;;MAKjDyC,OALiD,cAKjDA,OALiD;;MAMnD2G,eAAe,iBAAgB3G,OAAhB,EACQkD,GADR,CACY;WAAU0D,GAAV,SAAiB5G,QAAQ4G,GAAR,CAAjB;GADZ,EAEQvJ,IAFR,CAEa,GAFb,CAArB;;QAIM+E,WAAN,OAAsBqB,GAAtB,SAA6BkD,YAA7B,SAA6CrH,MAAMkB,QAAN,EAA7C,UAAkEiD,GAAlE;SACOzE,CAAP;;;ACXF,SAAS6H,cAAT,CAAwBC,IAAxB,EAA8B9H,CAA9B,EAAiC;MACzB+H,SAASC,SAASF,KAAKtH,IAAL,CAAU,QAAV,CAAT,EAA8B,EAA9B,CAAf;MACM+C,QAAQyE,SAASF,KAAKtH,IAAL,CAAU,OAAV,CAAT,EAA6B,EAA7B,KAAoC,EAAlD;;;;;MAKI,CAACuH,UAAU,EAAX,IAAiB,EAAjB,IAAuBxE,QAAQ,EAAnC,EAAuC;SAChC7B,MAAL;GADF,MAEO,IAAIqG,MAAJ,EAAY;;;;SAIZtH,UAAL,CAAgB,QAAhB;;;SAGKT,CAAP;;;;;AAKF,SAASiI,aAAT,CAAuBH,IAAvB,EAA6B9H,CAA7B,EAAgC;MAC1B8D,UAAUvE,IAAV,CAAeuI,KAAKtH,IAAL,CAAU,KAAV,CAAf,CAAJ,EAAsC;SAC/BkB,MAAL;;;SAGK1B,CAAP;;;AAGF,AAAe,SAASkI,WAAT,CAAqBC,QAArB,EAA+BnI,CAA/B,EAAkC;WACtCuB,IAAT,CAAc,KAAd,EAAqBpB,IAArB,CAA0B,UAACgB,KAAD,EAAQJ,GAAR,EAAgB;QAClC+G,OAAO9H,EAAEe,GAAF,CAAb;;mBAEe+G,IAAf,EAAqB9H,CAArB;kBACc8H,IAAd,EAAoB9H,CAApB;GAJF;;SAOOA,CAAP;;;ACnCa,SAASoI,aAAT,CAAuBC,OAAvB,EAAgCrI,CAAhC,EAA8C;MAAXsI,IAAW,uEAAJ,EAAI;;MACvDA,KAAK/F,MAAL,KAAgB,CAApB,EAAuB;WACdwB,iBAAP;;;IAGAuE,KAAKjK,IAAL,CAAU,GAAV,CAAF,EAAkBgK,OAAlB,EAA2B3G,MAA3B;;SAEO1B,CAAP;;;ACTF;;;;AAGA,AAAe,SAASuI,UAAT,CAAoBF,OAApB,EAA6BrI,CAA7B,EAAgC;MACvCwI,SAASxI,EAAE,IAAF,EAAQqI,OAAR,CAAf;;MAEIG,OAAOjG,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;;;ACZF,SAASyI,qBAAT,CAA+BN,QAA/B,EAAyC;;WAE9B5G,IAAT,CAAc,GAAd,EAAmBpB,IAAnB,CAAwB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;SAClCW,OAAL,GAAe,iBAAgBX,KAAKW,OAArB,EAA8B0H,MAA9B,CAAqC,UAACC,GAAD,EAAMnI,IAAN,EAAe;UAC7D8D,mBAAmB/E,IAAnB,CAAwBiB,IAAxB,CAAJ,EAAmC;4BACrBmI,GAAZ,sBAAkBnI,IAAlB,EAAyBH,KAAKW,OAAL,CAAaR,IAAb,CAAzB;;;aAGKmI,GAAP;KALa,EAMZ,EANY,CAAf;GADF;;;;;;;;;;AAkBF,AAAe,SAASC,eAAT,CAAyBT,QAAzB,EAAmC;wBAC1BA,QAAtB;;SAEOA,QAAP;;;AC3Ba,SAASU,WAAT,CAAqBV,QAArB,EAA+BnI,CAA/B,EAAkC;WACtCuB,IAAT,CAAc,GAAd,EAAmBpB,IAAnB,CAAwB,UAACgB,KAAD,EAAQ6F,CAAR,EAAc;QAC9B8B,KAAK9I,EAAEgH,CAAF,CAAX;QACI8B,GAAGC,IAAH,GAAUC,IAAV,OAAqB,EAAzB,EAA6BF,GAAGpH,MAAH;GAF/B;;SAKO1B,CAAP;;;ACNF;;;;;;AAMA,AAAO,IAAM6E,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,EASjC1G,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM2G,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,IAAI7G,MAAJ,QAAgB4G,yBAAuB3G,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;;;;AAMP,AAAO,IAAM4K,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,IAAM/D,gBAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,mBAAiB,IAAI/G,MAAJ,CAAW8G,cAAY7G,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAM+G,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,IAAIjH,MAAJ,CAAWgH,uBAAqB/G,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAM6K,sBAAoB,IAAI9K,MAAJ,CAAW,qBAAX,EAAkC,GAAlC,CAA1B;;;;;;AAMP,AAAO,IAAMkH,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,IAAInH,MAAJ,CAAWkH,uBAAqBjH,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA,AAGA,AAGA;;AAEA,AAAO,IAAMuH,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,IAAIzH,MAAJ,QAAgBwH,mBAAiBvH,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAMyH,wBAAsBjB,gCAA8BxG,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,IAAM2H,wBAAsBlB,gCAA8BzG,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,AAGA,AAAO,IAAM8K,yBAAuB,IAAI/K,MAAJ,CAAW,mBAAX,EAAgC,GAAhC,CAA7B;AACP,AAAO,IAAMgL,uBAAqB,IAAIhL,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAA3B;AACP,AAAO,IAAMiL,aAAW,IAAIjL,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB,CAEP;;AC3SA;AACA,AAAe,SAASkL,SAAT,CAAmBjJ,IAAnB,EAAyB;MAChC+F,UAAU/F,KAAKG,IAAL,CAAU,OAAV,CAAhB;MACM6F,KAAKhG,KAAKG,IAAL,CAAU,IAAV,CAAX;MACI+I,QAAQ,CAAZ;;MAEIlD,EAAJ,EAAQ;;QAEFhB,oBAAkB9F,IAAlB,CAAuB8G,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;QAEEd,oBAAkBhG,IAAlB,CAAuB8G,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;;;MAIAD,OAAJ,EAAa;QACPmD,UAAU,CAAd,EAAiB;;;UAGXlE,oBAAkB9F,IAAlB,CAAuB6G,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;UAEEb,oBAAkBhG,IAAlB,CAAuB6G,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;;;;;;QAOAjB,iBAAe5F,IAAf,CAAoB6G,OAApB,CAAJ,EAAkC;eACvB,EAAT;;;;;;;QAOE8C,oBAAkB3J,IAAlB,CAAuB6G,OAAvB,CAAJ,EAAqC;eAC1B,EAAT;;;;SAIGmD,KAAP;;;ACpDF;;;AAGA,AAAe,SAASC,QAAT,CAAkBlJ,KAAlB,EAAyB;SAC/BmJ,WAAWnJ,MAAME,IAAN,CAAW,OAAX,CAAX,KAAmC,IAA1C;;;ACJF;AACA,AAAe,SAASkJ,WAAT,CAAqBX,IAArB,EAA2B;SACjC,CAACA,KAAKY,KAAL,CAAW,IAAX,KAAoB,EAArB,EAAyBpH,MAAhC;;;ACFF,IAAMqH,QAAQ,IAAIxL,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAd;;AAEA,AAAe,SAASyL,WAAT,CAAqBC,UAArB,EAAgD;MAAfpH,OAAe,uEAAL,GAAK;;MACvDqH,SAASD,aAAa,EAA5B;;MAEIC,SAAS,CAAb,EAAgB;QACVC,oBAAJ;;;;;;;QAOIJ,MAAMrK,IAAN,CAAWmD,OAAX,CAAJ,EAAyB;oBACTqH,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,CAAwB/J,IAAxB,EAA8B;MACvCkJ,QAAQ,CAAZ;MACMR,OAAO1I,KAAK0I,IAAL,GAAYC,IAAZ,EAAb;MACMc,aAAaf,KAAKxG,MAAxB;;;MAGIuH,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,CAAkBhK,KAAlB,EAAyBN,CAAzB,EAA4BuJ,KAA5B,EAAmC;QAC1C/I,IAAN,CAAW,OAAX,EAAoB+I,KAApB;SACOjJ,KAAP;;;ACEa,SAASiK,QAAT,CAAkBjK,KAAlB,EAAyBN,CAAzB,EAA4BwK,MAA5B,EAAoC;MAC7C;QACIjB,QAAQkB,eAAenK,KAAf,EAAsBN,CAAtB,IAA2BwK,MAAzC;aACSlK,KAAT,EAAgBN,CAAhB,EAAmBuJ,KAAnB;GAFF,CAGE,OAAOmB,CAAP,EAAU;;;;SAILpK,KAAP;;;ACXF;AACA,AAAe,SAASqK,WAAT,CAAqBtK,IAArB,EAA2BL,CAA3B,EAA8BuJ,KAA9B,EAAqC;MAC5CqB,SAASvK,KAAKuK,MAAL,EAAf;MACIA,MAAJ,EAAY;aACDA,MAAT,EAAiB5K,CAAjB,EAAoBuJ,QAAQ,IAA5B;;;SAGKlJ,IAAP;;;ACFF;;;AAGA,AAAe,SAASoK,cAAT,CAAwBnK,KAAxB,EAA+BN,CAA/B,EAAsD;MAApB6K,WAAoB,uEAAN,IAAM;;MAC/DtB,QAAQC,SAASlJ,KAAT,CAAZ;;MAEIiJ,KAAJ,EAAW;WACFA,KAAP;;;UAGMuB,UAAUxK,KAAV,CAAR;;MAEIuK,WAAJ,EAAiB;aACNvB,UAAUhJ,KAAV,CAAT;;;cAGUA,KAAZ,EAAmBN,CAAnB,EAAsBuJ,KAAtB;;SAEOA,KAAP;;;AClBF;;AAEA,AAAe,SAASuB,SAAT,CAAmBxK,KAAnB,EAA0B;mBACnBA,MAAM/B,GAAN,CAAU,CAAV,CADmB;;MAC/BmE,OAD+B,cAC/BA,OAD+B;;;;;;MAMnCyG,uBAAqB5J,IAArB,CAA0BmD,OAA1B,CAAJ,EAAwC;WAC/B0H,eAAe9J,KAAf,CAAP;GADF,MAEO,IAAIoC,YAAY,KAAhB,EAAuB;WACrB,CAAP;GADK,MAEA,IAAI0G,qBAAmB7J,IAAnB,CAAwBmD,OAAxB,CAAJ,EAAsC;WACpC,CAAP;GADK,MAEA,IAAI2G,WAAS9J,IAAT,CAAcmD,OAAd,CAAJ,EAA4B;WAC1B,CAAC,CAAR;GADK,MAEA,IAAIA,YAAY,IAAhB,EAAsB;WACpB,CAAC,CAAR;;;SAGK,CAAP;;;ACjBF,SAAS4E,cAAT,CAAsBhH,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