diff --git a/.eslintrc b/.eslintrc
index 9b49b43c..0f8fa328 100644
--- a/.eslintrc
+++ b/.eslintrc
@@ -20,6 +20,7 @@
"func-names": 0,
// "no-useless-escape": 0,
"no-confusing-arrow": 0,
+ "camelcase": 0,
},
"settings": {
"import/resolver": {
diff --git a/dist/iris.js b/dist/iris.js
index 57cceb1c..3b4c3daa 100644
--- a/dist/iris.js
+++ b/dist/iris.js
@@ -3717,15 +3717,15 @@ var GenericExtractor = {
// This extractor is the default for all domains
domain: '*',
title: GenericTitleExtractor.extract,
- datePublished: GenericDatePublishedExtractor.extract,
+ date_published: GenericDatePublishedExtractor.extract,
author: GenericAuthorExtractor.extract,
content: GenericContentExtractor.extract.bind(GenericContentExtractor),
- leadImageUrl: GenericLeadImageUrlExtractor.extract,
+ lead_image_url: GenericLeadImageUrlExtractor.extract,
dek: GenericDekExtractor.extract,
- nextPageUrl: GenericNextPageUrlExtractor.extract,
- urlAndDomain: GenericUrlExtractor.extract,
+ next_page_url: GenericNextPageUrlExtractor.extract,
+ url_and_domain: GenericUrlExtractor.extract,
excerpt: GenericExcerptExtractor.extract,
- wordCount: GenericWordCountExtractor.extract,
+ word_count: GenericWordCountExtractor.extract,
extract: function extract(options) {
var html = options.html;
@@ -3737,33 +3737,33 @@ var GenericExtractor = {
}
var title = this.title(options);
- var datePublished = this.datePublished(options);
+ var date_published = this.date_published(options);
var author = this.author(options);
var content = this.content(_extends({}, options, { title: title }));
- var leadImageUrl = this.leadImageUrl(_extends({}, options, { content: content }));
+ var lead_image_url = this.lead_image_url(_extends({}, options, { content: content }));
var dek = this.dek(_extends({}, options, { content: content }));
- var nextPageUrl = this.nextPageUrl(options);
+ var next_page_url = this.next_page_url(options);
var excerpt = this.excerpt(_extends({}, options, { content: content }));
- var wordCount = this.excerpt(_extends({}, options, { content: content }));
+ var word_count = this.word_count(_extends({}, options, { content: content }));
- var _urlAndDomain = this.urlAndDomain(options);
+ var _url_and_domain = this.url_and_domain(options);
- var url = _urlAndDomain.url;
- var domain = _urlAndDomain.domain;
+ var url = _url_and_domain.url;
+ var domain = _url_and_domain.domain;
return {
title: title,
author: author,
- datePublished: datePublished || null,
+ date_published: date_published || null,
dek: dek,
- leadImageUrl: leadImageUrl,
+ lead_image_url: lead_image_url,
content: content,
- nextPageUrl: nextPageUrl,
+ next_page_url: next_page_url,
url: url,
domain: domain,
excerpt: excerpt,
- wordCount: wordCount
+ word_count: word_count
};
}
};
@@ -3910,17 +3910,17 @@ var RootExtractor = {
};
}
var title = extractResult(_extends({}, opts, { type: 'title' }));
- var datePublished = extractResult(_extends({}, opts, { type: 'datePublished' }));
+ var date_published = extractResult(_extends({}, opts, { type: 'date_published' }));
var author = extractResult(_extends({}, opts, { type: 'author' }));
- var nextPageUrl = extractResult(_extends({}, opts, { type: 'nextPageUrl' }));
+ var next_page_url = extractResult(_extends({}, opts, { type: 'next_page_url' }));
var content = extractResult(_extends({}, opts, { type: 'content', extractHtml: true, title: title
}));
- var leadImageUrl = extractResult(_extends({}, opts, { type: 'leadImageUrl', content: content }));
+ var lead_image_url = extractResult(_extends({}, opts, { type: 'lead_image_url', content: content }));
var dek = extractResult(_extends({}, opts, { type: 'dek', content: content }));
var excerpt = extractResult(_extends({}, opts, { type: 'excerpt', content: content }));
- var wordCount = extractResult(_extends({}, opts, { type: 'wordCount', content: content }));
+ var word_count = extractResult(_extends({}, opts, { type: 'word_count', content: content }));
- var _extractResult = extractResult(_extends({}, opts, { type: 'urlAndDomain' }));
+ var _extractResult = extractResult(_extends({}, opts, { type: 'url_and_domain' }));
var url = _extractResult.url;
var domain = _extractResult.domain;
@@ -3930,21 +3930,21 @@ var RootExtractor = {
title: title,
content: content,
author: author,
- datePublished: datePublished,
- leadImageUrl: leadImageUrl,
+ date_published: date_published,
+ lead_image_url: lead_image_url,
dek: dek,
- nextPageUrl: nextPageUrl,
+ next_page_url: next_page_url,
url: url,
domain: domain,
excerpt: excerpt,
- wordCount: wordCount
+ word_count: word_count
};
}
};
var collectAllPages = (function () {
var _ref = asyncToGenerator(regeneratorRuntime.mark(function _callee(_ref2) {
- var nextPageUrl = _ref2.nextPageUrl;
+ var next_page_url = _ref2.next_page_url;
var html = _ref2.html;
var $ = _ref2.$;
var metaCache = _ref2.metaCache;
@@ -3965,14 +3965,14 @@ var collectAllPages = (function () {
// likely gone wrong.
case 2:
- if (!(nextPageUrl && pages < 26)) {
+ if (!(next_page_url && pages < 26)) {
_context.next = 15;
break;
}
pages += 1;
_context.next = 6;
- return Resource.create(nextPageUrl);
+ return Resource.create(next_page_url);
case 6:
$ = _context.sent;
@@ -3980,7 +3980,7 @@ var collectAllPages = (function () {
html = $.html();
extractorOpts = {
- url: nextPageUrl,
+ url: next_page_url,
html: html,
$: $,
metaCache: metaCache,
@@ -3991,19 +3991,19 @@ var collectAllPages = (function () {
nextPageResult = RootExtractor.extract(Extractor, extractorOpts);
- previousUrls.push(nextPageUrl);
+ previousUrls.push(next_page_url);
result = _extends({}, result, {
content: '\n ' + result.content + '\n
\n Page ' + pages + '
\n ' + nextPageResult.content + '\n '
});
- nextPageUrl = nextPageResult.nextPageUrl;
+ next_page_url = nextPageResult.next_page_url;
_context.next = 2;
break;
case 15:
return _context.abrupt('return', _extends({}, result, {
- totalPages: pages,
- pagesRendered: pages
+ total_pages: pages,
+ pages_rendered: pages
}));
case 16:
@@ -4027,7 +4027,7 @@ var Iris = {
var opts = arguments.length <= 2 || arguments[2] === undefined ? {} : arguments[2];
return asyncToGenerator(regeneratorRuntime.mark(function _callee() {
- var _ref, _ref$fetchAllPages, fetchAllPages, Extractor, $, metaCache, result, _result, title, nextPageUrl;
+ var _ref, _ref$fetchAllPages, fetchAllPages, Extractor, $, metaCache, result, _result, title, next_page_url;
return regeneratorRuntime.wrap(function _callee$(_context) {
while (1) {
@@ -4055,11 +4055,11 @@ var Iris = {
result = RootExtractor.extract(Extractor, { url: url, html: html, $: $, metaCache: metaCache });
_result = result;
title = _result.title;
- nextPageUrl = _result.nextPageUrl;
+ next_page_url = _result.next_page_url;
- // Fetch more pages if nextPageUrl found
+ // Fetch more pages if next_page_url found
- if (!(fetchAllPages && nextPageUrl)) {
+ if (!(fetchAllPages && next_page_url)) {
_context.next = 19;
break;
}
@@ -4067,7 +4067,7 @@ var Iris = {
_context.next = 16;
return collectAllPages({
Extractor: Extractor,
- nextPageUrl: nextPageUrl,
+ next_page_url: next_page_url,
html: html,
$: $,
metaCache: metaCache,
@@ -4083,8 +4083,8 @@ var Iris = {
case 19:
result = _extends({}, result, {
- totalPages: 1,
- renderedPages: 1
+ total_pages: 1,
+ rendered_pages: 1
});
case 20:
diff --git a/dist/iris.js.map b/dist/iris.js.map
index 26a20e41..e1f6882b 100644
--- a/dist/iris.js.map
+++ b/dist/iris.js.map
@@ -1 +1 @@