,SAAS+K,UAAT,CAAoBzK,KAApB,EAA2BN,CAA3B,EAA8BuJ,KAA9B,EAAqC;MAC/BjJ,KAAJ,EAAW;mBACIA,KAAb,EAAoBN,CAApB;aACSM,KAAT,EAAgBN,CAAhB,EAAmBuJ,KAAnB;;;;AAIJ,SAASyB,OAAT,CAAiBhL,CAAjB,EAAoB6K,WAApB,EAAiC;IAC7B,QAAF,EAAY1E,GAAZ,CAAgB,SAAhB,EAA2BhG,IAA3B,CAAgC,UAACgB,KAAD,EAAQd,IAAR,EAAiB;;;QAG3CC,QAAQN,EAAEK,IAAF,CAAZ;YACQiK,SAAShK,KAAT,EAAgBN,CAAhB,EAAmByK,eAAenK,KAAf,EAAsBN,CAAtB,EAAyB6K,WAAzB,CAAnB,CAAR;;QAEMhI,UAAUvC,MAAMsK,MAAN,EAAhB;QACMK,WAAWH,UAAUxK,KAAV,CAAjB;;eAEWuC,OAAX,EAAoB7C,CAApB,EAAuBiL,QAAvB,EAAiCJ,WAAjC;QACIhI,OAAJ,EAAa;;;iBAGAA,QAAQ+H,MAAR,EAAX,EAA6B5K,CAA7B,EAAgCiL,WAAW,CAA3C,EAA8CJ,WAA9C;;GAbJ;;SAiBO7K,CAAP;;;;;AAKF,AAAe,SAASkL,YAAT,CAAsBlL,CAAtB,EAA6C;MAApB6K,WAAoB,uEAAN,IAAM;;;;4BAGlC5J,OAAxB,CAAgC,gBAAqC;;;QAAnCkK,cAAmC;QAAnBC,aAAmB;;MAC9DD,cAAL,SAAuBC,aAAvB,EAAwCjL,IAAxC,CAA6C,UAACgB,KAAD,EAAQd,IAAR,EAAiB;eACnDL,EAAEK,IAAF,EAAQuK,MAAR,CAAeO,cAAf,CAAT,EAAyCnL,CAAzC,EAA4C,EAA5C;KADF;GADF;;;;;;;UAWQA,CAAR,EAAW6K,WAAX;UACQ7K,CAAR,EAAW6K,WAAX;;SAEO7K,CAAP;;;ACpEF,IAAMqL,eAAe,SAArB;;AAEA,AAAe,SAASC,eAAT,CAAyBvC,IAAzB,EAA+B;SACrCA,KAAKvF,OAAL,CAAa6H,YAAb,EAA2B,GAA3B,EAAgCrC,IAAhC,EAAP;;;ACHF;;;;;AAKA,AAAe,SAASuC,cAAT,CAAwB/L,GAAxB,EAA6BgM,SAA7B,EAAwC;MAC/CC,UAAUD,UAAUjK,IAAV,CAAe;WAAMmK,GAAGnM,IAAH,CAAQC,GAAR,CAAN;GAAf,CAAhB;MACIiM,OAAJ,EAAa;WACJA,QAAQE,IAAR,CAAanM,GAAb,EAAkB,CAAlB,CAAP;;;SAGK,IAAP;;;ACXF;;;;;;;;;;;;;;;;AAgBA,AAAO,IAAMoM,kBAAkB,IAAIxN,MAAJ,CAAW,0EAAX,EAAuF,GAAvF,CAAxB;;AAEP,AAAO,IAAMyN,eAAe,QAArB;;AAEP,AAAO,IAAMC,cAAc,WAApB;AACP,AAAO,IAAMC,cAAc,WAApB;;ACnBQ,SAASC,cAAT,CAAwBxM,GAAxB,EAA6B;MACpCyM,UAAUzM,IAAImK,KAAJ,CAAUiC,eAAV,CAAhB;MACI,CAACK,OAAL,EAAc,OAAO,IAAP;;MAERC,UAAUlE,SAASiE,QAAQ,CAAR,CAAT,EAAqB,EAArB,CAAhB;;;;SAIOC,UAAU,GAAV,GAAgBA,OAAhB,GAA0B,IAAjC;;;ACVa,SAASC,YAAT,CAAsB3M,GAAtB,EAA2B;SACjCA,IAAI4M,KAAJ,CAAU,GAAV,EAAe,CAAf,EAAkB5I,OAAlB,CAA0B,KAA1B,EAAiC,EAAjC,CAAP;;;ACOF,SAAS6I,aAAT,CAAuBC,OAAvB,EAAgCnL,KAAhC,EAAuCoL,sBAAvC,EAA+D;MACzDC,cAAc,IAAlB;;;;MAIIrL,QAAQ,CAAR,IAAa4K,YAAYxM,IAAZ,CAAiB+M,OAAjB,CAAb,IAA0CA,QAAQ/J,MAAR,GAAiB,CAA/D,EAAkE;kBAClD,IAAd;;;;;MAKEpB,UAAU,CAAV,IAAemL,QAAQG,WAAR,OAA0B,OAA7C,EAAsD;kBACtC,KAAd;;;;;MAKEtL,QAAQ,CAAR,IAAamL,QAAQ/J,MAAR,GAAiB,CAA9B,IAAmC,CAACgK,sBAAxC,EAAgE;kBAChD,KAAd;;;SAGKC,WAAP;;;;;;AAMF,AAAe,SAASE,cAAT,CAAwBlN,GAAxB,EAA6BmN,MAA7B,EAAqC;MAC5ClN,YAAYkN,UAAUjN,IAAIC,KAAJ,CAAUH,GAAV,CAA5B;MACQoN,QAF0C,GAEjBnN,SAFiB,CAE1CmN,QAF0C;MAEhCC,IAFgC,GAEjBpN,SAFiB,CAEhCoN,IAFgC;MAE1BC,IAF0B,GAEjBrN,SAFiB,CAE1BqN,IAF0B;;;MAI9CP,yBAAyB,KAA7B;MACMQ,kBAAkBD,KAAKV,KAAL,CAAW,GAAX,EACvBY,OADuB,GAEvBtE,MAFuB,CAEhB,UAACC,GAAD,EAAMsE,UAAN,EAAkB9L,KAAlB,EAA4B;QAC9BmL,UAAUW,UAAd;;;QAGIX,QAAQpK,QAAR,CAAiB,GAAjB,CAAJ,EAA2B;2BACUoK,QAAQF,KAAR,CAAc,GAAd,CADV;;;;UAClBc,eADkB;UACDC,OADC;;UAErBrB,YAAYvM,IAAZ,CAAiB4N,OAAjB,CAAJ,EAA+B;kBACnBD,eAAV;;;;;;QAMAtB,gBAAgBrM,IAAhB,CAAqB+M,OAArB,KAAiCnL,QAAQ,CAA7C,EAAgD;gBACpCmL,QAAQ9I,OAAR,CAAgBoI,eAAhB,EAAiC,EAAjC,CAAV;;;;;;;QAOEzK,UAAU,CAAd,EAAiB;+BACU0K,aAAatM,IAAb,CAAkB+M,OAAlB,CAAzB;;;;QAIED,cAAcC,OAAd,EAAuBnL,KAAvB,EAA8BoL,sBAA9B,CAAJ,EAA2D;UACrDa,IAAJ,CAASd,OAAT;;;WAGK3D,GAAP;GAhCsB,EAiCrB,EAjCqB,CAAxB;;SAmCUiE,QAAV,UAAuBC,IAAvB,GAA8BE,gBAAgBC,OAAhB,GAA0B3O,IAA1B,CAA+B,GAA/B,CAA9B;;;AC3EF;;AAEA,IAAMgP,kBAAkB,IAAIjP,MAAJ,CAAW,QAAX,CAAxB;AACA,AAAe,SAASkP,cAAT,CAAwBvE,IAAxB,EAA8B;SACpCsE,gBAAgB9N,IAAhB,CAAqBwJ,IAArB,CAAP;;;ACKF;;;;;AAKA,AAAe,SAASwE,aAAT,CAAuBC,UAAvB,EAAmCC,QAAnC,EAA6CzN,CAA7C,EAAgD;MACzD,CAACwN,WAAW5C,MAAX,GAAoBrI,MAAzB,EAAiC;WACxBiL,UAAP;;;MAGIE,wBAAwBzD,KAAKE,GAAL,CAAS,EAAT,EAAasD,WAAW,IAAxB,CAA9B;MACME,cAAc3N,EAAE,aAAF,CAApB;;aAEW4K,MAAX,GAAoBtI,QAApB,GAA+BnC,IAA/B,CAAoC,UAACgB,KAAD,EAAQ2F,OAAR,EAAoB;QAChD8G,WAAW5N,EAAE8G,OAAF,CAAjB;;QAEI7B,4BAA0B1F,IAA1B,CAA+BuH,QAAQpE,OAAvC,CAAJ,EAAqD;aAC5C,IAAP;;;QAGImL,eAAerE,SAASoE,QAAT,CAArB;QACIC,YAAJ,EAAkB;UACZD,aAAaJ,UAAjB,EAA6B;oBACfrK,MAAZ,CAAmByK,QAAnB;OADF,MAEO;YACDE,eAAe,CAAnB;YACMC,UAAUC,YAAYJ,QAAZ,CAAhB;;;;YAIIG,UAAU,IAAd,EAAoB;0BACF,EAAhB;;;;;YAKEA,WAAW,GAAf,EAAoB;0BACF,EAAhB;;;;;YAKEH,SAASpN,IAAT,CAAc,OAAd,MAA2BgN,WAAWhN,IAAX,CAAgB,OAAhB,CAA/B,EAAyD;0BACvCiN,WAAW,GAA3B;;;YAGIQ,WAAWJ,eAAeC,YAAhC;;YAEIG,YAAYP,qBAAhB,EAAuC;iBAC9BC,YAAYxK,MAAZ,CAAmByK,QAAnB,CAAP;SADF,MAEO,IAAI9G,QAAQpE,OAAR,KAAoB,GAAxB,EAA6B;cAC5BwL,iBAAiBN,SAAS7E,IAAT,EAAvB;cACMoF,uBAAuBrE,WAAWoE,cAAX,CAA7B;;cAEIC,uBAAuB,EAAvB,IAA6BJ,UAAU,IAA3C,EAAiD;mBACxCJ,YAAYxK,MAAZ,CAAmByK,QAAnB,CAAP;WADF,MAEO,IAAIO,wBAAwB,EAAxB,IAA8BJ,YAAY,CAA1C,IACDT,eAAeY,cAAf,CADH,EACmC;mBACjCP,YAAYxK,MAAZ,CAAmByK,QAAnB,CAAP;;;;;;WAMD,IAAP;GAnDF;;SAsDOD,WAAP;;;ACxEF;;AAEA,AAAe,SAASS,gBAAT,CAA0BpO,CAA1B,EAA6B;MACtCwN,mBAAJ;MACIC,WAAW,CAAf;;IAEE,SAAF,EAAatN,IAAb,CAAkB,UAACgB,KAAD,EAAQd,IAAR,EAAiB;;QAE7B4E,4BAA0B1F,IAA1B,CAA+Bc,KAAKqC,OAApC,CAAJ,EAAkD;;;;QAI5CpC,QAAQN,EAAEK,IAAF,CAAd;QACMkJ,QAAQC,SAASlJ,KAAT,CAAd;;QAEIiJ,QAAQkE,QAAZ,EAAsB;iBACTlE,KAAX;mBACajJ,KAAb;;GAXJ;;;;MAiBI,CAACkN,UAAL,EAAiB;WACRxN,EAAE,MAAF,KAAaA,EAAE,GAAF,EAAOqO,KAAP,EAApB;;;eAGWd,cAAcC,UAAd,EAA0BC,QAA1B,EAAoCzN,CAApC,CAAb;;SAEOwN,UAAP;;;ACtBF,SAASc,mBAAT,CAA6BhO,KAA7B,EAAoCN,CAApC,EAAuCuO,MAAvC,EAA+C;;;;;MAKzCjO,MAAMkO,QAAN,CAAe,qBAAf,CAAJ,EAA2C;;;;MAIrCvM,UAAUqJ,gBAAgBhL,MAAMyI,IAAN,EAAhB,CAAhB;;MAEIW,YAAYzH,OAAZ,IAAuB,EAA3B,EAA+B;QACvBwM,SAASzO,EAAE,GAAF,EAAOM,KAAP,EAAciC,MAA7B;QACMmM,aAAa1O,EAAE,OAAF,EAAWM,KAAX,EAAkBiC,MAArC;;;QAGImM,aAAcD,SAAS,CAA3B,EAA+B;YACvB/M,MAAN;;;;QAIIpC,gBAAgB2C,QAAQM,MAA9B;QACMoM,WAAW3O,EAAE,KAAF,EAASM,KAAT,EAAgBiC,MAAjC;;;;QAIIjD,gBAAgB,EAAhB,IAAsBqP,aAAa,CAAvC,EAA0C;YAClCjN,MAAN;;;;QAIIqM,UAAUC,YAAY1N,KAAZ,CAAhB;;;;;QAKIiO,SAAS,EAAT,IAAeR,UAAU,GAAzB,IAAgCzO,gBAAgB,EAApD,EAAwD;YAChDoC,MAAN;;;;;;QAME6M,UAAU,EAAV,IAAgBR,UAAU,GAA9B,EAAmC;;;;UAI3BrL,UAAUpC,MAAM/B,GAAN,CAAU,CAAV,EAAamE,OAA7B;UACMkM,aAAalM,YAAY,IAAZ,IAAoBA,YAAY,IAAnD;UACIkM,UAAJ,EAAgB;YACRC,eAAevO,MAAMwO,IAAN,EAArB;YACID,gBAAgBvD,gBAAgBuD,aAAa9F,IAAb,EAAhB,EAAqCsB,KAArC,CAA2C,CAAC,CAA5C,MAAmD,GAAvE,EAA4E;;;;;YAKxE3I,MAAN;;;;QAIIqN,cAAc/O,EAAE,QAAF,EAAYM,KAAZ,EAAmBiC,MAAvC;;;QAGIwM,cAAc,CAAd,IAAmBzP,gBAAgB,GAAvC,EAA4C;YACpCoC,MAAN;;;;;;;;;;;;;AAaN,AAAe,SAASsN,SAAT,CAAmB7G,QAAnB,EAA6BnI,CAA7B,EAAgC;IAC3C0E,wBAAF,EAA4ByD,QAA5B,EAAsChI,IAAtC,CAA2C,UAACgB,KAAD,EAAQd,IAAR,EAAiB;QACpDC,QAAQN,EAAEK,IAAF,CAAd;QACIkO,SAAS/E,SAASlJ,KAAT,CAAb;QACI,CAACiO,MAAL,EAAa;eACF9D,eAAenK,KAAf,EAAsBN,CAAtB,CAAT;eACSM,KAAT,EAAgBN,CAAhB,EAAmBuO,MAAnB;;;;QAIEA,SAAS,CAAb,EAAgB;YACR7M,MAAN;KADF,MAEO;;0BAEepB,KAApB,EAA2BN,CAA3B,EAA8BuO,MAA9B;;GAbJ;;SAiBOvO,CAAP;;;ACrGa,SAASiP,YAAT,CAAsB9G,QAAtB,EAAgCnI,CAAhC,EAA+C;MAAZkP,KAAY,uEAAJ,EAAI;;IAC1DtK,eAAF,EAAmBuD,QAAnB,EAA6BhI,IAA7B,CAAkC,UAACgB,KAAD,EAAQgO,MAAR,EAAmB;QAC7CC,UAAUpP,EAAEmP,MAAF,CAAhB;;;;;QAKInP,EAAEoP,OAAF,EAAWjH,QAAX,EAAqBkH,OAArB,CAA6B,GAA7B,EAAkC9M,MAAlC,KAA6C,CAAjD,EAAoD;aAC3C6M,QAAQ1N,MAAR,EAAP;;;;QAIE4J,gBAAgBtL,EAAEmP,MAAF,EAAUpG,IAAV,EAAhB,MAAsCmG,KAA1C,EAAiD;aACxCE,QAAQ1N,MAAR,EAAP;;;;;QAKE4H,UAAUtJ,EAAEmP,MAAF,CAAV,IAAuB,CAA3B,EAA8B;aACrBC,QAAQ1N,MAAR,EAAP;;;WAGK0N,OAAP;GArBF;;SAwBOpP,CAAP;;;AC5BF;;;AAEA,AAAe,SAASsP,eAAT,CAAyBjH,OAAzB,EAAkCrI,CAAlC,EAAqC;;;;MAI9C0H,cAAc1H,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;MACI0H,cAAc1H,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;;SAEOA,CAAP;;;ACTF,SAASuP,UAAT,CAAoBvP,CAApB,EAAuBwP,OAAvB,EAAgChP,IAAhC,EAAsCiP,QAAtC,EAAgD;UACxCjP,IAAN,QAAeiP,QAAf,EAAyBtP,IAAzB,CAA8B,UAACC,CAAD,EAAIC,IAAJ,EAAa;QACnCb,MAAMa,KAAKW,OAAL,CAAaR,IAAb,CAAZ;QACMkP,cAAchQ,IAAIjB,OAAJ,CAAY+Q,OAAZ,EAAqBhQ,GAArB,CAApB;;SAEKwB,OAAL,CAAaR,IAAb,IAAqBkP,WAArB;GAJF;;;AAQF,AAAe,SAASC,iBAAT,CAA2BF,QAA3B,EAAqCzP,CAArC,EAAwCR,GAAxC,EAA6C;GACzD,MAAD,EAAS,KAAT,EAAgByB,OAAhB,CAAwB;WAAQsO,WAAWvP,CAAX,EAAcR,GAAd,EAAmBgB,IAAnB,EAAyBiP,QAAzB,CAAR;GAAxB;;SAEOA,QAAP;;;ACbK,SAAS3F,UAAT,CAAoBf,IAApB,EAA0B;SACxBA,KAAKC,IAAL,GACKxF,OADL,CACa,MADb,EACqB,GADrB,EAEKjB,MAFZ;;;;;;AAQF,AAAO,SAASyL,WAAT,CAAqB1N,KAArB,EAA4B;MAC3BsP,kBAAkB9F,WAAWxJ,MAAMyI,IAAN,EAAX,CAAxB;;MAEM8G,WAAWvP,MAAMiB,IAAN,CAAW,GAAX,EAAgBwH,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,CACb/P,CADa,EAEbgQ,SAFa,EAGbC,WAHa,EAKb;MADAjB,SACA,uEADY,IACZ;;MACMkB,aAAaF,UAAUvO,MAAV,CAAiB;WAAQwO,YAAYE,OAAZ,CAAoBC,IAApB,MAA8B,CAAC,CAAvC;GAAjB,CAAnB;;;;;;;;UAEWA,IAHX;;UAIQhP,OAAO,MAAb;UACMb,QAAQ,OAAd;;UAEM8P,QAAQrQ,YAAUoB,IAAV,UAAmBgP,IAAnB,QAAd;;;;;UAKME,SACJD,MAAMnM,GAAN,CAAU,UAAC/C,KAAD,EAAQd,IAAR;eAAiBL,EAAEK,IAAF,EAAQG,IAAR,CAAaD,KAAb,CAAjB;OAAV,EACMgQ,OADN,GAEM9O,MAFN,CAEa;eAAQsH,SAAS,EAAjB;OAFb,CADF;;;;;;UASIuH,OAAO/N,MAAP,KAAkB,CAAtB,EAAyB;YACnBiO,kBAAJ;;;YAGIxB,SAAJ,EAAe;sBACDyB,UAAUH,OAAO,CAAP,CAAV,EAAqBtQ,CAArB,CAAZ;SADF,MAEO;sBACOsQ,OAAO,CAAP,CAAZ;;;;aAGKE;;;;;sCA5BQN,UAAnB,4GAA+B;;;;;;;;;;;;;;;;;;;;;;SAiCxB,IAAP;;;AC3CF,SAASQ,UAAT,CAAoBpQ,KAApB,EAA2BqQ,WAA3B,EAAwC;;;MAGlCrQ,MAAMgC,QAAN,GAAiBC,MAAjB,GAA0BoO,WAA9B,EAA2C;WAClC,KAAP;;;MAGEC,cAActQ,KAAd,CAAJ,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;;;;AAMF,AAAe,SAASuQ,oBAAT,CACb7Q,CADa,EAEb8Q,SAFa,EAKb;MAFAH,WAEA,uEAFc,CAEd;MADAI,QACA,uEADW,IACX;;;;;;sCACuBD,SAAvB,4GAAkC;UAAvB3M,QAAuB;;UAC1BkM,QAAQrQ,EAAEmE,QAAF,CAAd;;;;UAIIkM,MAAM9N,MAAN,KAAiB,CAArB,EAAwB;YAChBjC,QAAQN,EAAEqQ,MAAM,CAAN,CAAF,CAAd;;YAEIK,WAAWpQ,KAAX,EAAkBqQ,WAAlB,CAAJ,EAAoC;cAC9B1O,gBAAJ;cACI8O,QAAJ,EAAc;sBACFzQ,MAAMyI,IAAN,EAAV;WADF,MAEO;sBACKzI,MAAM0Q,IAAN,EAAV;;;cAGE/O,OAAJ,EAAa;mBACJA,OAAP;;;;;;;;;;;;;;;;;;;;SAMD,IAAP;;;AChDF;AACA,AAAe,SAASwO,SAAT,CAAmB1H,IAAnB,EAAyB/I,CAAzB,EAA4B;;;MAGnCiR,YAAYjR,aAAW+I,IAAX,cAA0BA,IAA1B,EAAlB;SACOkI,cAAc,EAAd,GAAmBlI,IAAnB,GAA0BkI,SAAjC;;;ACLa,SAASL,aAAT,CAAuBtQ,KAAvB,EAA8B;MACrCwC,UAAUxC,MAAMwC,OAAN,GAAgByN,OAAhB,EAAhB;MACMW,gBAAgBpO,QAAQvB,IAAR,CAAa,UAACqJ,MAAD,EAAY;QACvCtE,aAAgBsE,OAAO5J,OAAP,CAAemQ,KAA/B,SAAwCvG,OAAO5J,OAAP,CAAeqF,EAA7D;WACOC,WAAWpE,QAAX,CAAoB,SAApB,CAAP;GAFoB,CAAtB;;SAKOgP,kBAAkBE,SAAzB;;;ACPF;;;;AAIA,AAAe,SAASC,gBAAT,CAA0B/Q,KAA1B,EAAiC;SACvCA,MAAMyI,IAAN,GAAaC,IAAb,GAAoBzG,MAApB,IAA8B,GAArC;;;ACHa,SAAS+O,WAAT,CAAqBtR,CAArB,EAAwB;SAC9BA,EAAEwF,cAAF,EAAkBjD,MAAlB,GAA2B,CAAlC;;;ACHF;AACA,AAAO,IAAMgP,kBAAkB,wCAAxB;;;;AAIP,AAAO,IAAMC,eAAe,IAAIpT,MAAJ,CAAW,aAAX,EAA0B,GAA1B,CAArB;AACP,AAYA,AASA;AACA,AAAO,IAAMqT,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,OAAOzT,IAAP,CAAY,GAAZ,CAAlB;AACA,IAAM2T,aAAa,qCAAnB;AACA,IAAMC,aAAa,wCAAnB;AACA,AAAO,IAAMC,oBACX,IAAI9T,MAAJ,OAAe4T,UAAf,WAA+BC,UAA/B,wBAA4DF,SAA5D,QAA0E,IAA1E,CADK;;;;;AAMP,AAAO,IAAMI,qBAAqB,gBAA3B;;AAEP,AAAO,IAAMC,oBACX,IAAIhU,MAAJ,CAAW,2BAAX,EAAwC,GAAxC,CADK;;ACxDP;;AAEA,AAAe,SAASiU,WAAT,CAAqBC,MAArB,EAA6B;SACnCA,OAAO9O,OAAP,CAAe+N,eAAf,EAAgC,IAAhC,EAAsCvI,IAAtC,EAAP;;;ACHa,SAASrH,OAAT,CAAe4Q,YAAf,EAA6B;iBAC3BA,aAAavJ,IAAb,EAAf;MACIwJ,SAASC,QAAT,CAAkBF,YAAlB,CAAJ,EAAqC;WAC5BA,YAAP;;;SAGK,IAAP;;;ACJF;;AAEA,AAAe,SAASG,QAAT,CAAkBC,GAAlB,QAA8B;MAAL3S,CAAK,QAALA,CAAK;;;MAEvC2S,IAAIpQ,MAAJ,GAAa,IAAb,IAAqBoQ,IAAIpQ,MAAJ,GAAa,CAAtC,EAAyC,OAAO,IAAP;;MAEnCqQ,UAAUnC,UAAUkC,GAAV,EAAe3S,CAAf,CAAhB;;;;MAIIwR,aAAajS,IAAb,CAAkBqT,OAAlB,CAAJ,EAAgC,OAAO,IAAP;;SAEzBA,QAAQ5J,IAAR,EAAP;;;ACfF;;;;AAIA,AASA,AAAO,SAAS6J,eAAT,CAAyBC,UAAzB,EAAqC;SACnC,CAACA,WAAWnJ,KAAX,CAAiBuI,iBAAjB,KAAuC,EAAxC,EACW7T,IADX,CACgB,GADhB,EAEWmF,OAFX,CAEmBqO,qBAFnB,EAE0C,GAF1C,EAGWrO,OAHX,CAGmBoO,sBAHnB,EAG2C,UAH3C,EAIWpO,OAJX,CAImBmO,oBAJnB,EAIyC,IAJzC,EAKW3I,IALX,EAAP;;;;;AAUF,AAAe,SAAS+J,kBAAT,CAA4BD,UAA5B,EAAwC;;MAEjDrB,eAAelS,IAAf,CAAoBuT,UAApB,KAAmCpB,gBAAgBnS,IAAhB,CAAqBuT,UAArB,CAAvC,EAAyE;iBAC1D9K,SAAS8K,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;;;AC1BF;;AACA,AAAe,SAASC,gBAAT,CACbhL,OADa,QASb;MANErI,CAMF,QANEA,CAMF;mCALEsT,kBAKF;MALEA,kBAKF,yCALuB,IAKvB;wBAJEpE,KAIF;MAJEA,KAIF,8BAJU,EAIV;sBAHE1P,GAGF;MAHEA,GAGF,4BAHQ,EAGR;iCAFE+T,cAEF;MAFEA,cAEF,uCAFmB,IAEnB;;;;kBAGgBlL,OAAhB,EAAyBrI,CAAzB;;;;;MAKIuT,cAAJ,EAAoBrL,YAAYG,OAAZ,EAAqBrI,CAArB;;;;gBAINqI,OAAd,EAAuBrI,CAAvB;;;;;aAKWqI,OAAX,EAAoBrI,CAApB;;;eAGaqI,OAAb,EAAsBrI,CAAtB,EAAyBkP,KAAzB;;;oBAGkB7G,OAAlB,EAA2BrI,CAA3B,EAA8BR,GAA9B;;;kBAGgB6I,OAAhB;;;;;;MAMIkL,cAAJ,EAAoBvE,UAAU3G,OAAV,EAAmBrI,CAAnB,EAAsBsT,kBAAtB;;;cAGRjL,OAAZ,EAAqBrI,CAArB;;SAEOqI,OAAP;;;ACtDa,SAASmL,UAAT,CAAoBtE,KAApB,QAAuC;MAAV1P,GAAU,QAAVA,GAAU;MAALQ,CAAK,QAALA,CAAK;;;;MAGhDmS,mBAAmB5S,IAAnB,CAAwB2P,KAAxB,CAAJ,EAAoC;YAC1BuE,kBAAkBvE,KAAlB,EAAyB1P,GAAzB,CAAR;;;;;MAKE0P,MAAM3M,MAAN,GAAe,GAAnB,EAAwB;;QAEhBmR,KAAK1T,EAAE,IAAF,CAAX;QACI0T,GAAGnR,MAAH,KAAc,CAAlB,EAAqB;cACXmR,GAAG3K,IAAH,EAAR;;;;;SAKG0H,UAAUvB,KAAV,EAAiBlP,CAAjB,EAAoBgJ,IAApB,EAAP;;;ACdF,SAAS2K,sBAAT,CAAgCC,UAAhC