-{"version":3,"file":null,"sources":["../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/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/utils/range.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/constants.js","../src/extractors/root-extractor.js","../src/extractors/collect-all-pages.js","../src/iris.js"],"sourcesContent":["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';\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 ${MAX_CONTENT_LENGTH}.`\n );\n }\n\n return true;\n}\n\n// Grabs the last two pieces of the URL and joins them back together\n// This is to get the 'livejournal.com' from 'erotictrains.livejournal.com'\nexport function baseDomain({ host }) {\n return host.split('.').slice(-2).join('.');\n}\n\n// Set our response attribute to the result of fetching our URL.\n// TODO: This should gracefully handle timeouts and raise the\n// proper exceptions on the many failure cases of HTTP.\n// TODO: Ensure we are not fetching something enormous. Always return\n// unicode content for HTML, with charset conversion.\n\nexport default async function fetchResource(url) {\n const parsedUrl = URL.parse(url);\n\n const options = {\n url: parsedUrl,\n headers: { ...REQUEST_HEADERS },\n timeout: FETCH_TIMEOUT,\n // Don't set encoding; this fixes issues\n // w/gzipped responses\n encoding: null,\n // Accept cookies\n jar: true,\n };\n\n const { response, body } = await get(options);\n\n try {\n validateResponse(response);\n return { body, response };\n } catch (e) {\n return e;\n }\n}\n","function convertMetaProp($, 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) {\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);\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// def __init__(self, url, parse_non_2xx=False, response=None):\n// \"\"\" Create a Resource.\n//\n// :param url: The URL for the document we should retrieve.\n// :param parse_non_2xx: If True, attempt to parse non-200 level\n// resources. If False, raise a RetrievalFailed\n// based exception. Default is False.\n// :param response: If not None, use as the response rather than\n// attempting to fetch it ourselves. Expects a\n// requests.models.Response object.\n// \"\"\"\n// self.url = url\n// self.parse_non_2xx = parse_non_2xx\n//\n// if response:\n// self.response = response\n// else:\n// self.response = self._fetch_resource()\n\n// Iris: Human-friendly content extraction.\n\n// import logging\n// import lxml\n// import re\n// import requests\n// import socket\n//\n// from django.conf import settings\n// from lxml.etree import XPathEvaluator\n// from lxml.html.clean import Cleaner\n// from urlparse import urlparse\n//\n// from utils.dom import extract_by_selector as ebs, convert_lazy_loaded_images\n// from utils.dom.attribmap import AttribMap\n// from utils.statsd import stats\n// from utils.text import is_text\n// from utils.html import get_charset_from_html, strip_content_encodings\n//\n// from . import exceptions\n//\n// logger = logging.getLogger(__name__)\n//\n// # Hosts that are allowed to use embeds and iframes. We should be very\n// # restrictive with this and only include top-tier video sites.\n// host_whitelist = ['www.youtube.com', 'www.vimeo.com']\n//\n// # The number of seconds to attempt to fetch a resource before timing out.\n// FETCH_TIMEOUT = 10\n//\n// cleaner = Cleaner(\n// style=True,\n// page_structure=False,\n// meta=False,\n// add_nofollow=False, # done by hand\n// remove_unknown_tags=False,\n// links=False,\n// host_whitelist=host_whitelist)\n//\n//\n//\n// class Resource(object):\n// \"\"\" A Resource is a wrapper class for an HTTP resource. Provides\n// functionality to fetch a resource as well as a handful of shortcut\n// methods to run xpath efficiently on HTML, etc.\n//\n// Uses requests and lxml internally for fetching and querying.\n// \"\"\"\n//\n//\n// def __init__(self, url, parse_non_2xx=False, response=None):\n// \"\"\" Create a Resource.\n//\n// :param url: The URL for the document we should retrieve.\n// :param parse_non_2xx: If True, attempt to parse non-200 level\n// resources. If False, raise a RetrievalFailed\n// based exception. Default is False.\n// :param response: If not None, use as the response rather than\n// attempting to fetch it ourselves. Expects a\n// requests.models.Response object.\n// \"\"\"\n// self.url = url\n// self.parse_non_2xx = parse_non_2xx\n//\n// if response:\n// self.response = response\n// else:\n// self.response = self._fetch_resource()\n//\n// def __unicode__(self):\n// return u''.format(self.url)\n//\n// def __repr__(self):\n// return \"\".format(self.url)\n//\n// @classmethod\n// def fabricate(kls, url, content, headers=None):\n// \"\"\" Given a URL and some content, create a fake Resource that looks\n// as though it has already fetched the content. Useful for using\n// Resource objects without having to do a GET.\n// \"\"\"\n//\n// if type(content) != unicode:\n// raise TypeError(\"Provided content must be unicode.\")\n//\n// if headers is None:\n// headers = {}\n//\n// try:\n// utf8_content = content.encode('utf-8', 'strict')\n// except UnicodeDecodeError:\n// logger.warning(\"Unable to encode content for url %s. Content \"\n// \"should be unicode and encodeable at this point.\")\n// utf8_content = content.encode('utf-8', 'replace')\n//\n// mocked_response_dict = {\n// \"cookies\": {},\n// \"_content\": utf8_content,\n// \"headers\": dict({\n// \"content-length\": len(content),\n// \"accept-ranges\": \"bytes\",\n// \"vary\": \"Accept-Encoding,Cookie\",\n// \"server\": \"Apache/2.2.21\",\n// \"content-type\": \"text/html; charset=UTF-8\"\n// }, **headers),\n// \"url\": url,\n// \"status_code\": 200,\n// \"_content_consumed\": False,\n// \"request\": None,\n// \"raw\": None,\n// \"error\": None,\n// \"config\": {\n// \"decode_unicode\": True,\n// \"pool_connections\": 10,\n// \"verbose\": None,\n// \"keep_alive\": True,\n// \"max_retries\": 0,\n// \"base_headers\": {\n// \"Accept-Encoding\": \"identity, deflate, compress, gzip\",\n// \"Accept\": \"|)}>#*\",\n// \"User-Agent\": \"python-requests/0.8.1\"\n// },\n// \"pool_maxsize\": 10,\n// \"safe_mode\": False,\n// \"max_redirects\": 30\n// },\n// \"history\": []\n// }\n// mocked_response = requests.Response()\n// for k, v in mocked_response_dict.items():\n// setattr(mocked_response, k, v)\n//\n// return Resource(\n// url = url,\n// response = mocked_response\n// )\n//\n//\n// @property\n// def url(self):\n// return self._url\n//\n//\n// @url.setter\n// def url(self, value):\n// parsed_url = urlparse(value)\n// if parsed_url.scheme not in ('http', 'https'):\n// raise ValueError(\"Resource only allows HTTP and HTTPS urls.\")\n//\n// if not parsed_url.netloc:\n// raise ValueError(\"Relative URLs are not allowed.\")\n//\n// self._url = value\n//\n// _parsed_url = None\n// @property\n// def parsed_url(self):\n// if self._parsed_url is None:\n// self._parsed_url = urlparse(self.url)\n// return self._parsed_url\n//\n// @property\n// def status_code(self):\n// return self.response.status_code\n//\n//\n// _content = None\n// @property\n// def content(self):\n// \"\"\"Return the content for a resource. Always returns unicode.\n//\n// \"\"\"\n// if self._content is None:\n// # Requests that come in without content-type encoding headers will\n// # default to iso-8859-1, which could be wrong\n// if (self.response.encoding and\n// self.response.encoding.lower() == 'iso-8859-1'):\n// # Dont send unicode, because it could have been decoded wrong\n// # by an incorrect content-type guess.\n// encoding = get_charset_from_html(self.response.content) or 'iso-8859-1'\n//\n// if encoding != self.response.encoding:\n// # First, try to use the encoding we found in the markup\n// try:\n// self._content = self.response.content.decode(encoding)\n// except (LookupError, UnicodeDecodeError):\n// stats.increment(\n// 'iris.resource.encoding.encoding_mismatch')\n// # That encoding might be wrong though, so if it is, use\n// # the one it reported since they could have the wrong\n// # one set in the markup. eg. sending the content over\n// # as iso but declaring it to be utf-8 like gq.com does.\n// # We may also end up with an invalid encoding type, at\n// # which point we should also just use the request\n// # encoding and replace silently.\n// self._content = self.response.content.decode(\n// self.response.encoding, 'replace')\n// else:\n// # If the encoding guess was right, just use the unicode\n// self._content = self.response.text\n//\n// else:\n// # Otherwise we trust the encoding\n// self._content = self.response.text\n//\n// return self._content\n//\n//\n// @property\n// def content_type(self):\n// return self.response.headers.get('content-type', '')\n//\n//\n// @property\n// def is_html(self):\n// if 'html' in self.content_type:\n// return True\n//\n// # Otherwise, just try parsing it and see if it succeeds\n// try:\n// return (self.doc is not None)\n// except:\n// return False\n//\n// @property\n// def is_plaintext(self):\n// if 'text/plain' in self.content_type:\n// return True\n//\n// return False\n//\n// @property\n// def is_image(self):\n// if 'image' in self.content_type:\n// return True\n//\n// return False\n//\n// @property\n// def is_pdf(self):\n// if 'pdf' in self.content_type:\n// return True\n//\n// return False\n//\n// _lxml_doc = None\n// @property\n// def doc(self):\n// if self._lxml_doc is None:\n// self._generate_lxml_doc()\n//\n// return self._lxml_doc\n//\n// _docxp = None\n// @property\n// def docxp(self):\n// \"\"\" Generate an XPath Evaluator for this doc. \"\"\"\n// if self._docxp is None:\n// self._docxp = XPathEvaluator(self.doc)\n//\n// return self._docxp\n//\n// _redocxp = None\n// @property\n// def redocxp(self):\n// \"\"\" Generate an XPath Evaluator for this doc, that includes the RE\n// namespace for regular expression matching.\n//\n// \"\"\"\n// if self._redocxp is None:\n// _rens = {'re':'http://exslt.org/regular-expressions'}\n// self._redocxp = XPathEvaluator(self.doc, namespaces=_rens)\n//\n// return self._redocxp\n//\n// def _generate_lxml_doc(self):\n// # First check if we have a text based resource\n// if (not 'html' in self.content_type and\n// not 'text' in self.content_type and\n// not is_text(self.content[:512])):\n// raise ValueError(\"Content does not appear to be text.\")\n//\n//\n// # Remove useless carriage returns which get parsed as
otherwise\n// content = re.sub(r'(\\n\\r|\\r\\n)', '\\n', self.content)\n//\n// # Dont pass any content encodings into lxml, it is dumb about them\n// content = strip_content_encodings(content)\n//\n// self._lxml_doc = lxml.html.fromstring(content)\n//\n//\n//\n//\n// if len(self._lxml_doc.getchildren()) == 0:\n// stats.increment('iris.resource.encoding.no_children')\n// raise ValueError(\"No children, likely a bad parse.\")\n//\n//\n// # Sometimes, lxml (or BeautifulSoup) will wrap the whole document\n// # in an extra html tag. This screws up a whole bunch of things in\n// # the parsing process. If this is the case, reset the doc to the\n// # ACTUAL root of the doc.\n// # Sample cases:\n// # * Strange Doctype causing issues: http://bit.ly/IATz0B\n// # * Messy markup causing double HTML tags: http://bit.ly/IGOq4o\n// # Also check for a body inside of our internal HTML tag, to determine\n// # that it's not just a junk HTML tag sibling at the bottom of the\n// # doc or something.\n// internal_html_tag = self._lxml_doc.find('html')\n// if (internal_html_tag is not None and\n// len(internal_html_tag.xpath('.//body')) > 0):\n// self._lxml_doc = internal_html_tag\n//\n// self._normalize_meta_tags()\n//\n// self._lxml_doc.make_links_absolute(self.url)\n//\n// # Convert any lazy loaded images into normal images before clean_html\n// # which will strip all other attributes\n// self._lxml_doc = convert_lazy_loaded_images(self._lxml_doc)\n//\n// # Clean the doc of anything malicious.\n// self._lxml_doc = cleaner.clean_html(self._lxml_doc)\n//\n// # Manually nofollow links so that we don't clobber rel author\n// # Workaround for https://bugs.launchpad.net/lxml/+bug/971754\n// for a in self.docxp('//a'):\n// if a.attrib.get('rel', None):\n// rel_attribs = set(a.attrib['rel'].split())\n// rel_attribs.add('nofollow')\n// a.attrib['rel'] = ' '.join(rel_attribs)\n// else:\n// a.attrib['rel'] = 'nofollow'\n//\n// # Re-relativize anchor links\n// anchor_link_xpath = (\"//a[starts-with(@href, '%s#')]\" %\n// self.url.replace(\"'\", \"%27\"))\n// for link in self.docxp(anchor_link_xpath):\n// link.attrib['href'] = link.attrib['href'].replace(self.url, '')\n//\n//\n// _attrib_map = None\n// @property\n// def attrib_map(self):\n// \"\"\" Create an AttribMap object for fast checking of class/id existence\n// in the document. Used in association with extract_by_selector.\n//\n// \"\"\"\n// if self._attrib_map is None:\n// self._attrib_map = AttribMap(self.doc)\n//\n// return self._attrib_map\n//\n//\n// def extract_by_selector(self, selector):\n// \" Shortcut to run extract_by_selector on our doc with our AttribMap. \"\n// return ebs(self.doc, selector, self.attrib_map, self.docxp)\n//\n//\n","const NYMagExtractor = {\n domain: 'nymag.com',\n content: {\n // Order by most likely. Extractor will stop on first occurence\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 && $children.get(0).tagName === 'img') {\n return 'figure';\n }\n\n return null;\n },\n },\n },\n\n title: {\n selectors: [\n 'h1.lede-feature-title',\n 'h1.headline-primary',\n 'h1',\n ],\n },\n\n author: {\n selectors: [\n '.by-authors',\n '.lede-feature-author',\n ],\n },\n\n dek: {\n selectors: [\n '.lede-feature-teaser',\n ],\n },\n\n datePublished: {\n selectors: [\n 'time.article-timestamp[datetime]',\n 'time.article-timestamp',\n ],\n },\n};\n\nexport default NYMagExtractor;\n","const BloggerExtractor = {\n domain: 'blogspot.com',\n content: {\n // Blogger is insane and does not load its content\n // initially in the page, but it's all there\n // in noscript\n selectors: [\n '.post-content noscript',\n ],\n\n // Selectors to remove from the extracted content\n clean: [\n ],\n\n // Convert the noscript tag to a div\n transforms: {\n noscript: 'div',\n },\n },\n\n author: {\n selectors: [\n '.post-author-name',\n ],\n },\n\n title: {\n selectors: [\n 'h2.title',\n ],\n },\n\n datePublished: {\n selectors: [\n 'span.publishdate',\n ],\n },\n};\n\nexport default BloggerExtractor;\n","const WikipediaExtractor = {\n domain: 'wikipedia.org',\n content: {\n selectors: [\n '#mw-content-text',\n ],\n\n // transform top infobox to an image with caption\n transforms: {\n '.infobox img': ($node) => {\n $node.parents('.infobox').prepend($node);\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 ],\n\n },\n\n author: 'Wikipedia Contributors',\n\n title: {\n selectors: [\n 'h2.title',\n ],\n },\n\n datePublished: {\n selectors: [\n '#footer-info-lastmod',\n ],\n },\n\n};\n\nexport default WikipediaExtractor;\n","import NYMagExtractor from './custom/nymag.com';\nimport BloggerExtractor from './custom/blogspot.com';\nimport WikipediaExtractor from './custom/wikipedia.org';\n\nconst Extractors = {\n 'nymag.com': NYMagExtractor,\n 'blogspot.com': BloggerExtractor,\n 'wikipedia.org': WikipediaExtractor,\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 => `[${selector}]`);\nexport const REMOVE_ATTR_LIST = REMOVE_ATTRS.join(',');\nexport const WHITELIST_ATTRS = ['src', 'href', 'class', 'id'];\nexport const WHITELIST_ATTRS_RE = new RegExp(`^(${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'].