,EAA4C7K,IAA5C,EAAkD;;;;MAI5C6K,WAAWrR,MAAX,IAAqB,CAAzB,EAA4B;;;;;UAIpBsR,aAAaD,WAAWlL,MAAX,CAAkB,UAACC,GAAD,EAAMmL,SAAN,EAAoB;YACnDA,SAAJ,IAAiBnL,IAAImL,SAAJ,IAAiBnL,IAAImL,SAAJ,IAAiB,CAAlC,GAAsC,CAAvD;eACOnL,GAAP;OAFiB,EAGhB,EAHgB,CAAnB;;kCAME,iBAAgBkL,UAAhB,EACQnL,MADR,CACe,UAACC,GAAD,EAAMf,GAAN,EAAc;YAChBe,IAAI,CAAJ,IAASkL,WAAWjM,GAAX,CAAb,EAA8B;iBACrB,CAACA,GAAD,EAAMiM,WAAWjM,GAAX,CAAN,CAAP;;;eAGKe,GAAP;OANT,EAOU,CAAC,CAAD,EAAI,CAAJ,CAPV,CAVwB;;;;UASnBoL,OATmB;UASVC,SATU;;;;;;;UAuBtBA,aAAa,CAAb,IAAkBD,QAAQxR,MAAR,IAAkB,CAAxC,EAA2C;qBAC5BwG,KAAKqD,KAAL,CAAW2H,OAAX,CAAb;;;UAGIE,YAAY,CAACL,WAAW,CAAX,CAAD,EAAgBA,WAAWvJ,KAAX,CAAiB,CAAC,CAAlB,CAAhB,CAAlB;UACM6J,aAAaD,UAAUvL,MAAV,CAAiB,UAACC,GAAD,EAAM/K,GAAN;eAAc+K,IAAIpG,MAAJ,GAAa3E,IAAI2E,MAAjB,GAA0BoG,GAA1B,GAAgC/K,GAA9C;OAAjB,EAAoE,EAApE,CAAnB;;UAEIsW,WAAW3R,MAAX,GAAoB,EAAxB,EAA4B;;aACnB2R;;;;;WAGFnL;;;;;;;SAGF,IAAP;;;AAGF,SAASoL,oBAAT,CAA8BP,UAA9B,EAA0CpU,GAA1C,EAA+C;;;;;;;mBAO5BE,IAAIC,KAAJ,CAAUH,GAAV,CAP4B;;MAOrCqN,IAPqC,cAOrCA,IAPqC;;MAQvCuH,cAAcvH,KAAKrJ,OAAL,CAAa4O,iBAAb,EAAgC,EAAhC,CAApB;;MAEMiC,YAAYT,WAAW,CAAX,EAAcnH,WAAd,GAA4BjJ,OAA5B,CAAoC,GAApC,EAAyC,EAAzC,CAAlB;MACM8Q,iBAAiBC,MAAMC,WAAN,CAAkBH,SAAlB,EAA6BD,WAA7B,CAAvB;;MAEIE,iBAAiB,GAAjB,IAAwBD,UAAU9R,MAAV,GAAmB,CAA/C,EAAkD;WACzCqR,WAAWvJ,KAAX,CAAiB,CAAjB,EAAoBhM,IAApB,CAAyB,EAAzB,CAAP;;;MAGIoW,UAAUb,WAAWvJ,KAAX,CAAiB,CAAC,CAAlB,EAAqB,CAArB,EAAwBoC,WAAxB,GAAsCjJ,OAAtC,CAA8C,GAA9C,EAAmD,EAAnD,CAAhB;MACMkR,eAAeH,MAAMC,WAAN,CAAkBC,OAAlB,EAA2BL,WAA3B,CAArB;;MAEIM,eAAe,GAAf,IAAsBD,QAAQlS,MAAR,IAAkB,CAA5C,EAA+C;WACtCqR,WAAWvJ,KAAX,CAAiB,CAAjB,EAAoB,CAAC,CAArB,EAAwBhM,IAAxB,CAA6B,EAA7B,CAAP;;;SAGK,IAAP;;;;;AAKF,AAAe,SAASoV,iBAAT,CAA2BvE,KAA3B,EAA4C;MAAV1P,GAAU,uEAAJ,EAAI;;;;MAGnDoU,aAAa1E,MAAM9C,KAAN,CAAY+F,kBAAZ,CAAnB;MACIyB,WAAWrR,MAAX,KAAsB,CAA1B,EAA6B;WACpB2M,KAAP;;;MAGEyF,WAAWhB,uBAAuBC,UAAvB,EAAmC1E,KAAnC,CAAf;MACIyF,QAAJ,EAAc,OAAOA,QAAP;;aAEHR,qBAAqBP,UAArB,EAAiCpU,GAAjC,CAAX;MACImV,QAAJ,EAAc,OAAOA,QAAP;;;;SAIPzF,KAAP;;;AC3FF,IAAM0F,WAAW;UACPvC,WADO;kBAECwC,OAFD;OAGVnC,QAHU;kBAICK,kBAJD;WAKN+B,gBALM;SAMRtB;CANT,CAUA,AAEA,AACA,AACA,AACA,AACA,AACA,AACA;;ACfA;;;;;;;;;;;AAWA,AAAe,SAASuB,eAAT,CAAyB/U,CAAzB,EAA4BgV,IAA5B,EAAkC;;;;;;;MAO3CA,KAAK9O,uBAAT,EAAkC;QAC5BA,wBAAwBlG,CAAxB,CAAJ;;;MAGEyH,oBAAoBzH,CAApB,CAAJ;MACIkL,aAAalL,CAAb,EAAgBgV,KAAKnK,WAArB,CAAJ;MACMoK,gBAAgB7G,iBAAiBpO,CAAjB,CAAtB;;SAEOiV,aAAP;;;AC3BF,IAAMC,0BAA0B;eACjB;6BACc,IADd;iBAEE,IAFF;wBAGS;GAJQ;;;;;;;;;;;;;;;;;;;;;SAAA,yBA0BGF,IA1BH,EA0BS;QAA7BhV,CAA6B,QAA7BA,CAA6B;QAA1BgR,IAA0B,QAA1BA,IAA0B;QAApB9B,KAAoB,QAApBA,KAAoB;QAAb1P,GAAa,QAAbA,GAAa;;wBACzB,KAAK2V,WAAjB,EAAiCH,IAAjC;;QAEIhV,KAAKmC,QAAQC,IAAR,CAAa4O,IAAb,CAAT;;;;QAII3Q,OAAO,KAAK+U,cAAL,CAAoBpV,CAApB,EAAuBkP,KAAvB,EAA8B1P,GAA9B,EAAmCwV,IAAnC,CAAX;;QAEI3D,iBAAiBhR,IAAjB,CAAJ,EAA4B;aACnB,KAAKgV,kBAAL,CAAwBhV,IAAxB,EAA8BL,CAA9B,CAAP;;;;;;;;;;wCAKgB,iBAAgBgV,IAAhB,EAAsBvT,MAAtB,CAA6B;eAAKuT,KAAKM,CAAL,MAAY,IAAjB;OAA7B,CAAlB,4GAAuE;YAA5D1N,GAA4D;;aAChEA,GAAL,IAAY,KAAZ;YACIzF,QAAQC,IAAR,CAAa4O,IAAb,CAAJ;;eAEO,KAAKoE,cAAL,CAAoBpV,CAApB,EAAuBkP,KAAvB,EAA8B1P,GAA9B,EAAmCwV,IAAnC,CAAP;;YAEI3D,iBAAiBhR,IAAjB,CAAJ,EAA4B;;;;;;;;;;;;;;;;;;;WAKvB,KAAKgV,kBAAL,CAAwBhV,IAAxB,EAA8BL,CAA9B,CAAP;GApD4B;;;;gBAAA,0BAwDfA,CAxDe,EAwDZkP,KAxDY,EAwDL1P,GAxDK,EAwDAwV,IAxDA,EAwDM;WAC3BF,iBACGC,gBAAgB/U,CAAhB,EAAmBgV,IAAnB,CADH,EAEL;UAAA;0BAEsBA,KAAK1B,kBAF3B;kBAAA;;KAFK,CAAP;GAzD4B;;;;;;oBAAA,8BAsEXjT,IAtEW,EAsELL,CAtEK,EAsEF;QACtB,CAACK,IAAL,EAAW;aACF,IAAP;;;WAGKiL,gBAAgBtL,EAAEgR,IAAF,CAAO3Q,IAAP,CAAhB,CAAP;;;;;;;CA3EJ,CAqFA;;AC9FA;;;;;;;AAOA,AAAO,IAAMkV,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;QAArB3V,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAboW,SAAa,QAAbA,SAAa;;;;QAGzB1G,cAAJ;;YAEQa,gBAAgB/P,CAAhB,EAAmBuV,sBAAnB,EAA2CK,SAA3C,CAAR;QACI1G,KAAJ,EAAW,OAAOsE,WAAWtE,KAAX,EAAkB,EAAE1P,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;;YAIH6Q,qBAAqB7Q,CAArB,EAAwByV,sBAAxB,CAAR;QACIvG,KAAJ,EAAW,OAAOsE,WAAWtE,KAAX,EAAkB,EAAE1P,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;YAGH+P,gBAAgB/P,CAAhB,EAAmBwV,oBAAnB,EAAyCI,SAAzC,CAAR;QACI1G,KAAJ,EAAW,OAAOsE,WAAWtE,KAAX,EAAkB,EAAE1P,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;YAGH6Q,qBAAqB7Q,CAArB,EAAwB0V,oBAAxB,CAAR;QACIxG,KAAJ,EAAW,OAAOsE,WAAWtE,KAAX,EAAkB,EAAE1P,QAAF,EAAOQ,IAAP,EAAlB,CAAP;;;WAGJ,EAAP;;CAvBJ,CA2BA;;ACxCA;;;;;;AAMA,AAAO,IAAM6V,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;QAAhBlW,CAAgB,QAAhBA,CAAgB;QAAb4V,SAAa,QAAbA,SAAa;;QACpBtD,eAAJ;;;;aAISvC,gBAAgB/P,CAAhB,EAAmB6V,gBAAnB,EAAqCD,SAArC,CAAT;QACItD,UAAUA,OAAO/P,MAAP,GAAgBuT,iBAA9B,EAAiD;aACxCzD,YAAYC,MAAZ,CAAP;;;;aAIOzB,qBAAqB7Q,CAArB,EAAwB+V,gBAAxB,EAA0C,CAA1C,CAAT;QACIzD,UAAUA,OAAO/P,MAAP,GAAgBuT,iBAA9B,EAAiD;aACxCzD,YAAYC,MAAZ,CAAP;;;;;;;;;;wCAK8B2D,mBAAhC,4GAAqD;;;;;YAAzC9R,QAAyC;YAA/BgS,KAA+B;;YAC7C9V,OAAOL,EAAEmE,QAAF,CAAb;YACI9D,KAAKkC,MAAL,KAAgB,CAApB,EAAuB;cACfwG,OAAO1I,KAAK0I,IAAL,EAAb;cACIoN,MAAM5W,IAAN,CAAWwJ,IAAX,CAAJ,EAAsB;mBACbsJ,YAAYtJ,IAAZ,CAAP;;;;;;;;;;;;;;;;;;;WAKC,IAAP;;CA7BJ,CAiCA;;AC9CA;;;;AAIA,AAAO,IAAMqN,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,IAAInY,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAFoC;;;;AAMpC,IAAIA,MAAJ,CAAW,6BAAX,EAA0C,GAA1C,CANoC;;AAQpC,IAAIA,MAAJ,iBAAyBkY,eAAzB,kBAAuD,GAAvD,CARoC,CAA/B;;ACrCP,IAAME,gCAAgC;SAAA,yBACL;QAArBxW,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAboW,SAAa,QAAbA,SAAa;;QACzBa,sBAAJ;;;;oBAIgB1G,gBAAgB/P,CAAhB,EAAmBoW,wBAAnB,EAA6CR,SAA7C,EAAwD,KAAxD,CAAhB;QACIa,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;;oBAIH5F,qBAAqB7Q,CAArB,EAAwBqW,wBAAxB,CAAhB;QACII,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;oBAGHlL,eAAe/L,GAAf,EAAoB+W,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,IAAI1Y,MAAJ,CAAWyY,8BAA8BxY,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAM0Y,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,IAAI5Y,MAAJ,CAAW2Y,8BAA8B1Y,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAM4Y,SAAS,gBAAf;AACP,AAAO,IAAMC,SAAS,kBAAf;;AC3CP,SAASC,MAAT,CAAgB7W,KAAhB,EAAuB;UACXA,MAAME,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCF,MAAME,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