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',\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// 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\n// ## NOTES:\n// This is a working first pass, but if/when we start optimizing\n// this is a good candidate. - AP\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 = $div.children(DIV_TO_P_BLOCK_TAGS).length === 0;\n // .not(DIV_TO_P_BLOCK_TAGS).length === 0;\n if (convertable) {\n convertNodeTo($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 { 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, $) {\n $(STRIP_OUTPUT_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 = $.find('h1')\n const $hOnes = $('h1', article);\n if ($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 * 0.2);\n const wrappingDiv = $('');\n\n $candidate.parent().children().each((index, child) => {\n const $child = $(child);\n // Ignore tags like BR, HR, etc\n if (NON_TOP_CANDIDATE_TAGS_RE.test(child.tagName)) {\n return null;\n }\n\n const childScore = getScore($child);\n if (childScore) {\n if ($child === $candidate) {\n wrappingDiv.append($child);\n } else {\n let contentBonus = 0;\n // extract to scoreLinkDensity() TODO\n const density = linkDensity($child);\n\n // If sibling has a very low link density,\n // give it a small bonus\n if (density < 0.05) {\n contentBonus += 20;\n }\n\n // If sibling has a high link density,\n // give it a penalty\n if (density >= 0.5) {\n contentBonus -= 20;\n }\n\n // If sibling node has the same class as\n // candidate, give it a bonus\n if ($child.attr('class') === $candidate.attr('class')) {\n contentBonus += topScore * 0.2;\n }\n\n const newScore = getScore($child) + contentBonus;\n\n if (newScore >= siblingScoreThreshold) {\n return wrappingDiv.append($child);\n } else if (child.tagName === 'p') {\n const childContent = $child.text();\n const childContentLength = textLength(childContent);\n\n if (childContentLength > 80 && density < 0.25) {\n return wrappingDiv.append($child);\n } else if (childContentLength <= 80 && density === 0 &&\n hasSentenceEnd(childContent)) {\n return wrappingDiv.append($child);\n }\n }\n }\n }\n\n return null;\n });\n\n return wrappingDiv;\n}\n","import { NON_TOP_CANDIDATE_TAGS_RE } from './constants';\nimport { getScore } from './index';\nimport mergeSiblings from './merge-siblings';\n\n// After we've calculated scores, loop through all of the possible\n// candidate nodes we found and find the one with the highest score.\nexport default function findTopCandidate($) {\n let $candidate;\n let topScore = 0;\n\n $('[score]').each((index, node) => {\n const $node = $(node);\n const score = getScore($node);\n // Ignore tags like BR, HR, etc\n if (NON_TOP_CANDIDATE_TAGS_RE.test(node.tagName)) {\n return;\n }\n\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, $content));\n\n // console.log($content.html())\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 $(IS_WP_SELECTOR).length > 0;\n}\n","// CLEAN AUTHOR CONSTANTS\nexport const CLEAN_AUTHOR_RE = /^\\s*(posted |written )?by\\s*:?\\s*(.*)/i;\n // author = re.sub(r'^\\s*(posted |written )?by\\s*:?\\s*(.*)(?i)',\n\n// CLEAN DEK CONSTANTS\nexport const TEXT_LINK_RE = new RegExp('http(s)?://', 'i');\n// An ordered list of meta tag names that denote likely article deks.\n// From most distinct to least distinct.\n//\n// NOTE: There are currently no meta tags that seem to provide the right\n// content consistenty enough. Two options were:\n// - og:description\n// - dc.description\n// However, these tags often have SEO-specific junk in them that's not\n// header-worthy like a dek is. Excerpt material at best.\nexport const DEK_META_TAGS = [\n];\n\n// An ordered list of Selectors to find likely article deks. From\n// most explicit to least explicit.\n//\n// Should be more restrictive than not, as a failed dek can be pretty\n// detrimental to the aesthetics of an article.\nexport const DEK_SELECTORS = [\n '.entry-summary',\n];\n\n// CLEAN DATE PUBLISHED CONSTANTS\nexport const CLEAN_DATE_STRING_RE = /^\\s*published\\s*:?\\s*(.*)/i;\nexport const TIME_MERIDIAN_SPACE_RE = /(.*\\d)(am|pm)(.*)/i;\nexport const TIME_MERIDIAN_DOTS_RE = /\\.m\\./i;\nconst months = [\n 'jan',\n 'feb',\n 'mar',\n 'apr',\n 'may',\n 'jun',\n 'jul',\n 'aug',\n 'sep',\n 'oct',\n 'nov',\n 'dec',\n];\nconst allMonths = months.join('|');\nconst timestamp1 = '[0-9]{1,2}:[0-9]{2,2}( ?[ap].?m.?)?';\nconst timestamp2 = '[0-9]{1,2}[/-][0-9]{1,2}[/-][0-9]{2,4}';\nexport const SPLIT_DATE_STRING =\n new RegExp(`(${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 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 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 }\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 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 style or align 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 cleanTags(article, $, cleanConditionally);\n\n // Remove empty paragraph nodes\n removeEmpty(article, $);\n\n return article;\n}\n // headers = doc.xpath('.//h2 | .//h3 | .//h4 | .//h5 | .//h6')\n // for header in headers:\n // drop_header = False\n //\n // # Remove any headers that are before any 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 // no_previous_ps = int(header.xpath(\"count(preceding::p[1])\")) == 0\n // if no_previous_ps:\n // similar_header_count = int(doc.xpath('count(.//%s)' % header.tag))\n // if similar_header_count < 3:\n // drop_header = True\n //\n // # Remove any headers that match the title exactly.\n // if inner_text(header) == self.title:\n // drop_header = True\n //\n // # If this header has a negative weight, it's probably junk.\n // # Get rid of it.\n // if self.get_weight(header) < 0:\n // drop_header = True\n //\n // if drop_header:\n // try:\n // header.drop_tree()\n // except AssertionError:\n // # No parent exists for this node, so just blank it out.\n // header.text = ''\n //\n // if clean_conditionally:\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 // self._clean_conditionally(doc, ['ul', 'ol', 'table', 'div'])\n //\n // return doc\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 leadImageUrl: cleanImage,\n dek: cleanDek,\n datePublished: 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|>([^|]|$)|»([^|]|$))', 'i');\n\n// Match any link text/classname/id that looks like it is an end link: things\n// like \"first\", \"last\", \"end\", etc.\nexport const CAP_LINK_TEXT_RE = new RegExp('(first|last|end)', 'i');\n\n// Match any link text/classname/id that looks like it means the previous\n// page.\nexport const PREV_LINK_TEXT_RE = new RegExp('(prev|earl|old|new|<|«)', 'i');\n\n// Match any phrase that looks like it could be page, or paging, or pagination\nexport const PAGE_RE = new RegExp('pag(e|ing|inat)', 'i');\n\n","import { EXTRANEOUS_LINK_HINTS_RE } from '../constants';\n\nexport default function scoreExtraneousLinks(href) {\n // If the URL itself contains extraneous values, give a penalty.\n if (EXTRANEOUS_LINK_HINTS_RE.test(href)) {\n return -25;\n }\n\n return 0;\n}\n","export default function* range(start = 1, end = 1) {\n while (start <= end) {\n yield start += 1;\n }\n}\n","import { range } from 'utils';\nimport {\n NEGATIVE_SCORE_RE,\n POSITIVE_SCORE_RE,\n PAGE_RE,\n} from 'utils/dom/constants';\nimport { EXTRANEOUS_LINK_HINTS_RE } from '../constants';\n\nfunction makeSig($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';\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 datePublished: GenericDatePublishedExtractor.extract,\n author: GenericAuthorExtractor.extract,\n content: GenericContentExtractor.extract.bind(GenericContentExtractor),\n leadImageUrl: GenericLeadImageUrlExtractor.extract,\n dek: GenericDekExtractor.extract,\n nextPageUrl: GenericNextPageUrlExtractor.extract,\n urlAndDomain: GenericUrlExtractor.extract,\n excerpt: GenericExcerptExtractor.extract,\n wordCount: GenericWordCountExtractor.extract,\n\n extract(options) {\n const { html } = options;\n\n if (html) {\n const $ = cheerio.load(html);\n options.$ = $;\n }\n\n const title = this.title(options);\n const datePublished = this.datePublished(options);\n const author = this.author(options);\n const content = this.content({ ...options, title });\n const leadImageUrl = this.leadImageUrl({ ...options, content });\n const dek = this.dek({ ...options, content });\n const nextPageUrl = this.nextPageUrl(options);\n const excerpt = this.excerpt({ ...options, content });\n const wordCount = this.excerpt({ ...options, content });\n const { url, domain } = this.urlAndDomain(options);\n\n return {\n title,\n author,\n datePublished: datePublished || null,\n dek,\n leadImageUrl,\n content,\n nextPageUrl,\n url,\n domain,\n excerpt,\n wordCount,\n };\n },\n};\n\nexport default GenericExtractor;\n","import URL from 'url';\n\nimport Extractors from './all';\nimport GenericExtractor from './generic';\n\nexport default function getExtractor(url) {\n const parsedUrl = URL.parse(url);\n const { hostname } = parsedUrl;\n const baseDomain = hostname.split('.').slice(-2).join('.');\n\n return Extractors[hostname] || Extractors[baseDomain] || GenericExtractor;\n}\n","export const ATTR_RE = /\\[([\\w-]+)\\]/;\n","import 'babel-polyfill';\n\nimport Cleaners from 'cleaners';\nimport { convertNodeTo } from 'utils/dom';\nimport GenericExtractor from './generic';\nimport { ATTR_RE } from './constants';\n\n// Remove elements by an array of selectors\nexport function cleanBySelectors($content, $, { clean }) {\n if (!clean) return null;\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 null;\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\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 } = extractionOpts;\n\n const matchingSelector = selectors.find(selector => $(selector).length === 1 && $(selector).text().trim() !== '');\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);\n\n return $.html($content);\n }\n // if selector includes an attr (e.g., img[src]),\n // extract the attr\n const attr = matchingSelector.match(ATTR_RE);\n let result;\n\n if (attr) {\n result = $(matchingSelector).attr(attr[1]);\n } else {\n // otherwise use the text of the node\n result = $(matchingSelector).text();\n }\n return Cleaners[type](result, opts);\n}\n\nfunction extractResult(opts) {\n const { type, extractor } = opts;\n\n // If nothing matches the selector,\n // run the Generic extraction\n return select({ ...opts, extractionOpts: extractor[type] }) ||\n GenericExtractor[type](opts);\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 datePublished = extractResult({ ...opts, type: 'datePublished' });\n const author = extractResult({ ...opts, type: 'author' });\n const nextPageUrl = extractResult({ ...opts, type: 'nextPageUrl' });\n const content = extractResult({\n ...opts, type: 'content', extractHtml: true, title,\n });\n const leadImageUrl = extractResult({ ...opts, type: 'leadImageUrl', content });\n const dek = extractResult({ ...opts, type: 'dek', content });\n const excerpt = extractResult({ ...opts, type: 'excerpt', content });\n const wordCount = extractResult({ ...opts, type: 'wordCount', content });\n const { url, domain } = extractResult({ ...opts, type: 'urlAndDomain' });\n\n return {\n title,\n content,\n author,\n datePublished,\n leadImageUrl,\n dek,\n nextPageUrl,\n url,\n domain,\n excerpt,\n wordCount,\n };\n },\n};\n\nexport default RootExtractor;\n","import 'babel-polyfill';\nimport { removeAnchor } from 'utils/text';\nimport RootExtractor from 'extractors/root-extractor';\nimport Resource from 'resource';\n\nexport default async function collectAllPages(\n {\n nextPageUrl,\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 (nextPageUrl && pages < 26) {\n pages += 1;\n $ = await Resource.create(nextPageUrl);\n html = $.html();\n\n const extractorOpts = {\n url: nextPageUrl,\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(nextPageUrl);\n result = {\n ...result,\n content: `\n ${result.content}\n <hr>\n <h4>Page ${pages}</h4>\n ${nextPageResult.content}\n `,\n };\n\n nextPageUrl = nextPageResult.nextPageUrl;\n }\n\n return {\n ...result,\n totalPages: pages,\n pagesRendered: pages,\n };\n}\n","import Resource from 'resource';\nimport getExtractor from 'extractors/get-extractor';\nimport RootExtractor from 'extractors/root-extractor';\nimport collectAllPages from 'extractors/collect-all-pages';\n\nconst Iris = {\n async parse(url, html, opts = {}) {\n const { fetchAllPages = true } = opts || true;\n const Extractor = getExtractor(url);\n // console.log(`Using extractor for ${Extractor.domain}`);\n\n const $ = await Resource.create(url, html);\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 });\n const { title, nextPageUrl } = result;\n\n // Fetch more pages if nextPageUrl found\n if (fetchAllPages && nextPageUrl) {\n result = await collectAllPages(\n {\n Extractor,\n nextPageUrl,\n html,\n $,\n metaCache,\n result,\n title,\n url,\n }\n );\n } else {\n result = {\n ...result,\n totalPages: 1,\n renderedPages: 1,\n };\n }\n\n return result;\n },\n\n};\n\nexport default Iris;\n"],"names":["REQUEST_HEADERS","FETCH_TIMEOUT","BAD_CONTENT_TYPES","BAD_CONTENT_TYPES_RE","RegExp","join","MAX_CONTENT_LENGTH","get","options","Promise","resolve","reject","err","response","body","validateResponse","parseNon2xx","statusMessage","statusCode","Error","error","headers","contentType","contentLength","test","url","URL","parse","parsedUrl","fetchResource","convertMetaProp","$","from","to","each","_","node","$node","value","attr","removeAttr","normalizeMetaTags","IS_LINK","IS_IMAGE","TAGS_TO_REMOVE","convertLazyLoadedImages","img","ownKeys","attribs","forEach","isComment","index","type","cleanComments","root","find","contents","filter","remove","clean","Resource","preparedResponse","validResponse","generateDoc","result","content","includes","cheerio","load","normalizeWhitespace","children","length","NYMagExtractor","$children","tagName","BloggerExtractor","WikipediaExtractor","parents","prepend","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","replaceWith","convertDivs","div","$div","convertable","convertSpans","span","$span","convertToParagraphs","convertNodeTo","attribString","Reflect","key","cleanForHeight","$img","height","parseInt","width","removeSpacers","cleanImages","$article","stripJunkTags","article","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","$parent","rawScore","scoreContent","parentSelector","childSelector","NORMALIZE_RE","normalizeSpaces","replace","extractFromUrl","regexList","matchRe","re","exec","PAGE_IN_HREF_RE","HAS_ALPHA_RE","IS_ALPHA_RE","IS_DIGIT_RE","pageNumFromUrl","matches","pageNum","removeAnchor","split","isGoodSegment","segment","firstSegmentHasLetters","goodSegment","toLowerCase","articleBaseUrl","parsed","protocol","host","path","cleanedSegments","reverse","rawSegment","possibleSegment","fileExt","push","SENTENCE_END_RE","hasSentenceEnd","mergeSiblings","$candidate","topScore","siblingScoreThreshold","wrappingDiv","child","$child","childScore","append","contentBonus","density","linkDensity","newScore","childContent","childContentLength","findTopCandidate","first","removeUnlessContent","weight","hasClass","pCount","inputCount","imgCount","nodeIsList","previousNode","prev","scriptCount","cleanTags","cleanHeaders","title","header","$header","prevAll","rewriteTopLevel","absolutize","rootUrl","$content","absoluteUrl","makeLinksAbsolute","totalTextLength","linkText","linkLength","extractFromMeta","metaNames","cachedNames","foundNames","indexOf","name","nodes","values","toArray","metaValue","stripTags","isGoodNode","maxChildren","withinComment","extractFromSelectors","selectors","textOnly","html","cleanText","commentParent","class","undefined","nodeIsSufficient","isWordpress","CLEAN_AUTHOR_RE","TEXT_LINK_RE","CLEAN_DATE_STRING_RE","TIME_MERIDIAN_SPACE_RE","TIME_MERIDIAN_DOTS_RE","months","allMonths","timestamp1","timestamp2","SPLIT_DATE_STRING","TITLE_SPLITTERS_RE","DOMAIN_ENDINGS_RE","cleanAuthor","author","leadImageUrl","validUrl","isWebUri","cleanDek","dek","dekText","cleanDateString","dateString","cleanDatePublished","date","moment","Date","isValid","toISOString","extractCleanNode","cleanConditionally","cleanTitle","resolveSplitTitle","h1","extractBreadcrumbTitle","splitTitle","termCounts","titleText","maxTerm","termCount","splitEnds","longestEnd","end","cleanDomainFromTitle","nakedDomain","startSlug","startSlugRatio","wuzzy","levenshtein","endSlug","endSlugRatio","newTitle","Cleaners","cleanImage","cleanContent","extractBestNode","opts","$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