;;AAIF,AAAO,SAAS4W,aAAT,CAAuB5X,GAAvB,EAA4B;QAC3BA,IAAIwJ,IAAJ,EAAN;MACIO,QAAQ,CAAZ;;MAEIuN,iCAAiCvX,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;MAGEwX,iCAAiCzX,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;;;MAKEyX,OAAO1X,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;MAGE0X,OAAO3X,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;;;SAKK+J,KAAP;;;;AAIF,AAAO,SAAS8N,SAAT,CAAmBvP,IAAnB,EAAyB;MAC1BA,KAAKtH,IAAL,CAAU,KAAV,CAAJ,EAAsB;WACb,CAAP;;;SAGK,CAAP;;;;;AAKF,AAAO,SAAS8W,cAAT,CAAwBxP,IAAxB,EAA8B;MAC/ByB,QAAQ,CAAZ;MACMgO,aAAazP,KAAKhF,OAAL,CAAa,QAAb,EAAuBuL,KAAvB,EAAnB;;MAEIkJ,WAAWhV,MAAX,KAAsB,CAA1B,EAA6B;aAClB,EAAT;;;MAGIM,UAAUiF,KAAK8C,MAAL,EAAhB;MACI4M,iBAAJ;MACI3U,QAAQN,MAAR,KAAmB,CAAvB,EAA0B;eACbM,QAAQ+H,MAAR,EAAX;;;GAGD/H,OAAD,EAAU2U,QAAV,EAAoBvW,OAApB,CAA4B,UAACX,KAAD,EAAW;QACjC6E,iBAAe5F,IAAf,CAAoB4X,OAAO7W,KAAP,CAApB,CAAJ,EAAwC;eAC7B,EAAT;;GAFJ;;SAMOiJ,KAAP;;;;;AAKF,AAAO,SAASkO,cAAT,CAAwB3P,IAAxB,EAA8B;MAC/ByB,QAAQ,CAAZ;MACMqE,WAAW9F,KAAKnB,IAAL,EAAjB;MACMG,UAAU8G,SAASrP,GAAT,CAAa,CAAb,CAAhB;;MAEIuI,WAAWA,QAAQpE,OAAR,KAAoB,YAAnC,EAAiD;aACtC,EAAT;;;MAGEyC,iBAAe5F,IAAf,CAAoB4X,OAAOvJ,QAAP,CAApB,CAAJ,EAA2C;aAChC,EAAT;;;SAGKrE,KAAP;;;AAGF,AAAO,SAASmO,iBAAT,CAA2B5P,IAA3B,EAAiC;MAClCyB,QAAQ,CAAZ;;MAEMhG,QAAQkG,WAAW3B,KAAKtH,IAAL,CAAU,OAAV,CAAX,CAAd;MACMuH,SAAS0B,WAAW3B,KAAKtH,IAAL,CAAU,QAAV,CAAX,CAAf;MACM8C,MAAMwE,KAAKtH,IAAL,CAAU,KAAV,CAAZ;;;MAGI+C,SAASA,SAAS,EAAtB,EAA0B;aACf,EAAT;;;;MAIEwE,UAAUA,UAAU,EAAxB,EAA4B;aACjB,EAAT;;;MAGExE,SAASwE,MAAT,IAAmB,CAACzE,IAAIpB,QAAJ,CAAa,QAAb,CAAxB,EAAgD;QACxCyV,OAAOpU,QAAQwE,MAArB;QACI4P,OAAO,IAAX,EAAiB;;eACN,GAAT;KADF,MAEO;eACI1N,KAAK2N,KAAL,CAAWD,OAAO,IAAlB,CAAT;;;;SAIGpO,KAAP;;;AAGF,AAAO,SAASsO,eAAT,CAAyBC,KAAzB,EAAgC3W,KAAhC,EAAuC;SACpC2W,MAAMvV,MAAN,GAAe,CAAhB,GAAqBpB,KAA5B;;;ACxGF;;;;;;;;AAQA,IAAM4W,+BAA+B;SAAA,yBACA;QAAzB/X,CAAyB,QAAzBA,CAAyB;QAAtBiC,OAAsB,QAAtBA,OAAsB;QAAb2T,SAAa,QAAbA,SAAa;;QAC7BoC,iBAAJ;;;;;;QAMMC,WACJlI,gBACE/P,CADF,EAEE2W,wBAFF,EAGEf,SAHF,EAIE,KAJF,CADF;;QAQIqC,QAAJ,EAAc;iBACDpD,QAAWoD,QAAX,CAAX;;UAEID,QAAJ,EAAc,OAAOA,QAAP;;;;;;QAMVE,OAAOlY,EAAE,KAAF,EAASiC,OAAT,EAAkBsO,OAAlB,EAAb;QACM4H,YAAY,EAAlB;;SAEKlX,OAAL,CAAa,UAACF,GAAD,EAAMI,KAAN,EAAgB;UACrB2G,OAAO9H,EAAEe,GAAF,CAAb;UACMuC,MAAMwE,KAAKtH,IAAL,CAAU,KAAV,CAAZ;;UAEI,CAAC8C,GAAL,EAAU;;UAENiG,QAAQ6N,cAAc9T,GAAd,CAAZ;eACS+T,UAAUvP,IAAV,CAAT;eACSwP,eAAexP,IAAf,CAAT;eACS2P,eAAe3P,IAAf,CAAT;eACS4P,kBAAkB5P,IAAlB,CAAT;eACS+P,gBAAgBK,IAAhB,EAAsB/W,KAAtB,CAAT;;gBAEUmC,GAAV,IAAiBiG,KAAjB;KAbF;;gCAiBE,iBAAgB4O,SAAhB,EAA2BzP,MAA3B,CAAkC,UAACC,GAAD,EAAMf,GAAN;aAChCuQ,UAAUvQ,GAAV,IAAiBe,IAAI,CAAJ,CAAjB,GAA0B,CAACf,GAAD,EAAMuQ,UAAUvQ,GAAV,CAAN,CAA1B,GAAkDe,GADlB;KAAlC,EAEE,CAAC,IAAD,EAAO,CAAP,CAFF,CA5C+B;;;;QA2C1ByP,MA3C0B;QA2ClB3K,QA3CkB;;;QAgD7BA,WAAW,CAAf,EAAkB;iBACLoH,QAAWuD,MAAX,CAAX;;UAEIJ,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;wCAKOpB,wBAAvB,4GAAiD;YAAtCzS,QAAsC;;YACzC7D,QAAQN,EAAEmE,QAAF,EAAYkK,KAAZ,EAAd;YACM/K,MAAMhD,MAAME,IAAN,CAAW,KAAX,CAAZ;YACI8C,GAAJ,EAAS;qBACIuR,QAAWvR,GAAX,CAAX;cACI0U,QAAJ,EAAc,OAAOA,QAAP;;;YAGVK,OAAO/X,MAAME,IAAN,CAAW,MAAX,CAAb;YACI6X,IAAJ,EAAU;qBACGxD,QAAWwD,IAAX,CAAX;cACIL,QAAJ,EAAc,OAAOA,QAAP;;;YAGVzX,QAAQD,MAAME,IAAN,CAAW,OAAX,CAAd;YACID,KAAJ,EAAW;qBACEsU,QAAWtU,KAAX,CAAX;cACIyX,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;;;;;;;;;WAIX,IAAP;;CA9EJ,CAkFA;;AC3Ge,SAASM,eAAT,CAAyB/O,KAAzB,EAAgCgP,UAAhC,EAA4CF,IAA5C,EAAkD;;;;;;MAM3D9O,QAAQ,CAAZ,EAAe;QACPiP,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;WACOrP,QAAQsP,YAAf;;;SAGK,CAAP;;;ACnBa,SAASC,aAAT,CAAuBjJ,QAAvB,EAAiC3D,OAAjC,EAA0C;;;;;MAKnD3C,QAAQ,CAAZ;;MAEIwC,YAAYxM,IAAZ,CAAiBsQ,SAAS7G,IAAT,EAAjB,CAAJ,EAAuC;QAC/B+P,gBAAgB/Q,SAAS6H,QAAT,EAAmB,EAAnB,CAAtB;;;;QAIIkJ,gBAAgB,CAApB,EAAuB;cACb,CAAC,EAAT;KADF,MAEO;cACG9O,KAAKE,GAAL,CAAS,CAAT,EAAY,KAAK4O,aAAjB,CAAR;;;;;;QAME7M,WAAWA,WAAW6M,aAA1B,EAAyC;eAC9B,EAAT;;;;SAIGxP,KAAP;;;AC5Ba,SAASyP,eAAT,CAAyB9M,OAAzB,EAAkC+M,IAAlC,EAAwC;;;;MAIjD/M,WAAW,CAAC+M,IAAhB,EAAsB;WACb,EAAP;;;SAGK,CAAP;;;ACRK,IAAMC,aAAW,IAAjB;;;;AAIP,AAAO,IAAMzT,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,IAAItH,MAAJ,CAAWqH,wBAAsBpH,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;;;AAKP,AAAO,IAAM8a,sBAAoB,IAAI/a,MAAJ,CAAW,4CAAX,EAAyD,GAAzD,CAA1B;;;;AAIP,AAAO,IAAMgb,qBAAmB,IAAIhb,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAzB;;;;AAIP,AAAO,IAAMib,sBAAoB,IAAIjb,MAAJ,CAAW,yBAAX,EAAsC,GAAtC,CAA1B,CAEP;;ACjCe,SAASkb,oBAAT,CAA8BjB,IAA9B,EAAoC;;MAE7C3S,2BAAyBnG,IAAzB,CAA8B8Y,IAA9B,CAAJ,EAAyC;WAChC,CAAC,EAAR;;;SAGK,CAAP;;;ACAF,SAASkB,SAAT,CAAiBC,KAAjB,EAAwB;UACZA,MAAMhZ,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCgZ,MAAMhZ,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;AAGF,AAAe,SAAS8W,gBAAT,CAAwBkC,KAAxB,EAA+B;;;;MAIxC3W,UAAU2W,MAAM5O,MAAN,EAAd;MACI6O,gBAAgB,KAApB;MACIC,gBAAgB,KAApB;MACInQ,QAAQ,CAAZ;;cAEW7L,MAAM,CAAN,EAAS,CAAT,CAAX,EAAwBuD,OAAxB,CAAgC,YAAM;QAChC4B,QAAQN,MAAR,KAAmB,CAAvB,EAA0B;;;;QAIpBoX,aAAaJ,UAAQ1W,OAAR,EAAiB,GAAjB,CAAnB;;;;QAII,CAAC4W,aAAD,IAAkB9T,QAAQpG,IAAR,CAAaoa,UAAb,CAAtB,EAAgD;sBAC9B,IAAhB;eACS,EAAT;;;;;;QAME,CAACD,aAAD,IAAkBnU,kBAAkBhG,IAAlB,CAAuBoa,UAAvB,CAAlB,IACEjU,2BAAyBnG,IAAzB,CAA8Boa,UAA9B,CADN,EACiD;UAC3C,CAACtU,kBAAkB9F,IAAlB,CAAuBoa,UAAvB,CAAL,EAAyC;wBACvB,IAAhB;iBACS,EAAT;;;;cAIM9W,QAAQ+H,MAAR,EAAV;GAzBF;;SA4BOrB,KAAP;;;AC/Ca,SAASqQ,aAAT,CAAuBC,QAAvB,EAAiC;;;MAG1CR,oBAAkB9Z,IAAlB,CAAuBsa,QAAvB,CAAJ,EAAsC;WAC7B,CAAC,GAAR;;;SAGK,CAAP;;;ACFa,SAASC,WAAT,CACbzB,IADa,EAEbE,UAFa,EAGbwB,OAHa,EAIbta,SAJa,EAKboQ,QALa,EAMbmK,YANa,EAOb;;MAEIA,aAAazY,IAAb,CAAkB;WAAO8W,SAAS7Y,GAAhB;GAAlB,MAA2C4R,SAA/C,EAA0D;WACjD,KAAP;;;;;MAKE,CAACiH,IAAD,IAASA,SAASE,UAAlB,IAAgCF,SAAS0B,OAA7C,EAAsD;WAC7C,KAAP;;;MAGMjc,QAZR,GAYqB2B,SAZrB,CAYQ3B,QAZR;;mBAa+B4B,IAAIC,KAAJ,CAAU0Y,IAAV,CAb/B;;MAakB4B,QAblB,cAaQnc,QAbR;;;;MAgBImc,aAAanc,QAAjB,EAA2B;WAClB,KAAP;;;;;MAKIoc,WAAW7B,KAAK7U,OAAL,CAAauW,OAAb,EAAsB,EAAtB,CAAjB;MACI,CAACb,WAAS3Z,IAAT,CAAc2a,QAAd,CAAL,EAA8B;WACrB,KAAP;;;;;MAKExU,2BAAyBnG,IAAzB,CAA8BsQ,QAA9B,CAAJ,EAA6C;WACpC,KAAP;;;;MAIEA,SAAStN,MAAT,GAAkB,EAAtB,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;ACpDa,SAAS4X,YAAT,CAAsB9B,IAAtB,EAA4B+B,SAA5B,EAAuC;;;;;MAKhD,CAACA,UAAU7a,IAAV,CAAe8Y,IAAf,CAAL,EAA2B;WAClB,CAAC,EAAR;;;SAGK,CAAP;;;ACPa,SAASgC,iBAAT,CAA2BR,QAA3B,EAAqC;;MAE9CV,oBAAkB5Z,IAAlB,CAAuBsa,QAAvB,CAAJ,EAAsC;WAC7B,EAAP;;;SAGK,CAAP;;;ACHa,SAASS,aAAT,CAAuBT,QAAvB,EAAiC;;MAE1CT,mBAAiB7Z,IAAjB,CAAsBsa,QAAtB,CAAJ,EAAqC;;;;;QAK/BV,oBAAkB5Z,IAAlB,CAAuBsa,QAAvB,CAAJ,EAAsC;aAC7B,CAAC,EAAR;;;;SAIG,CAAP;;;ACKK,SAASU,aAAT,CAAuBR,OAAvB,EAAgC;SAC9B,IAAI3b,MAAJ,OAAe2b,OAAf,EAA0B,GAA1B,CAAP;;;AAGF,SAASR,OAAT,CAAiBC,KAAjB,EAAwB3J,QAAxB,EAAkC;UACtBA,YAAY2J,MAAMzQ,IAAN,EAAtB,WAAsCyQ,MAAMhZ,IAAN,CAAW,OAAX,KAAuB,EAA7D,WAAmEgZ,MAAMhZ,IAAN,CAAW,IAAX,KAAoB,EAAvF;;;AAGF,AAAe,SAASga,UAAT,OAOZ;MANDC,KAMC,QANDA,KAMC;MALDlC,UAKC,QALDA,UAKC;MAJDwB,OAIC,QAJDA,OAIC;MAHDta,SAGC,QAHDA,SAGC;MAFDO,CAEC,QAFDA,CAEC;+BADDga,YACC;MADDA,YACC,qCADc,EACd;;cACWva,aAAaC,IAAIC,KAAJ,CAAU4Y,UAAV,CAAzB;MACM6B,YAAYG,cAAcR,OAAd,CAAlB;MACMd,OAAO3H,YAAYtR,CAAZ,CAAb;;;;;;;;;MASM0a,cAAcD,MAAM/R,MAAN,CAAa,UAACiS,aAAD,EAAgBC,IAAhB,EAAyB;;;;QAIlDvC,OAAOlM,aAAayO,KAAK5Z,OAAL,CAAaqX,IAA1B,CAAb;QACMmB,QAAQxZ,EAAE4a,IAAF,CAAd;QACM/K,WAAW2J,MAAMzQ,IAAN,EAAjB;;QAEI,CAAC+Q,YAAYzB,IAAZ,EAAkBE,UAAlB,EAA8BwB,OAA9B,EAAuCta,SAAvC,EAAkDoQ,QAAlD,EAA4DmK,YAA5D,CAAL,EAAgF;aACvEW,aAAP;;;;QAIE,CAACA,cAActC,IAAd,CAAL,EAA0B;oBACVA,IAAd,IAAsB;eACb,CADa;0BAAA;;OAAtB;KADF,MAMO;oBACSA,IAAd,EAAoBxI,QAApB,GAAkC8K,cAActC,IAAd,EAAoBxI,QAAtD,SAAkEA,QAAlE;;;QAGIgL,eAAeF,cAActC,IAAd,CAArB;QACMwB,WAAWN,QAAQC,KAAR,EAAe3J,QAAf,CAAjB;QACM3D,UAAUF,eAAeqM,IAAf,CAAhB;;QAEI9O,QAAQ4Q,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,gBAAgB9M,OAAhB,EAAyB+M,IAAzB,CAAT;aACSH,cAAcjJ,QAAd,EAAwB3D,OAAxB,CAAT;aACSoM,gBAAgB/O,KAAhB,EAAuBgP,UAAvB,EAAmCF,IAAnC,CAAT;;iBAEa9O,KAAb,GAAqBA,KAArB;;WAEOoR,aAAP;GAvCkB,EAwCjB,EAxCiB,CAApB;;SA0CO,iBAAgBD,WAAhB,EAA6BnY,MAA7B,KAAwC,CAAxC,GAA4C,IAA5C,GAAmDmY,WAA1D;;;AClFF;;AAEA,IAAMI,8BAA8B;SAAA,yBACgB;QAAxC9a,CAAwC,QAAxCA,CAAwC;QAArCR,GAAqC,QAArCA,GAAqC;QAAhCC,SAAgC,QAAhCA,SAAgC;iCAArBua,YAAqB;QAArBA,YAAqB,qCAAN,EAAM;;gBACpCva,aAAaC,IAAIC,KAAJ,CAAUH,GAAV,CAAzB;;QAEM+Y,aAAapM,aAAa3M,GAAb,CAAnB;QACMua,UAAUrN,eAAelN,GAAf,EAAoBC,SAApB,CAAhB;;QAEMgb,QAAQza,EAAE,SAAF,EAAauQ,OAAb,EAAd;;QAEMwK,cAAcP,WAAW;kBAAA;4BAAA;sBAAA;0BAAA;UAAA;;KAAX,CAApB;;;QAUI,CAACO,WAAL,EAAkB,OAAO,IAAP;;;;QAIZC,UAAU,iBAAgBD,WAAhB,EAA6BrS,MAA7B,CAAoC,UAACC,GAAD,EAAMiS,IAAN,EAAe;UAC3DK,aAAaF,YAAYH,IAAZ,CAAnB;aACOK,WAAW1R,KAAX,GAAmBZ,IAAIY,KAAvB,GAA+B0R,UAA/B,GAA4CtS,GAAnD;KAFc,EAGb,EAAEY,OAAO,CAAC,GAAV,EAHa,CAAhB;;;;QAOIyR,QAAQzR,KAAR,IAAiB,EAArB,EAAyB;aAChByR,QAAQ3C,IAAf;;;WAGK,IAAP;;CAlCJ,CAuCA;;AClDO,IAAM6C,2BAA2B,CACtC,QADsC,CAAjC;;ACKP,SAASC,WAAT,CAAqB3b,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;YAEG2b,YAAY3b,GAAZ;GAFV;;;AAMF,IAAM4b,sBAAsB;SAAA,yBACK;QAArBpb,CAAqB,QAArBA,CAAqB;QAAlBR,GAAkB,QAAlBA,GAAkB;QAAboW,SAAa,QAAbA,SAAa;;QACvByF,aAAarb,EAAE,qBAAF,CAAnB;QACIqb,WAAW9Y,MAAX,KAAsB,CAA1B,EAA6B;UACrB8V,OAAOgD,WAAW7a,IAAX,CAAgB,MAAhB,CAAb;UACI6X,IAAJ,EAAU;eACDtW,OAAOsW,IAAP,CAAP;;;;QAIEiD,UAAUvL,gBAAgB/P,CAAhB,EAAmBkb,wBAAnB,EAA6CtF,SAA7C,CAAhB;QACI0F,OAAJ,EAAa;aACJvZ,OAAOuZ,OAAP,CAAP;;;WAGKvZ,OAAOvC,GAAP,CAAP;;CAfJ,CAoBA;;ACtCO,IAAM+b,yBAAyB,CACpC,gBADoC,EAEpC,qBAFoC,CAA/B;;ACSA,SAAS5Z,OAAT,CAAeM,OAAf,EAAwBjC,CAAxB,EAA4C;MAAjBwb,SAAiB,uEAAL,GAAK;;YACvCvZ,QAAQuB,OAAR,CAAgB,UAAhB,EAA4B,GAA5B,EAAiCwF,IAAjC,EAAV;SACOyS,UAAUxZ,OAAV,EAAmBuZ,SAAnB,EAA8B,EAAEE,SAAS,UAAX,EAA9B,CAAP;;;AAGF,IAAMC,0BAA0B;SAAA,yBACK;QAAzB3b,CAAyB,QAAzBA,CAAyB;QAAtBiC,OAAsB,QAAtBA,OAAsB;QAAb2T,SAAa,QAAbA,SAAa;;QAC3BgG,UAAU7L,gBAAgB/P,CAAhB,EAAmBub,sBAAnB,EAA2C3F,SAA3C,CAAhB;QACIgG,OAAJ,EAAa;aACJja,QAAM8O,UAAUmL,OAAV,EAAmB5b,CAAnB,CAAN,CAAP;;;QAGIwb,YAAY,GAAlB;QACMK,eAAe5Z,QAAQoI,KAAR,CAAc,CAAd,EAAiBmR,YAAY,CAA7B,CAArB;WACO7Z,QAAM3B,EAAE6b,YAAF,EAAgB9S,IAAhB,EAAN,EAA8B/I,CAA9B,EAAiCwb,SAAjC,CAAP;;CATJ,CAaA;;ACvBA,IAAMM,4BAA4B;SAAA,yBACX;QAAX7Z,OAAW,QAAXA,OAAW;;QACbjC,IAAImC,QAAQC,IAAR,CAAaH,OAAb,CAAV;;QAEM8G,OAAOuC,gBAAgBtL,EAAE,KAAF,EAASqO,KAAT,GAAiBtF,IAAjB,EAAhB,CAAb;WACOA,KAAKqD,KAAL,CAAW,IAAX,EAAiB7J,MAAxB;;CALJ,CASA;;ACCA,IAAMwZ,mBAAmB;;UAEf,GAFe;SAGhBpG,sBAAsBqG,OAHN;kBAIPxF,8BAA8BwF,OAJvB;UAKf9F,uBAAuB8F,OALR;WAMd9G,wBAAwB8G,OAAxB,CAAgCC,IAAhC,CAAqC/G,uBAArC,CANc;kBAOP6C,6BAA6BiE,OAPtB;OAQlBtF,oBAAoBsF,OARF;iBASRlB,4BAA4BkB,OATpB;kBAUPZ,oBAAoBY,OAVb;WAWdL,wBAAwBK,OAXV;cAYXF,0BAA0BE,OAZf;aAaZ;QAAG9M,KAAH,QAAGA,KAAH;WAAegN,gBAAgBC,YAAhB,CAA6BjN,KAA7B,CAAf;GAbY;;SAAA,mBAef1Q,OAfe,EAeN;QACPwS,IADO,GACExS,OADF,CACPwS,IADO;;;QAGXA,IAAJ,EAAU;UACFhR,IAAImC,QAAQC,IAAR,CAAa4O,IAAb,CAAV;cACQhR,CAAR,GAAYA,CAAZ;;;QAGIkP,QAAQ,KAAKA,KAAL,CAAW1Q,OAAX,CAAd;QACM4d,iBAAiB,KAAKA,cAAL,CAAoB5d,OAApB,CAAvB;QACM8T,SAAS,KAAKA,MAAL,CAAY9T,OAAZ,CAAf;QACMyD,UAAU,KAAKA,OAAL,cAAkBzD,OAAlB,IAA2B0Q,YAA3B,IAAhB;QACMmN,iBAAiB,KAAKA,cAAL,cAAyB7d,OAAzB,IAAkCyD,gBAAlC,IAAvB;QACM0Q,MAAM,KAAKA,GAAL,cAAcnU,OAAd,IAAuByD,gBAAvB,IAAZ;QACMqa,gBAAgB,KAAKA,aAAL,CAAmB9d,OAAnB,CAAtB;QACMod,UAAU,KAAKA,OAAL,cAAkBpd,OAAlB,IAA2ByD,gBAA3B,IAAhB;QACMsa,aAAa,KAAKA,UAAL,cAAqB/d,OAArB,IAA8ByD,gBAA9B,IAAnB;QACMua,YAAY,KAAKA,SAAL,CAAe,EAAEtN,YAAF,EAAf,CAAlB;;0BACwB,KAAKuN,cAAL,CAAoBje,OAApB,CAlBT;;QAkBPgB,GAlBO,mBAkBPA,GAlBO;QAkBFkd,MAlBE,mBAkBFA,MAlBE;;;WAoBR;kBAAA;oBAAA;sBAGWN,kBAAkB,IAH7B;cAAA;oCAAA;sBAAA;kCAAA;cAAA;oBAAA;sBAAA;4BAAA;;KAAP;;CAnCJ,CAoDA;;AC7De,SAASO,YAAT,CAAsBnd,GAAtB,EAA2BC,SAA3B,EAAsC;cACvCA,aAAaC,IAAIC,KAAJ,CAAUH,GAAV,CAAzB;mBACqBC,SAF8B;MAE3C3B,QAF2C,cAE3CA,QAF2C;;MAG7C8e,aAAa9e,SAASsO,KAAT,CAAe,GAAf,EAAoB/B,KAApB,CAA0B,CAAC,CAA3B,EAA8BhM,IAA9B,CAAmC,GAAnC,CAAnB;;SAEOwF,WAAW/F,QAAX,KAAwB+F,WAAW+Y,UAAX,CAAxB,IAAkDb,gBAAzD;;;ACJF;AACA,AAAO,SAASc,gBAAT,CAA0BpN,QAA1B,EAAoCzP,CAApC,QAAkD;MAAT2B,KAAS,QAATA,KAAS;;MACnD,CAACA,KAAL,EAAY,OAAO8N,QAAP;;IAEV9N,MAAMtD,IAAN,CAAW,GAAX,CAAF,EAAmBoR,QAAnB,EAA6B/N,MAA7B;;SAEO+N,QAAP;;;;AAIF,AAAO,SAASqN,iBAAT,CAA2BrN,QAA3B,EAAqCzP,CAArC,SAAwD;MAAd+c,UAAc,SAAdA,UAAc;;MACzD,CAACA,UAAL,EAAiB,OAAOtN,QAAP;;mBAEDsN,UAAhB,EAA4B9b,OAA5B,CAAoC,UAAC2G,GAAD,EAAS;QACrCoV,WAAWhd,EAAE4H,GAAF,EAAO6H,QAAP,CAAjB;QACMlP,QAAQwc,WAAWnV,GAAX,CAAd;;;QAGI,OAAOrH,KAAP,KAAiB,QAArB,EAA+B;eACpBJ,IAAT,CAAc,UAACgB,KAAD,EAAQd,IAAR,EAAiB;sBACfL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0B+c,WAAWnV,GAAX,CAA1B;OADF;KADF,MAIO,IAAI,OAAOrH,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;;SAqBO0N,QAAP;;;AAGF,SAASwN,oBAAT,CAA8Bjd,CAA9B,EAAiC8Q,SAAjC,EAA4C;SACnCA,UAAUvP,IAAV,CAAe,UAAC4C,QAAD,EAAc;QAC9B+Y,MAAMC,OAAN,CAAchZ,QAAd,CAAJ,EAA6B;qCACTA,QADS;;UACpBiZ,CADoB;UACjB5c,IADiB;;aAEpBR,EAAEod,CAAF,EAAK7a,MAAL,KAAgB,CAAhB,IAAqBvC,EAAEod,CAAF,EAAK5c,IAAL,CAAUA,IAAV,CAArB,IAAwCR,EAAEod,CAAF,EAAK5c,IAAL,CAAUA,IAAV,EAAgBwI,IAAhB,OAA2B,EAA1E;;;WAGKhJ,EAAEmE,QAAF,EAAY5B,MAAZ,KAAuB,CAAvB,IAA4BvC,EAAEmE,QAAF,EAAY4E,IAAZ,GAAmBC,IAAnB,OAA8B,EAAjE;GANK,CAAP;;;AAUF,AAAO,SAASqU,MAAT,CAAgBrI,IAAhB,EAAsB;MACnBhV,CADmB,GAC8BgV,IAD9B,CACnBhV,CADmB;MAChBoB,IADgB,GAC8B4T,IAD9B,CAChB5T,IADgB;MACVkc,cADU,GAC8BtI,IAD9B,CACVsI,cADU;0BAC8BtI,IAD9B,CACMuI,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,CASR/J,cATQ;MASRA,cATQ,yCASS,IATT;;;MAWrBiK,mBAAmBP,qBAAqBjd,CAArB,EAAwB8Q,SAAxB,CAAzB;;MAEI,CAAC0M,gBAAL,EAAuB,OAAO,IAAP;;;;;;;;MAQnBD,WAAJ,EAAiB;QACX9N,WAAWzP,EAAEwd,gBAAF,CAAf;;;aAGSC,IAAT,CAAczd,EAAE,aAAF,CAAd;eACWyP,SAAS7E,MAAT,EAAX;;eAEWkS,kBAAkBrN,QAAlB,EAA4BzP,CAA5B,EAA+Bsd,cAA/B,CAAX;eACWT,iBAAiBpN,QAAjB,EAA2BzP,CAA3B,EAA8Bsd,cAA9B,CAAX;;eAEW1I,SAASxT,IAAT,EAAeqO,QAAf,eAA8BuF,IAA9B,IAAoCzB,8BAApC,IAAX;;WAEOvT,EAAEgR,IAAF,CAAOvB,QAAP,CAAP;;;MAGE1N,eAAJ;;;;MAIImb,MAAMC,OAAN,CAAcK,gBAAd,CAAJ,EAAqC;2CACVA,gBADU;;QAC5BrZ,QAD4B;QAClB3D,IADkB;;aAE1BR,EAAEmE,QAAF,EAAY3D,IAAZ,CAAiBA,IAAjB,EAAuBwI,IAAvB,EAAT;GAFF,MAGO;aACIhJ,EAAEwd,gBAAF,EAAoBzU,IAApB,GAA2BC,IAA3B,EAAT;;;;;MAKEuK,cAAJ,EAAoB;WACXqB,SAASxT,IAAT,EAAeW,MAAf,EAAuBiT,IAAvB,CAAP;;;SAGKjT,MAAP;;;AAGF,SAAS2b,aAAT,CAAuB1I,IAAvB,EAA6B;MACnB5T,IADmB,GACkB4T,IADlB,CACnB5T,IADmB;MACbuc,SADa,GACkB3I,IADlB,CACb2I,SADa;uBACkB3I,IADlB,CACF4I,QADE;MACFA,QADE,kCACS,IADT;;;MAGrB7b,SAASsb,oBAAYrI,IAAZ,IAAkBsI,gBAAgBK,UAAUvc,IAAV,CAAlC,IAAf;;;MAGIW,MAAJ,EAAY;WACHA,MAAP;;;;;MAKE6b,QAAJ,EAAc,OAAO7B,iBAAiB3a,IAAjB,EAAuB4T,IAAvB,CAAP;;SAEP,IAAP;;;AAGF,IAAM6I,gBAAgB;SAAA,qBACwB;QAApCF,SAAoC,uEAAxB5B,gBAAwB;QAAN/G,IAAM;gBACFA,IADE;QAClC8I,WADkC,SAClCA,WADkC;QACrBC,cADqB,SACrBA,cADqB;;;QAGtCJ,UAAUjB,MAAV,KAAqB,GAAzB,EAA8B,OAAOiB,UAAU3B,OAAV,CAAkBhH,IAAlB,CAAP;;wBAGzBA,IADL;;;;QAKI8I,WAAJ,EAAiB;UACT7b,WAAUyb,2BACX1I,IADW,IACL5T,MAAM,SADD,EACYmc,aAAa,IADzB,EAC+BrO,OAAO6O;SADtD;aAGO;;OAAP;;QAII7O,QAAQwO,2BAAmB1I,IAAnB,IAAyB5T,MAAM,OAA/B,IAAd;QACMgb,iBAAiBsB,2BAAmB1I,IAAnB,IAAyB5T,MAAM,gBAA/B,IAAvB;QACMkR,SAASoL,2BAAmB1I,IAAnB,IAAyB5T,MAAM,QAA/B,IAAf;QACMkb,gBAAgBoB,2BAAmB1I,IAAnB,IAAyB5T,MAAM,eAA/B,IAAtB;QACMa,UAAUyb,2BACX1I,IADW,IACL5T,MAAM,SADD,EACYmc,aAAa,IADzB,EAC+BrO;OAD/C;QAGMmN,iBAAiBqB,2BAAmB1I,IAAnB,IAAyB5T,MAAM,gBAA/B,EAAiDa,gBAAjD,IAAvB;QACM0Q,MAAM+K,2BAAmB1I,IAAnB,IAAyB5T,MAAM,KAA/B,EAAsCa,gBAAtC,IAAZ;QACM2Z,UAAU8B,2BAAmB1I,IAAnB,IAAyB5T,MAAM,SAA/B,EAA0Ca,gBAA1C,IAAhB;QACMsa,aAAamB,2BAAmB1I,IAAnB,IAAyB5T,MAAM,YAA/B,EAA6Ca,gBAA7C,IAAnB;QACMua,YAAYkB,2BAAmB1I,IAAnB,IAAyB5T,MAAM,WAA/B,EAA4C8N,YAA5C,IAAlB;;gBAEEwO,2BAAmB1I,IAAnB,IAAyB5T,MAAM,gBAA/B,QAAsD,EAAE5B,KAAK,IAAP,EAAakd,QAAQ,IAArB,EA/Bd;;QA8BlCld,GA9BkC,SA8BlCA,GA9BkC;QA8B7Bkd,MA9B6B,SA8B7BA,MA9B6B;;;WAiCnC;kBAAA;sBAAA;oBAAA;oCAAA;oCAAA;cAAA;kCAAA;cAAA;oBAAA;sBAAA;4BAAA;;KAAP;;CAlCJ,CAmDA;;AC5KA;wDAAe;QAEXJ,aAFW,SAEXA,aAFW;QAGXtL,IAHW,SAGXA,IAHW;QAIXhR,CAJW,SAIXA,CAJW;QAKX4V,SALW,SAKXA,SALW;QAMX7T,MANW,SAMXA,MANW;QAOXic,SAPW,SAOXA,SAPW;QAQX9O,KARW,SAQXA,KARW;QASX1P,GATW,SASXA,GATW;;;;;;;iBAAA,GAaD,CAbC;wBAAA,GAcQ,CAAC2M,aAAa3M,GAAb,CAAD,CAdR;;;;;;kBAkBN8c,iBAAiB2B,QAAQ,EAlBnB;;;;;qBAmBF,CAAT;;mBACUrc,SAASsc,MAAT,CAAgB5B,aAAhB,CApBC;;;aAAA;;mBAqBJtc,EAAEgR,IAAF,EAAP;;yBArBW,GAuBW;mBACfsL,aADe;wBAAA;kBAAA;kCAAA;2BAKP,IALO;8BAMJpN,KANI;;aAvBX;0BAAA,GAiCY2O,cAAc7B,OAAd,CAAsBgC,SAAtB,EAAiCG,aAAjC,CAjCZ;;;yBAmCE/Q,IAAb,CAAkBkP,aAAlB;kCAEKva,MADL;sCAGMA,OAAOE,OADX,yCAGagc,KAHb,uBAIIG,eAAenc,OAJnB;;;4BAQcmc,eAAe9B,aAA/B;;;;;sBA9CW,GAiDMP,iBAAiBQ,UAAjB,CAA4B,EAAEta,mBAAiBF,OAAOE,OAAxB,WAAF,EAA5B,CAjDN;0DAmDRF,MAnDQ;2BAoDEkc,KApDF;8BAqDKA,KArDL;;;;;;;;;;GAAf;;WAA8BI,eAA9B;;;;SAA8BA,eAA9B;;;ACKA,IAAMC,UAAU;OAAA,iBACF9e,GADE,EACGwR,IADH,EACoB;;;QAAXgE,IAAW,uEAAJ,EAAI;;;;;;;;oCAI5BA,IAJ4B,CAE9BuJ,aAF8B;2BAAA,uCAEd,IAFc;+BAI5BvJ,IAJ4B,CAG9B4I,QAH8B;sBAAA,kCAGnB,IAHmB;uBAAA,GAMdle,IAAIC,KAAJ,CAAUH,GAAV,CANc;;kBAQ3B3B,YAAY4B,SAAZ,CAR2B;;;;;+CASvB1B,OAAO8B,MATgB;;;uBAAA,GAYd8c,aAAand,GAAb,EAAkBC,SAAlB,CAZc;;;;qBAehBmC,SAASsc,MAAT,CAAgB1e,GAAhB,EAAqBwR,IAArB,EAA2BvR,SAA3B,CAfgB;;;eAAA;;mBAkB5BO,EAAEb,KAlB0B;;;;;+CAmBvBa,CAnBuB;;;;qBAsBzBA,EAAEgR,IAAF,EAAP;;;;uBAtBgC,GA0BdhR,EAAE,MAAF,EAAUkE,GAAV,CAAc,UAAC9D,CAAD,EAAIC,IAAJ;uBAAaL,EAAEK,IAAF,EAAQG,IAAR,CAAa,MAAb,CAAb;eAAd,EAAiD+P,OAAjD,EA1Bc;oBAAA,GA4BnBsN,cAAc7B,OAAd,CAAsBgC,SAAtB,EAAiC,EAAExe,QAAF,EAAOwR,UAAP,EAAahR,IAAb,EAAgB4V,oBAAhB,EAA2BnW,oBAA3B,EAAsCme,kBAAtC,EAAjC,CA5BmB;wBA6BC7b,MA7BD;mBAAA,WA6BxBmN,KA7BwB;2BAAA,WA6BjBoN,aA7BiB;;;;oBAgC5BiC,iBAAiBjC,aAhCW;;;;;;qBAiCf+B,gBACb;oCAAA;4CAAA;0BAAA;oBAAA;oCAAA;8BAAA;4BAAA;;eADa,CAjCe;;;oBAAA;;;;;oCA+CzBtc,MADL;6BAEe,CAFf;gCAGkB;;;;+CAIbA,MArDyB;;;;;;;;;GADpB;;;;;eAAA,yBA2DMvC,GA3DN,EA2DW;;;;;;;;;qBACVoC,SAASsc,MAAT,CAAgB1e,GAAhB,CADU;;;;;;;;;;;;;CA3D3B,CAiEA;;"} \ No newline at end of file diff --git a/fixtures/www.msn.com/1475506925474.html b/fixtures/www.msn.com/1475506925474.html new file mode 100644 index 00000000..0d4f2b3a --- /dev/null +++ b/fixtures/www.msn.com/1475506925474.html @@ -0,0 +1 @@ +<?xml version="1.0" encoding="UTF-8" ?> <!DOCTYPE HTML PUBLIC "-//WAPFORUM//DTD XHTML Mobile 1.2//EN" "http://www.openmobilealliance.org/tech/DTD/xhtml-mobile12.dtd"> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en-US" dir="ltr"> <head> <meta name="application-name" value="This Is Your Brain On Sad Movies; Plus 5 Films To Cry To"> <link rel="canonical" href="http://www.medicaldaily.com/why-we-love-sad-movies-watching-tearjerkers-boosts-pain-tolerance-and-social-398615"> <link rel="alternate" hreflang="en-us" href="http://www.msn.com/en-us/health/wellness/this-is-your-brain-on-sad-movies-plus-5-films-to-cry-to/ar-BBwsPWG?li=BBnb2gg"> <title>This Is Your Brain On Sad Movies; Plus 5 Films To Cry To
You are using an older browser version. Please use a supported version for the best MSN experience.