","$sibling","scoreByDimensions","src","area","round","scoreByPosition","$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","range","start","makeSig","$link","positiveMatch","negativeMatch","parentData","scorePrevLink","linkData","shouldScore","baseUrl","previousUrls","hostname","linkHost","fragment","scoreBaseUrl","baseRegex","scoreNextLinkText","scoreCapLinks","makeBaseRegex","scoreLinks","links","scoredPages","possiblePages","link","possiblePage","GenericNextPageUrlExtractor","scoredLinks","topPage","scoredLink","CANONICAL_META_SELECTORS","parseDomain","GenericUrlExtractor","$canonical","metaUrl","EXCERPT_META_SELECTORS","maxLength","ellipsize","ellipse","GenericExcerptExtractor","excerpt","shortContent","GenericWordCountExtractor","GenericExtractor","extract","bind","nextPageUrl","wordCount","urlAndDomain","domain","getExtractor","baseDomain","ATTR_RE","cleanBySelectors","transformElements","transforms","$matches","select","extractionOpts","extractHtml","matchingSelector","wrap","extractResult","extractor","RootExtractor","contentOnly","extractedTitle","Extractor","pages","create","extractorOpts","nextPageResult","collectAllPages","Iris","fetchAllPages"],"mappings":";;;;;;;;;;;;;;AAAO,IAAMA,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;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;ACnBA,SAASC,GAAT,CAAaC,OAAb,EAAsB;SACb,IAAIC,OAAJ,CAAY,UAACC,OAAD,EAAUC,MAAV,EAAqB;YAC9BH,OAAR,EAAiB,UAACI,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,yDAAP,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;;;;MAoB1DpB,qBAAqBqB,IAArB,CAA0BF,WAA1B,CAAJ,EAA4C;UACpC,IAAIH,KAAJ,yCACkCG,WADlC,0BAAN;;;;MAMEC,gBAAgBjB,kBAApB,EAAwC;UAChC,IAAIa,KAAJ,yEACkEb,kBADlE,OAAN;;;SAKK,IAAP;;;AAGF,AAMA;;;;;;AAMA;uDAAe,iBAA6BmB,GAA7B;;;;;;;qBAAA,GACKC,IAAIC,KAAJ,CAAUF,GAAV,CADL;mBAAA,GAGG;mBACTG,SADS;oCAEA5B,eAAd,CAFc;uBAGLC,aAHK;;;wBAMJ,IANI;;mBAQT;aAXM;;mBAcoBM,IAAIC,OAAJ,CAdpB;;;;oBAAA,SAcLK,QAdK;gBAAA,SAcKC,IAdL;;;6BAiBMD,QAAjB;6CACO,EAAEC,UAAF,EAAQD,kBAAR,EAlBI;;;;;;;;;;;;;GAAf;;WAA8BgB,aAA9B;;;;SAA8BA,aAA9B;;;AC7EA,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,IAAItC,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAhB;AACP,AAAO,IAAMuC,WAAW,IAAIvC,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB;;AAEP,AAAO,IAAMwC,iBAAiB,CAC5B,QAD4B,EAE5B,OAF4B,EAG5B,MAH4B,EAI5BvC,IAJ4B,CAIvB,GAJuB,CAAvB;;ACIP;;;;;AAKA,AAAe,SAASwC,uBAAT,CAAiCd,CAAjC,EAAoC;IAC/C,KAAF,EAASG,IAAT,CAAc,UAACC,CAAD,EAAIW,GAAJ,EAAY;YAChBC,OAAR,CAAgBD,IAAIE,OAApB,EAA6BC,OAA7B,CAAqC,UAACV,IAAD,EAAU;UACvCD,QAAQQ,IAAIE,OAAJ,CAAYT,IAAZ,CAAd;;UAEIA,SAAS,KAAT,IAAkBG,QAAQlB,IAAR,CAAac,KAAb,CAAlB,IACAK,SAASnB,IAAT,CAAcc,KAAd,CADJ,EAC0B;UACtBQ,GAAF,EAAOP,IAAP,CAAY,KAAZ,EAAmBD,KAAnB;;KALJ;GADF;;SAWOP,CAAP;;;ACtBF,SAASmB,SAAT,CAAmBC,KAAnB,EAA0Bf,IAA1B,EAAgC;SACvBA,KAAKgB,IAAL,KAAc,SAArB;;;AAGF,SAASC,aAAT,CAAuBtB,CAAvB,EAA0B;IACtBuB,IAAF,GAASC,IAAT,CAAc,GAAd,EACSC,QADT,GAESC,MAFT,CAEgBP,SAFhB,EAGSQ,MAHT;;SAKO3B,CAAP;;;AAGF,AAAe,SAAS4B,KAAT,CAAe5B,CAAf,EAAkB;IAC7Ba,cAAF,EAAkBc,MAAlB;;MAEIL,cAActB,CAAd,CAAJ;SACOA,CAAP;;;ACRF,IAAM6B,WAAW;;;;;;;;QAAA,kBAQFnC,GARE,EAQGoC,gBARH,EAQqB;;;;;;;;;oBAAA;;mBAG9BA,gBAH8B;;;;;2BAAA,GAIV;+BACL,IADK;4BAER,GAFQ;yBAGX;kCACS,WADT;oCAEW;;eATU;;;uBAavB,EAAE/C,MAAM+C,gBAAR,EAA0BhD,UAAUiD,aAApC,EAAT;;;;;;qBAEejC,cAAcJ,GAAd,CAfiB;;;oBAAA;;;+CAiB3B,MAAKsC,WAAL,CAAiBC,MAAjB,CAjB2B;;;;;;;;;GARrB;aAAA,6BA4B0B;QAArBC,OAAqB,QAA3BnD,IAA2B;QAAZD,QAAY,QAAZA,QAAY;QACfS,WADe,GACCT,SAASQ,OADV,CAC/B,cAD+B;;;;;QAKnC,CAACC,YAAY4C,QAAZ,CAAqB,MAArB,CAAD,IACA,CAAC5C,YAAY4C,QAAZ,CAAqB,MAArB,CADL,EACmC;YAC3B,IAAI/C,KAAJ,CAAU,qCAAV,CAAN;;;QAGEY,IAAIoC,QAAQC,IAAR,CAAaH,OAAb,EAAsB,EAAEI,qBAAqB,IAAvB,EAAtB,CAAR;;QAEItC,EAAEuB,IAAF,GAASgB,QAAT,GAAoBC,MAApB,KAA+B,CAAnC,EAAsC;YAC9B,IAAIpD,KAAJ,CAAU,kCAAV,CAAN;;;QAGEsB,kBAAkBV,CAAlB,CAAJ;QACIc,wBAAwBd,CAAxB,CAAJ;QACI4B,MAAM5B,CAAN,CAAJ;;WAEOA,CAAP;;CAhDJ;;AAoDA,AACA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AChEA,IAAMyC,iBAAiB;UACb,WADa;WAEZ;;eAEI,CACT,qBADS,EAET,cAFS,EAGT,iBAHS,CAFJ;;;WASA,CACL,KADK,EAEL,uBAFK,CATA;;;;;;;;gBAoBK;;UAEN,IAFM;;;gBAKA,kBAACnC,KAAD,EAAW;YACboC,YAAYpC,MAAMiC,QAAN,EAAlB;YACIG,UAAUF,MAAV,KAAqB,CAArB,IAA0BE,UAAUlE,GAAV,CAAc,CAAd,EAAiBmE,OAAjB,KAA6B,KAA3D,EAAkE;iBACzD,QAAP;;;eAGK,IAAP;;;GAjCe;;SAsCd;eACM,CACT,uBADS,EAET,qBAFS,EAGT,IAHS;GAvCQ;;UA8Cb;eACK,CACT,aADS,EAET,sBAFS;GA/CQ;;OAqDhB;eACQ,CACT,sBADS;GAtDQ;;iBA2DN;eACF,CACT,kCADS,EAET,wBAFS;;CA5Df,CAmEA;;ACnEA,IAAMC,mBAAmB;UACf,cADe;WAEd;;;;eAII,CACT,wBADS,CAJJ;;;WASA,EATA;;;gBAaK;gBACA;;GAhBS;;UAoBf;eACK,CACT,mBADS;GArBU;;SA0BhB;eACM,CACT,UADS;GA3BU;;iBAgCR;eACF,CACT,kBADS;;CAjCf,CAuCA;;ACvCA,IAAMC,qBAAqB;UACjB,eADiB;WAEhB;eACI,CACT,kBADS,CADJ;;;gBAMK;sBACM,oBAACvC,KAAD,EAAW;cACnBwC,OAAN,CAAc,UAAd,EAA0BC,OAA1B,CAAkCzC,KAAlC;OAFQ;0BAIU,YAJV;kBAKE;KAXP;;;WAeA,CACL,iBADK,EAEL,oCAFK,EAGL,MAHK;;GAjBgB;;UAyBjB,wBAzBiB;;SA2BlB;eACM,CACT,UADS;GA5BY;;iBAiCV;eACF,CACT,sBADS;;;CAlCf,CAyCA;;ACrCA,IAAM0C,aAAa;eACJP,cADI;kBAEDG,gBAFC;mBAGAC;CAHnB,CAMA;;ACVA;AACA,AAAO,IAAMI,YAAY,IAAI5E,MAAJ,CAAW,gCAAX,EAA6C,GAA7C,CAAlB;;;AAGP,AAAO,IAAM6E,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,aAAa7E,IAAb,CAAkB,GAAlB,CAAzB;AACP,AAAO,IAAMkF,kBAAkB,CAAC,KAAD,EAAQ,MAAR,EAAgB,OAAhB,EAAyB,IAAzB,CAAxB;AACP,AAAO,IAAMC,qBAAqB,IAAIpF,MAAJ,QAAgBmF,gBAAgBlF,IAAhB,CAAqB,GAArB,CAAhB,SAA+C,GAA/C,CAA3B;;;AAGP,AAAO,IAAMoF,oBAAoB,CAAC,GAAD,CAA1B;AACP,AAAO,IAAMC,yBAAyBD,kBAAkBL,GAAlB,CAAsB;SAAUO,GAAV;CAAtB,EAA6CtF,IAA7C,CAAkD,GAAlD,CAA/B;;;AAGP,AAAO,IAAMuF,2BAA2B,CAAC,IAAD,EAAO,IAAP,EAAa,OAAb,EAAsB,KAAtB,EAA6BvF,IAA7B,CAAkC,GAAlC,CAAjC;;;AAGP,IAAMwF,cAAc,CAAC,IAAD,EAAO,IAAP,EAAa,IAAb,EAAmB,IAAnB,EAAyB,IAAzB,CAApB;AACA,AAAO,IAAMC,kBAAkBD,YAAYxF,IAAZ,CAAiB,GAAjB,CAAxB;;;;;;;;AASP,AAAO,IAAM0F,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,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,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,EASjC5F,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM6F,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,IAAI/F,MAAJ,QAAgB8F,uBAAuB7F,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;AAGP,AAYA,AAAO,IAAM+F,cAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,iBAAiB,IAAIjG,MAAJ,CAAWgG,YAAY/F,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAMiG,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,IAAInG,MAAJ,CAAWkG,qBAAqBjG,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA;;;;AAIA,AAAO,IAAMmG,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,IAAIrG,MAAJ,CAAWoG,qBAAqBnG,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMqG,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,IAAIxG,MAAJ,CAAWuG,sBAAsBtG,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;AAGP,AAAO,IAAMwG,UAAU,IAAIzG,MAAJ,CAAW,iBAAX,EAA8B,GAA9B,CAAhB;;AAEP,AAMA,AAIA,AAIA,AAGA,AAGA;;AAEA,AAAO,IAAM0G,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,IAAI3G,MAAJ,QAAgB0G,iBAAiBzG,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAM2G,sBAAsBjB,8BAA8B1F,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM4G,uBAAuB,IAAI7G,MAAJ,CAAW4G,mBAAX,EAAgC,GAAhC,CAA7B;;AAEP,IAAME,sBAAsBlB,8BAA8B3F,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM8G,uBAAuB,IAAI/G,MAAJ,CAAW8G,mBAAX,EAAgC,GAAhC,CAA7B,CAEP,AAGA,AACA,AACA,AAEA;;ACzXA;;;;AAIA,AAAe,SAASE,uBAAT,CAAiCrF,CAAjC,EAAoC;;;;;;;;;;IAU/C,GAAF,EAAOsF,GAAP,CAAW,GAAX,EAAgBnF,IAAhB,CAAqB,UAACiB,KAAD,EAAQf,IAAR,EAAiB;QAC9BC,QAAQN,EAAEK,IAAF,CAAd;QACMkF,UAAUjF,MAAME,IAAN,CAAW,OAAX,CAAhB;QACMgF,KAAKlF,MAAME,IAAN,CAAW,IAAX,CAAX;QACI,CAACgF,EAAD,IAAO,CAACD,OAAZ,EAAqB;;QAEfE,cAAgBF,WAAW,EAA3B,WAAiCC,MAAM,EAAvC,CAAN;QACIJ,qBAAqB3F,IAArB,CAA0BgG,UAA1B,CAAJ,EAA2C;;KAA3C,MAEO,IAAIP,qBAAqBzF,IAArB,CAA0BgG,UAA1B,CAAJ,EAA2C;YAC1C9D,MAAN;;GAVJ;;SAcO3B,CAAP;;;AC/BF;;;;;;;;;;AAUA,AAAe,SAAS0F,OAAT,CAAiB1F,CAAjB,EAAoB;MAC7B2F,aAAa,KAAjB;IACE,IAAF,EAAQxF,IAAR,CAAa,UAACiB,KAAD,EAAQwE,OAAR,EAAoB;QACzBC,cAAc7F,EAAE4F,OAAF,EAAWE,IAAX,GAAkBtH,GAAlB,CAAsB,CAAtB,CAApB;;QAEIqH,eAAeA,YAAYlD,OAAZ,KAAwB,IAA3C,EAAiD;mBAClC,IAAb;QACEiD,OAAF,EAAWjE,MAAX;KAFF,MAGO,IAAIgE,UAAJ,EAAgB;mBACR,KAAb;;mBAEaC,OAAb,EAAsB5F,CAAtB,EAAyB,IAAzB;;GATJ;;SAaOA,CAAP;;;ACzBF;;;;;;;;;;;AAWA,AAAe,SAAS+F,YAAT,CAAsB1F,IAAtB,EAA4BL,CAA5B,EAA2C;MAAZgG,EAAY,yDAAP,KAAO;;MAClD1F,QAAQN,EAAEK,IAAF,CAAd;;MAEI2F,EAAJ,EAAQ;QACFC,UAAU5F,KAAK6F,WAAnB;QACMC,IAAInG,EAAE,SAAF,CAAV;;;;WAIOiG,WAAW,EAAEA,QAAQtD,OAAR,IAAmBqC,oBAAoBvF,IAApB,CAAyBwG,QAAQtD,OAAjC,CAArB,CAAlB,EAAmF;UAC3EuD,cAAcD,QAAQC,WAA5B;QACED,OAAF,EAAWG,QAAX,CAAoBD,CAApB;gBACUD,WAAV;;;UAGIG,WAAN,CAAkBF,CAAlB;UACMxE,MAAN;WACO3B,CAAP;;;SAGKA,CAAP;;;AC7BF,SAASsG,WAAT,CAAqBtG,CAArB,EAAwB;IACpB,KAAF,EAASG,IAAT,CAAc,UAACiB,KAAD,EAAQmF,GAAR,EAAgB;QACtBC,OAAOxG,EAAEuG,GAAF,CAAb;QACME,cAAcD,KAAKjE,QAAL,CAAc2B,mBAAd,EAAmC1B,MAAnC,KAA8C,CAAlE;;QAEIiE,WAAJ,EAAiB;oBACDD,IAAd,EAAoBxG,CAApB,EAAuB,GAAvB;;GALJ;;SASOA,CAAP;;;AAGF,SAAS0G,YAAT,CAAsB1G,CAAtB,EAAyB;IACrB,MAAF,EAAUG,IAAV,CAAe,UAACiB,KAAD,EAAQuF,IAAR,EAAiB;QACxBC,QAAQ5G,EAAE2G,IAAF,CAAd;QACMF,cAAcG,MAAM9D,OAAN,CAAc,QAAd,EAAwBN,MAAxB,KAAmC,CAAvD;QACIiE,WAAJ,EAAiB;oBACDG,KAAd,EAAqB5G,CAArB,EAAwB,GAAxB;;GAJJ;;SAQOA,CAAP;;;;;;;;;;;;;;;AAeF,AAAe,SAAS6G,mBAAT,CAA6B7G,CAA7B,EAAgC;MACzC0F,QAAQ1F,CAAR,CAAJ;MACIsG,YAAYtG,CAAZ,CAAJ;MACI0G,aAAa1G,CAAb,CAAJ;;SAEOA,CAAP;;;AC5Ca,SAAS8G,aAAT,CAAuBxG,KAAvB,EAA8BN,CAA9B,EAA4C;MAAX4D,GAAW,yDAAL,GAAK;;mBACrCtD,MAAM9B,GAAN,CAAU,CAAV,CADqC;;MACjDyC,OADiD,cACjDA,OADiD;;MAEnD8F,eAAeC,QAAQhG,OAAR,CAAgBC,OAAhB,EACQoC,GADR,CACY;WAAU4D,GAAV,SAAiBhG,QAAQgG,GAAR,CAAjB;GADZ,EAEQ3I,IAFR,CAEa,GAFb,CAArB;;QAIM+H,WAAN,OAAsBzC,GAAtB,SAA6BmD,YAA7B,SAA6CzG,MAAMmB,QAAN,EAA7C,UAAkEmC,GAAlE;SACO5D,CAAP;;;ACPF,SAASkH,cAAT,CAAwBC,IAAxB,EAA8BnH,CAA9B,EAAiC;MACzBoH,SAASC,SAASF,KAAK3G,IAAL,CAAU,QAAV,CAAT,EAA8B,EAA9B,CAAf;MACM8G,QAAQD,SAASF,KAAK3G,IAAL,CAAU,OAAV,CAAT,EAA6B,EAA7B,KAAoC,EAAlD;;;;;MAKI,CAAC4G,UAAU,EAAX,IAAiB,EAAjB,IAAuBE,QAAQ,EAAnC,EAAuC;SAChC3F,MAAL;GADF,MAEO,IAAIyF,MAAJ,EAAY;;;;SAIZ3G,UAAL,CAAgB,QAAhB;;;SAGKT,CAAP;;;;;AAKF,SAASuH,aAAT,CAAuBJ,IAAvB,EAA6BnH,CAA7B,EAAgC;MAC1BiD,UAAUxD,IAAV,CAAe0H,KAAK3G,IAAL,CAAU,KAAV,CAAf,CAAJ,EAAsC;SAC/BmB,MAAL;;;SAGK3B,CAAP;;;AAGF,AAAe,SAASwH,WAAT,CAAqBC,QAArB,EAA+BzH,CAA/B,EAAkC;WACtCwB,IAAT,CAAc,KAAd,EAAqBrB,IAArB,CAA0B,UAACiB,KAAD,EAAQL,GAAR,EAAgB;QAClCoG,OAAOnH,EAAEe,GAAF,CAAb;;mBAEeoG,IAAf,EAAqBnH,CAArB;kBACcmH,IAAd,EAAoBnH,CAApB;GAJF;;SAOOA,CAAP;;;ACnCa,SAAS0H,aAAT,CAAuBC,OAAvB,EAAgC3H,CAAhC,EAAmC;IAC9CkD,kBAAkB5E,IAAlB,CAAuB,GAAvB,CAAF,EAA+BqJ,OAA/B,EAAwChG,MAAxC;;SAEO3B,CAAP;;;ACLF;;;;AAGA,AAAe,SAAS4H,UAAT,CAAoBD,OAApB,EAA6B3H,CAA7B,EAAgC;;MAEvC6H,SAAS7H,EAAE,IAAF,EAAQ2H,OAAR,CAAf;MACIE,OAAOrF,MAAP,GAAgB,CAApB,EAAuB;WACdrC,IAAP,CAAY,UAACiB,KAAD,EAAQf,IAAR;aAAiBL,EAAEK,IAAF,EAAQsB,MAAR,EAAjB;KAAZ;GADF,MAEO;WACExB,IAAP,CAAY,UAACiB,KAAD,EAAQf,IAAR,EAAiB;oBACbL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0B,IAA1B;KADF;;;SAKKA,CAAP;;;ACZF,SAAS8H,qBAAT,CAA+BL,QAA/B,EAAyC;;WAE9BjG,IAAT,CAAc,GAAd,EAAmBrB,IAAnB,CAAwB,UAACiB,KAAD,EAAQf,IAAR,EAAiB;SAClCY,OAAL,GAAe+F,QAAQhG,OAAR,CAAgBX,KAAKY,OAArB,EAA8B8G,MAA9B,CAAqC,UAACC,GAAD,EAAMxH,IAAN,EAAe;UAC7DiD,mBAAmBhE,IAAnB,CAAwBe,IAAxB,CAAJ,EAAmC;4BACrBwH,GAAZ,qBAAkBxH,IAAlB,EAAyBH,KAAKY,OAAL,CAAaT,IAAb,CAAzB;;;aAGKwH,GAAP;KALa,EAMZ,EANY,CAAf;GADF;;;;;;;;;;AAkBF,AAAe,SAASC,eAAT,CAAyBR,QAAzB,EAAmC;wBAC1BA,QAAtB;;SAEOA,QAAP;;;AC3Ba,SAASS,WAAT,CAAqBT,QAArB,EAA+BzH,CAA/B,EAAkC;WACtCwB,IAAT,CAAc,GAAd,EAAmBrB,IAAnB,CAAwB,UAACiB,KAAD,EAAQ+E,CAAR,EAAc;QAC9BgC,KAAKnI,EAAEmG,CAAF,CAAX;QACIgC,GAAGC,IAAH,GAAUC,IAAV,OAAqB,EAAzB,EAA6BF,GAAGxG,MAAH;GAF/B;;SAKO3B,CAAP;;;ACNF;;;;;;AAMA,AAAO,IAAMgE,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,EASjC5F,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM6F,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,IAAI/F,MAAJ,QAAgB8F,yBAAuB7F,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;;;;AAMP,AAAO,IAAMgK,4BAA0B,CACrC,CAAC,SAAD,EAAY,gBAAZ,CADqC,EAErC,CAAC,OAAD,EAAU,gBAAV,CAFqC,EAGrC,CAAC,QAAD,EAAW,gBAAX,CAHqC,EAIrC,CAAC,OAAD,EAAU,WAAV,CAJqC,EAKrC,CAAC,OAAD,EAAU,YAAV,CALqC,EAMrC,CAAC,OAAD,EAAU,YAAV,CANqC,CAAhC;;AASP,AAAO,IAAMjE,gBAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,mBAAiB,IAAIjG,MAAJ,CAAWgG,cAAY/F,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAMiG,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,IAAInG,MAAJ,CAAWkG,uBAAqBjG,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMiK,sBAAoB,IAAIlK,MAAJ,CAAW,qBAAX,EAAkC,GAAlC,CAA1B;;;;;;AAMP,AAAO,IAAMoG,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,IAAIrG,MAAJ,CAAWoG,uBAAqBnG,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA,AAGA,AAGA;;AAEA,AAAO,IAAMyG,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,IAAI3G,MAAJ,QAAgB0G,mBAAiBzG,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAM2G,wBAAsBjB,gCAA8B1F,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,IAAM6G,wBAAsBlB,gCAA8B3F,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,AAGA,AAAO,IAAMkK,yBAAuB,IAAInK,MAAJ,CAAW,mBAAX,EAAgC,GAAhC,CAA7B;AACP,AAAO,IAAMoK,uBAAqB,IAAIpK,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAA3B;AACP,AAAO,IAAMqK,aAAW,IAAIrK,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB,CAEP;;AC3SA;AACA,AAAe,SAASsK,SAAT,CAAmBtI,IAAnB,EAAyB;MAChCkF,UAAUlF,KAAKG,IAAL,CAAU,OAAV,CAAhB;MACMgF,KAAKnF,KAAKG,IAAL,CAAU,IAAV,CAAX;MACIoI,QAAQ,CAAZ;;MAEIpD,EAAJ,EAAQ;;QAEFhB,oBAAkB/E,IAAlB,CAAuB+F,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;QAEEd,oBAAkBjF,IAAlB,CAAuB+F,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;;;MAIAD,OAAJ,EAAa;QACPqD,UAAU,CAAd,EAAiB;;;UAGXpE,oBAAkB/E,IAAlB,CAAuB8F,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;UAEEb,oBAAkBjF,IAAlB,CAAuB8F,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;;;;;;QAOAjB,iBAAe7E,IAAf,CAAoB8F,OAApB,CAAJ,EAAkC;eACvB,EAAT;;;;;;;QAOEgD,oBAAkB9I,IAAlB,CAAuB8F,OAAvB,CAAJ,EAAqC;eAC1B,EAAT;;;;SAIGqD,KAAP;;;ACpDF;;;AAGA,AAAe,SAASC,QAAT,CAAkBvI,KAAlB,EAAyB;SAC/BwI,WAAWxI,MAAME,IAAN,CAAW,OAAX,CAAX,KAAmC,IAA1C;;;ACJF;AACA,AAAe,SAASuI,WAAT,CAAqBX,IAArB,EAA2B;SACjC,CAACA,KAAKY,KAAL,CAAW,IAAX,KAAoB,EAArB,EAAyBxG,MAAhC;;;ACFF,IAAMyG,QAAQ,IAAI5K,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAd;;AAEA,AAAe,SAAS6K,WAAT,CAAqBC,UAArB,EAAgD;MAAfxG,OAAe,yDAAL,GAAK;;MACvDyG,SAASD,aAAa,EAA5B;;MAEIC,SAAS,CAAb,EAAgB;QACVC,oBAAJ;;;;;;;QAOIJ,MAAMxJ,IAAN,CAAWkD,OAAX,CAAJ,EAAyB;oBACTyG,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,CAAwBpJ,IAAxB,EAA8B;MACvCuI,QAAQ,CAAZ;MACMR,OAAO/H,KAAK+H,IAAL,GAAYC,IAAZ,EAAb;MACMc,aAAaf,KAAK5F,MAAxB;;;MAGI2G,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,CAAkBrJ,KAAlB,EAAyBN,CAAzB,EAA4B4I,KAA5B,EAAmC;QAC1CpI,IAAN,CAAW,OAAX,EAAoBoI,KAApB;SACOtI,KAAP;;;ACEa,SAASsJ,QAAT,CAAkBtJ,KAAlB,EAAyBN,CAAzB,EAA4B6J,MAA5B,EAAoC;MAC7C;QACIjB,QAAQkB,eAAexJ,KAAf,EAAsBN,CAAtB,IAA2B6J,MAAzC;aACSvJ,KAAT,EAAgBN,CAAhB,EAAmB4I,KAAnB;GAFF,CAGE,OAAOmB,CAAP,EAAU;;;;SAILzJ,KAAP;;;ACXF;AACA,AAAe,SAAS0J,WAAT,CAAqB3J,IAArB,EAA2BL,CAA3B,EAA8B4I,KAA9B,EAAqC;MAC5CqB,SAAS5J,KAAK4J,MAAL,EAAf;MACIA,MAAJ,EAAY;aACDA,MAAT,EAAiBjK,CAAjB,EAAoB4I,QAAQ,IAA5B;;;SAGKvI,IAAP;;;ACFF;;;AAGA,AAAe,SAASyJ,cAAT,CAAwBxJ,KAAxB,EAA+BN,CAA/B,EAAsD;MAApBkK,WAAoB,yDAAN,IAAM;;MAC/DtB,QAAQC,SAASvI,KAAT,CAAZ;;MAEIsI,KAAJ,EAAW;WACFA,KAAP;;;UAGMuB,UAAU7J,KAAV,CAAR;;MAEI4J,WAAJ,EAAiB;aACNvB,UAAUrI,KAAV,CAAT;;;cAGUA,KAAZ,EAAmBN,CAAnB,EAAsB4I,KAAtB;;SAEOA,KAAP;;;AClBF;;AAEA,AAAe,SAASuB,SAAT,CAAmB7J,KAAnB,EAA0B;mBACnBA,MAAM9B,GAAN,CAAU,CAAV,CADmB;;MAC/BmE,OAD+B,cAC/BA,OAD+B;;;;;;MAMnC6F,uBAAqB/I,IAArB,CAA0BkD,OAA1B,CAAJ,EAAwC;WAC/B8G,eAAenJ,KAAf,CAAP;GADF,MAEO,IAAIqC,YAAY,KAAhB,EAAuB;WACrB,CAAP;GADK,MAEA,IAAI8F,qBAAmBhJ,IAAnB,CAAwBkD,OAAxB,CAAJ,EAAsC;WACpC,CAAP;GADK,MAEA,IAAI+F,WAASjJ,IAAT,CAAckD,OAAd,CAAJ,EAA4B;WAC1B,CAAC,CAAR;GADK,MAEA,IAAIA,YAAY,IAAhB,EAAsB;WACpB,CAAC,CAAR;;;SAGK,CAAP;;;ACjBF,SAAS+D,cAAT,CAAsBpG,KAAtB,EAA6BN,CAA7B,EAAgC;MAC1BM,MAAM9B,GAAN,CAAU,CAAV,CAAJ,EAAkB;qBACI8B,MAAM9B,GAAN,CAAU,CAAV,CADJ;;QACRmE,OADQ,cACRA,OADQ;;;QAGZA,YAAY,MAAhB,EAAwB;;oBAERrC,KAAd,EAAqBN,CAArB,EAAwB,KAAxB;;;;;AAKN,SAASoK,UAAT,CAAoB9J