This Is Your Brain On Sad Movies; Plus 5 Films To Cry To

Medical Daily Medical Daily 9/21/2016 Lizette Borreli

© Archive Photos/Moveipix/Getty Images The psychological reason why we love to watch sad movies is linked to the release of endorphins.

Most of us have sat through a screening of the film Titanic, once, twice, or perhaps more times than we're willing to admit — with a box of tissues. Despite its being one of Hollywood's most tragic love stories, we continue to rewatch the tale of two star-crossed lovers meeting on the ill-fated boat. Now researchers at the University of Oxford suggest our masochistic movie tendencies have surprising benefits: boosts in social bonding, and our threshold for physical pain.

"But why should we be just as engaged by emotionally stirring plots that ‘reduce us to tears’ (i.e. tragedies)?" wrote the researchers, in the paper.

Previous research has found our enjoyment for comedy might stem from the fact that laughter activates the endorphin system, giving us a sense of reward and pleasure. Endorphins are feel-good chemicals secreted within the brain and nervous system affecting a number of physiological functions. They act as analgesics and increase the tolerance for pain. This is why comedy makes us laugh out loud and leads to an increase in our pain tolerance, but why do we keep repeatedly returning to tearjerker movies?

The researchers found watching a sad movie may harness the same system as comedies do. Films like Titanic boost feelings of social bonding, as well as increasing pain tolerance by upping endorphin levels in the brain. The emotional pain derived from watching tragedy — which is different from actually experiencing a tragedy in real life — triggers the endorphin system, which as a result, leads us to experience happiness.

In the study, published in the journal Royal Society Open Science, Robin Dunbar, co-author of the study and professor of evolutionary psychology at the University of Oxford, and his colleagues sought to explore whether our love for storytelling is underpinned by an endorphin-related bonding mechanism.

The psychological reason why we love to watch sad movies is linked to the release of endorphins. © Photo courtesy of Pexels, Public Domain The psychological reason why we love to watch sad movies is linked to the release of endorphins.

A total of 169 participants were recruited to watch Stuart: A Life Backwards , a made-for-TV film based on the true story of a disabled, homeless drug addict and alcoholic. A control group of 68 individuals were shown two documentaries back to back: an episode of The Museum of Life, a behind-the-scenes look at the Natural History Museum in London, and Landscape Mysteries “In Search of Irish Gold,” which explores Irish geology and archaeology.

Before and after watching the films, all the participants took two tests: One measured their sense of belonging or bonding with their fellow audience members. Another was a measure of pain sensitivity, called the Roman chair — a well-established proxy for endorphin release, according to Dunbar. In it, participants brace themselves unsupported in a chairlike stance against a wall until their leg muscles burn painfully. The higher a person’s endorphin level, the longer they should be able to sustain the posture.

The findings revealed those who had watched Stuart: A Life Backwards were able to maintain the Roman chair roughly 18 percent longer than they had in their initial baseline test, compared with those who had watched the documentaries. There was also a correlation between sad movies and a sense of social bonding that was not seen in the control group.

It's important to note the Roman chair pain sensitivity test did not directly measure endorphin release, which leaves the possibility for other factors to influence the increase in social bonding. For example, not everyone who saw the Stuart: A Life Backwards experienced an increase in their sense of bonding; it actually made them more sensitive to pain. This is not surprising since movies are subjective to individual taste.

A similar 2012 study in the journal Communication Research suggests we enjoy watching sad movies like Titanic because they deliver an unexpected emotion: short-term happiness. Watching a tragedy causes us to think about our own close relationships, which in turn boosts our life happiness. In other words, it seems that a negative experience can make us happier by making us more appreciative of the positive aspects in our own lives.

So, for those looking to have a good cry and boost social bonding here is our list of the five must-watch tearjerkers: Schindler's List; Titanic; Life Is Beautiful; The Notebook; and 12 Years a Slave .

Source: Dunbar RIM, Teasdale B, Thompson J et al. Emotional arousal when watching drama increases pain threshold and social bonding. Royal Society Open Science. 2016.

image beaconimage beaconimage beacon
diff --git a/src/extractors/all.js b/src/extractors/all.js index 6c2699c2..2f009e96 100644 --- a/src/extractors/all.js +++ b/src/extractors/all.js @@ -6,6 +6,8 @@ import { NYTimesExtractor } from './custom/www.nytimes.com'; import { TheAtlanticExtractor } from './custom/www.theatlantic.com'; import { NewYorkerExtractor } from './custom/www.newyorker.com'; import { WiredExtractor } from './custom/www.wired.com'; +import { MSNExtractor } from './custom/www.msn.com'; + const Extractors = { 'nymag.com': NYMagExtractor, @@ -16,6 +18,7 @@ const Extractors = { 'www.theatlantic.com': TheAtlanticExtractor, 'www.newyorker.com': NewYorkerExtractor, 'www.wired.com': WiredExtractor, + 'www.msn.com': MSNExtractor, }; diff --git a/src/extractors/custom/www.msn.com/index.js b/src/extractors/custom/www.msn.com/index.js new file mode 100644 index 00000000..558db25c --- /dev/null +++ b/src/extractors/custom/www.msn.com/index.js @@ -0,0 +1,61 @@ +// Rename CustomExtractor +// to fit your publication +// (e.g., NYTimesExtractor) +export const MSNExtractor = { + domain: 'www.msn.com', + title: { + selectors: [ + 'h1', + // enter title selectors + ], + }, + + author: { + selectors: [ + 'span.authorname-txt', + // enter author selectors + ], + }, + + content: { + selectors: [ + 'div.richtext', + // enter content selectors + ], + + // Is there anything in the content you selected that needs transformed + // before it's consumable content? E.g., unusual lazy loaded images + transforms: [ + ], + + // Is there anything that is in the result that shouldn't be? + // The clean selectors will remove anything that matches from + // the result + clean: [ + 'span.caption', + + ], + }, + + date_published: { + selectors: [ + 'span.time', + ], + }, + + lead_image_url: { + selectors: [ + + ], + }, + + dek: { + selectors: [ + ['meta[name="description"]', 'value'], + ], + }, + + next_page_url: null, + + excerpt: null, +}; diff --git a/src/extractors/custom/www.msn.com/index.test.js b/src/extractors/custom/www.msn.com/index.test.js new file mode 100644 index 00000000..d9e3271d --- /dev/null +++ b/src/extractors/custom/www.msn.com/index.test.js @@ -0,0 +1,134 @@ +import assert from 'assert'; +import fs from 'fs'; +import URL from 'url'; +import cheerio from 'cheerio'; + +import Mercury from 'mercury'; +import getExtractor from 'extractors/get-extractor'; + +// Rename CustomExtractor +describe('MSNExtractor', () => { + it('is selected properly', () => { + // To pass this test, rename your extractor in + // ./src/extractors/custom/www.msn.com/index.js + // (e.g., CustomExtractor => NYTimesExtractor) + // then add your new extractor to + // src/extractors/all.js + const url = + 'http://www.msn.com/en-us/health/wellness/this-is-your-brain-on-sad-movies-plus-5-films-to-cry-to/ar-BBwsPWG?li=BBnb2gg'; + const extractor = getExtractor(url); + assert.equal(extractor.domain, URL.parse(url).hostname); + }); + + it('returns the title', ((async)) () => { + // To pass this test, fill out the title selector + // in ./src/extractors/custom/www.msn.com/index.js. + const html = + fs.readFileSync('./fixtures/www.msn.com/1475505231797.html'); + const articleUrl = + 'http://www.msn.com/en-us/health/wellness/this-is-your-brain-on-sad-movies-plus-5-films-to-cry-to/ar-BBwsPWG?li=BBnb2gg'; + + const { title } = + await Mercury.parse(articleUrl, html, { fallback: false }); + + // Update these values with the expected values from + // the article. + assert.equal(title, 'This Is Your Brain On Sad Movies; Plus 5 Films To Cry To'); + }); + + + it('returns the author', ((async)) () => { + // To pass this test, fill out the author selector + // in ./src/extractors/custom/www.msn.com/index.js. + const html = + fs.readFileSync('./fixtures/www.msn.com/1475505231797.html'); + const articleUrl = + 'http://www.msn.com/en-us/health/wellness/this-is-your-brain-on-sad-movies-plus-5-films-to-cry-to/ar-BBwsPWG?li=BBnb2gg'; + + const { author } = + await Mercury.parse(articleUrl, html, { fallback: false }); + + // Update these values with the expected values from + // the article. + assert.equal(author, 'Lizette Borreli'); + }); + + + it('returns the date_published', ((async)) () => { + // To pass this test, fill out the date_published selector + // in ./src/extractors/custom/www.msn.com/index.js. + const html = + fs.readFileSync('./fixtures/www.msn.com/1475505231797.html'); + const articleUrl = + 'http://www.msn.com/en-us/health/wellness/this-is-your-brain-on-sad-movies-plus-5-films-to-cry-to/ar-BBwsPWG?li=BBnb2gg'; + + const { date_published } = + await Mercury.parse(articleUrl, html, { fallback: false }); + + // Update these values with the expected values from + // the article. + assert.equal(date_published, '2016-09-21T04:00:00.000Z'); + }); + + + it('returns the dek', ((async)) () => { + // To pass this test, fill out the dek selector + // in ./src/extractors/custom/www.msn.com/index.js. + const html = + fs.readFileSync('./fixtures/www.msn.com/1475505231797.html'); + const articleUrl = + 'http://www.msn.com/en-us/health/wellness/this-is-your-brain-on-sad-movies-plus-5-films-to-cry-to/ar-BBwsPWG?li=BBnb2gg'; + + const { dek } = + await Mercury.parse(articleUrl, html, { fallback: false }); + + // Update these values with the expected values from + // the article. + assert.equal(dek, 'The psychological reason why we love to watch sad movies is linked to the release of endorphins.'); + }); + + + it('returns the lead_image_url', ((async)) () => { + // To pass this test, fill out the lead_image_url selector + // in ./src/extractors/custom/www.msn.com/index.js. + const html = + fs.readFileSync('./fixtures/www.msn.com/1475505231797.html'); + const articleUrl = + 'http://www.msn.com/en-us/health/wellness/this-is-your-brain-on-sad-movies-plus-5-films-to-cry-to/ar-BBwsPWG?li=BBnb2gg'; + + const { lead_image_url } = + await Mercury.parse(articleUrl, html, { fallback: false }); + + // Update these values with the expected values from + // the article. + assert.equal(lead_image_url, ''); + }); + + + it('returns the content', ((async)) () => { + // To pass this test, fill out the content selector + // in ./src/extractors/custom/www.msn.com/index.js. + // You may also want to make use of the clean and transform + // options. + const html = + fs.readFileSync('./fixtures/www.msn.com/1475505231797.html'); + const url = + 'http://www.msn.com/en-us/health/wellness/this-is-your-brain-on-sad-movies-plus-5-films-to-cry-to/ar-BBwsPWG?li=BBnb2gg'; + + const { content } = + await Mercury.parse(url, html, { fallback: false }); + + const $ = cheerio.load(content || ''); + + const first13 = $('*').first() + .text() + .trim() + .split(/\s+/) + .slice(0, 13) + .join(' '); + + // Update these values with the expected values from + // the article. + assert.equal(first13, 'The psychological reason why we love to watch sad movies is linked to'); + }); +});