,KAApB,EAA2BN,CAA3B,EAA8B4I,KAA9B,EAAqC;MAC/BtI,KAAJ,EAAW;mBACIA,KAAb,EAAoBN,CAApB;aACSM,KAAT,EAAgBN,CAAhB,EAAmB4I,KAAnB;;;;AAIJ,SAASyB,OAAT,CAAiBrK,CAAjB,EAAoBkK,WAApB,EAAiC;IAC7B,QAAF,EAAY5E,GAAZ,CAAgB,SAAhB,EAA2BnF,IAA3B,CAAgC,UAACiB,KAAD,EAAQf,IAAR,EAAiB;;;QAG3CC,QAAQN,EAAEK,IAAF,CAAZ;YACQsJ,SAASrJ,KAAT,EAAgBN,CAAhB,EAAmB8J,eAAexJ,KAAf,EAAsBN,CAAtB,EAAyBkK,WAAzB,CAAnB,CAAR;;QAEMI,UAAUhK,MAAM2J,MAAN,EAAhB;QACMM,WAAWJ,UAAU7J,KAAV,CAAjB;;eAEWgK,OAAX,EAAoBtK,CAApB,EAAuBuK,QAAvB,EAAiCL,WAAjC;QACII,OAAJ,EAAa;;;iBAGAA,QAAQL,MAAR,EAAX,EAA6BjK,CAA7B,EAAgCuK,WAAW,CAA3C,EAA8CL,WAA9C;;GAbJ;;SAiBOlK,CAAP;;;;;AAKF,AAAe,SAASwK,YAAT,CAAsBxK,CAAtB,EAA6C;MAApBkK,WAAoB,yDAAN,IAAM;;;;4BAGlChJ,OAAxB,CAAgC,gBAAqC;;;QAAnCuJ,cAAmC;QAAnBC,aAAmB;;MAC9DD,cAAL,SAAuBC,aAAvB,EAAwCvK,IAAxC,CAA6C,UAACiB,KAAD,EAAQf,IAAR,EAAiB;eACnDL,EAAEK,IAAF,EAAQ4J,MAAR,CAAeQ,cAAf,CAAT,EAAyCzK,CAAzC,EAA4C,EAA5C;KADF;GADF;;;;;;;MAWIqK,QAAQrK,CAAR,EAAWkK,WAAX,CAAJ;MACIG,QAAQrK,CAAR,EAAWkK,WAAX,CAAJ;;SAEOlK,CAAP;;;ACpEF,IAAM2K,eAAe,SAArB;;AAEA,AAAe,SAASC,eAAT,CAAyBxC,IAAzB,EAA+B;SACrCA,KAAKyC,OAAL,CAAaF,YAAb,EAA2B,GAA3B,EAAgCtC,IAAhC,EAAP;;;ACHF;;;;;AAKA,AAAe,SAASyC,cAAT,CAAwBpL,GAAxB,EAA6BqL,SAA7B,EAAwC;MAC/CC,UAAUD,UAAUvJ,IAAV,CAAe;WAAMyJ,GAAGxL,IAAH,CAAQC,GAAR,CAAN;GAAf,CAAhB;MACIsL,OAAJ,EAAa;WACJA,QAAQE,IAAR,CAAaxL,GAAb,EAAkB,CAAlB,CAAP;;;SAGK,IAAP;;;ACXF;;;;;;;;;;;;;;;;AAgBA,AAAO,IAAMyL,kBAAkB,IAAI9M,MAAJ,CAAW,0EAAX,EAAuF,GAAvF,CAAxB;;AAEP,AAAO,IAAM+M,eAAe,QAArB;;AAEP,AAAO,IAAMC,cAAc,WAApB;AACP,AAAO,IAAMC,cAAc,WAApB;;ACnBQ,SAASC,cAAT,CAAwB7L,GAAxB,EAA6B;MACpC8L,UAAU9L,IAAIsJ,KAAJ,CAAUmC,eAAV,CAAhB;MACI,CAACK,OAAL,EAAc,OAAO,IAAP;;MAERC,UAAUpE,SAASmE,QAAQ,CAAR,CAAT,EAAqB,EAArB,CAAhB;;;;SAIOC,UAAU,GAAV,GAAgBA,OAAhB,GAA0B,IAAjC;;;ACVa,SAASC,YAAT,CAAsBhM,GAAtB,EAA2B;SACjCA,IAAIiM,KAAJ,CAAU,GAAV,EAAe,CAAf,EAAkBd,OAAlB,CAA0B,KAA1B,EAAiC,EAAjC,CAAP;;;ACOF,SAASe,aAAT,CAAuBC,OAAvB,EAAgCzK,KAAhC,EAAuC0K,sBAAvC,EAA+D;MACzDC,cAAc,IAAlB;;;;MAII3K,QAAQ,CAAR,IAAakK,YAAY7L,IAAZ,CAAiBoM,OAAjB,CAAb,IAA0CA,QAAQrJ,MAAR,GAAiB,CAA/D,EAAkE;kBAClD,IAAd;;;;;MAKEpB,UAAU,CAAV,IAAeyK,QAAQG,WAAR,OAA0B,OAA7C,EAAsD;kBACtC,KAAd;;;;;MAKE5K,QAAQ,CAAR,IAAayK,QAAQrJ,MAAR,GAAiB,CAA9B,IAAmC,CAACsJ,sBAAxC,EAAgE;kBAChD,KAAd;;;SAGKC,WAAP;;;;;;AAMF,AAAe,SAASE,cAAT,CAAwBvM,GAAxB,EAA6BwM,MAA7B,EAAqC;MAC5CrM,YAAYqM,UAAUvM,IAAIC,KAAJ,CAAUF,GAAV,CAA5B;MACQyM,QAF0C,GAEjBtM,SAFiB,CAE1CsM,QAF0C;MAEhCC,IAFgC,GAEjBvM,SAFiB,CAEhCuM,IAFgC;MAE1BC,IAF0B,GAEjBxM,SAFiB,CAE1BwM,IAF0B;;;MAI9CP,yBAAyB,KAA7B;MACMQ,kBAAkBD,KAAKV,KAAL,CAAW,GAAX,EACvBY,OADuB,GAEvBxE,MAFuB,CAEhB,UAACC,GAAD,EAAMwE,UAAN,EAAkBpL,KAAlB,EAA4B;QAC9ByK,UAAUW,UAAd;;;QAGIX,QAAQ1J,QAAR,CAAiB,GAAjB,CAAJ,EAA2B;2BACU0J,QAAQF,KAAR,CAAc,GAAd,CADV;;;;UAClBc,eADkB;UACDC,OADC;;UAErBrB,YAAY5L,IAAZ,CAAiBiN,OAAjB,CAAJ,EAA+B;kBACnBD,eAAV;;;;;;QAMAtB,gBAAgB1L,IAAhB,CAAqBoM,OAArB,KAAiCzK,QAAQ,CAA7C,EAAgD;gBACpCyK,QAAQhB,OAAR,CAAgBM,eAAhB,EAAiC,EAAjC,CAAV;;;;;;;QAOE/J,UAAU,CAAd,EAAiB;+BACUgK,aAAa3L,IAAb,CAAkBoM,OAAlB,CAAzB;;;;QAIED,cAAcC,OAAd,EAAuBzK,KAAvB,EAA8B0K,sBAA9B,CAAJ,EAA2D;UACrDa,IAAJ,CAASd,OAAT;;;WAGK7D,GAAP;GAhCsB,EAiCrB,EAjCqB,CAAxB;;SAmCUmE,QAAV,UAAuBC,IAAvB,GAA8BE,gBAAgBC,OAAhB,GAA0BjO,IAA1B,CAA+B,GAA/B,CAA9B;;;AC3EF;;AAEA,IAAMsO,kBAAkB,IAAIvO,MAAJ,CAAW,QAAX,CAAxB;AACA,AAAe,SAASwO,cAAT,CAAwBzE,IAAxB,EAA8B;SACpCwE,gBAAgBnN,IAAhB,CAAqB2I,IAArB,CAAP;;;ACKF;;;;;AAKA,AAAe,SAAS0E,aAAT,CAAuBC,UAAvB,EAAmCC,QAAnC,EAA6ChN,CAA7C,EAAgD;MACzD,CAAC+M,WAAW9C,MAAX,GAAoBzH,MAAzB,EAAiC;WACxBuK,UAAP;;;MAGIE,wBAAwB3D,KAAKE,GAAL,CAAS,EAAT,EAAawD,WAAW,GAAxB,CAA9B;MACME,cAAclN,EAAE,aAAF,CAApB;;aAEWiK,MAAX,GAAoB1H,QAApB,GAA+BpC,IAA/B,CAAoC,UAACiB,KAAD,EAAQ+L,KAAR,EAAkB;QAC9CC,SAASpN,EAAEmN,KAAF,CAAf;;QAEI/I,4BAA0B3E,IAA1B,CAA+B0N,MAAMxK,OAArC,CAAJ,EAAmD;aAC1C,IAAP;;;QAGI0K,aAAaxE,SAASuE,MAAT,CAAnB;QACIC,UAAJ,EAAgB;UACVD,WAAWL,UAAf,EAA2B;oBACbO,MAAZ,CAAmBF,MAAnB;OADF,MAEO;YACDG,eAAe,CAAnB;;YAEMC,UAAUC,YAAYL,MAAZ,CAAhB;;;;YAIII,UAAU,IAAd,EAAoB;0BACF,EAAhB;;;;;YAKEA,WAAW,GAAf,EAAoB;0BACF,EAAhB;;;;;YAKEJ,OAAO5M,IAAP,CAAY,OAAZ,MAAyBuM,WAAWvM,IAAX,CAAgB,OAAhB,CAA7B,EAAuD;0BACrCwM,WAAW,GAA3B;;;YAGIU,WAAW7E,SAASuE,MAAT,IAAmBG,YAApC;;YAEIG,YAAYT,qBAAhB,EAAuC;iBAC9BC,YAAYI,MAAZ,CAAmBF,MAAnB,CAAP;SADF,MAEO,IAAID,MAAMxK,OAAN,KAAkB,GAAtB,EAA2B;cAC1BgL,eAAeP,OAAOhF,IAAP,EAArB;cACMwF,qBAAqBzE,WAAWwE,YAAX,CAA3B;;cAEIC,qBAAqB,EAArB,IAA2BJ,UAAU,IAAzC,EAA+C;mBACtCN,YAAYI,MAAZ,CAAmBF,MAAnB,CAAP;WADF,MAEO,IAAIQ,sBAAsB,EAAtB,IAA4BJ,YAAY,CAAxC,IACDX,eAAec,YAAf,CADH,EACiC;mBAC/BT,YAAYI,MAAZ,CAAmBF,MAAnB,CAAP;;;;;;WAMD,IAAP;GApDF;;SAuDOF,WAAP;;;ACzEF;;AAEA,AAAe,SAASW,gBAAT,CAA0B7N,CAA1B,EAA6B;MACtC+M,mBAAJ;MACIC,WAAW,CAAf;;IAEE,SAAF,EAAa7M,IAAb,CAAkB,UAACiB,KAAD,EAAQf,IAAR,EAAiB;QAC3BC,QAAQN,EAAEK,IAAF,CAAd;QACMuI,QAAQC,SAASvI,KAAT,CAAd;;QAEI8D,4BAA0B3E,IAA1B,CAA+BY,KAAKsC,OAApC,CAAJ,EAAkD;;;;;;QAM9CiG,QAAQoE,QAAZ,EAAsB;iBACTpE,KAAX;mBACatI,KAAb;;GAZJ;;;;MAkBI,CAACyM,UAAL,EAAiB;WACR/M,EAAE,MAAF,KAAaA,EAAE,GAAF,EAAO8N,KAAP,EAApB;;;eAGWhB,cAAcC,UAAd,EAA0BC,QAA1B,EAAoChN,CAApC,CAAb;;SAEO+M,UAAP;;;ACvBF,SAASgB,mBAAT,CAA6BzN,KAA7B,EAAoCN,CAApC,EAAuCgO,MAAvC,EAA+C;;;;;MAKzC1N,MAAM2N,QAAN,CAAe,qBAAf,CAAJ,EAA2C;;;;MAIrC/L,UAAU0I,gBAAgBtK,MAAM8H,IAAN,EAAhB,CAAhB;;MAEIW,YAAY7G,OAAZ,IAAuB,EAA3B,EAA+B;QACvBgM,SAASlO,EAAE,GAAF,EAAOM,KAAP,EAAckC,MAA7B;QACM2L,aAAanO,EAAE,OAAF,EAAWM,KAAX,EAAkBkC,MAArC;;;QAGI2L,aAAcD,SAAS,CAA3B,EAA+B;YACvBvM,MAAN;;;;QAIInC,gBAAgB0C,QAAQM,MAA9B;QACM4L,WAAWpO,EAAE,KAAF,EAASM,KAAT,EAAgBkC,MAAjC;;;;QAIIhD,gBAAgB,EAAhB,IAAsB4O,aAAa,CAAvC,EAA0C;YAClCzM,MAAN;;;;QAII6L,UAAUC,YAAYnN,KAAZ,CAAhB;;;;;QAKI0N,SAAS,EAAT,IAAeR,UAAU,GAAzB,IAAgChO,gBAAgB,EAApD,EAAwD;YAChDmC,MAAN;;;;;;QAMEqM,UAAU,EAAV,IAAgBR,UAAU,GAA9B,EAAmC;;;;UAI3B7K,UAAUrC,MAAM9B,GAAN,CAAU,CAAV,EAAamE,OAA7B;UACM0L,aAAa1L,YAAY,IAAZ,IAAoBA,YAAY,IAAnD;UACI0L,UAAJ,EAAgB;YACRC,eAAehO,MAAMiO,IAAN,EAArB;YACID,gBAAgB1D,gBAAgB0D,aAAalG,IAAb,EAAhB,EAAqCsB,KAArC,CAA2C,CAAC,CAA5C,MAAmD,GAAvE,EAA4E;;;;;YAKxE/H,MAAN;;;;QAII6M,cAAcxO,EAAE,QAAF,EAAYM,KAAZ,EAAmBkC,MAAvC;;;QAGIgM,cAAc,CAAd,IAAmBhP,gBAAgB,GAAvC,EAA4C;YACpCmC,MAAN;;;;;;;;;;;;;AAaN,AAAe,SAAS8M,SAAT,CAAmBhH,QAAnB,EAA6BzH,CAA7B,EAAgC;IAC3C6D,wBAAF,EAA4B4D,QAA5B,EAAsCtH,IAAtC,CAA2C,UAACiB,KAAD,EAAQf,IAAR,EAAiB;QACpDC,QAAQN,EAAEK,IAAF,CAAd;QACI2N,SAASnF,SAASvI,KAAT,CAAb;QACI,CAAC0N,MAAL,EAAa;eACFlE,eAAexJ,KAAf,EAAsBN,CAAtB,CAAT;eACSM,KAAT,EAAgBN,CAAhB,EAAmBgO,MAAnB;;;;QAIEA,SAAS,CAAb,EAAgB;YACRrM,MAAN;KADF,MAEO;;0BAEerB,KAApB,EAA2BN,CAA3B,EAA8BgO,MAA9B;;GAbJ;;SAiBOhO,CAAP;;;ACrGa,SAAS0O,YAAT,CAAsBjH,QAAtB,EAAgCzH,CAAhC,EAA+C;MAAZ2O,KAAY,yDAAJ,EAAI;;IAC1D5K,eAAF,EAAmB0D,QAAnB,EAA6BtH,IAA7B,CAAkC,UAACiB,KAAD,EAAQwN,MAAR,EAAmB;QAC7CC,UAAU7O,EAAE4O,MAAF,CAAhB;;;;;QAKI5O,EAAE6O,OAAF,EAAWpH,QAAX,EAAqBqH,OAArB,CAA6B,GAA7B,EAAkCtM,MAAlC,KAA6C,CAAjD,EAAoD;aAC3CqM,QAAQlN,MAAR,EAAP;;;;QAIEiJ,gBAAgB5K,EAAE4O,MAAF,EAAUxG,IAAV,EAAhB,MAAsCuG,KAA1C,EAAiD;aACxCE,QAAQlN,MAAR,EAAP;;;;;QAKEgH,UAAU3I,EAAE4O,MAAF,CAAV,IAAuB,CAA3B,EAA8B;aACrBC,QAAQlN,MAAR,EAAP;;;WAGKkN,OAAP;GArBF;;SAwBO7O,CAAP;;;AC5BF;;;AAEA,AAAe,SAAS+O,eAAT,CAAyBpH,OAAzB,EAAkC3H,CAAlC,EAAqC;;;;MAI9C8G,cAAc9G,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;MACI8G,cAAc9G,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;;SAEOA,CAAP;;;ACTF,SAASgP,UAAT,CAAoBhP,CAApB,EAAuBiP,OAAvB,EAAgCzO,IAAhC,EAAsC0O,QAAtC,EAAgD;UACxC1O,IAAN,QAAe0O,QAAf,EAAyB/O,IAAzB,CAA8B,UAACC,CAAD,EAAIC,IAAJ,EAAa;QACnCX,MAAMW,KAAKY,OAAL,CAAaT,IAAb,CAAZ;QACM2O,cAAcxP,IAAIhB,OAAJ,CAAYsQ,OAAZ,EAAqBvP,GAArB,CAApB;;SAEKuB,OAAL,CAAaT,IAAb,IAAqB2O,WAArB;GAJF;;;AAQF,AAAe,SAASC,iBAAT,CAA2BF,QAA3B,EAAqClP,CAArC,EAAwCN,GAAxC,EAA6C;GACzD,MAAD,EAAS,KAAT,EAAgBwB,OAAhB,CAAwB;WAAQ8N,WAAWhP,CAAX,EAAcN,GAAd,EAAmBc,IAAnB,EAAyB0O,QAAzB,CAAR;GAAxB;;;SAGOA,QAAP;;;ACdK,SAAS/F,UAAT,CAAoBf,IAApB,EAA0B;SACxBA,KAAKC,IAAL,GACKwC,OADL,CACa,MADb,EACqB,GADrB,EAEKrI,MAFZ;;;;;;AAQF,AAAO,SAASiL,WAAT,CAAqBnN,KAArB,EAA4B;MAC3B+O,kBAAkBlG,WAAW7I,MAAM8H,IAAN,EAAX,CAAxB;;MAEMkH,WAAWhP,MAAMkB,IAAN,CAAW,GAAX,EAAgB4G,IAAhB,EAAjB;MACMmH,aAAapG,WAAWmG,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,CACbxP,CADa,EAEbyP,SAFa,EAGbC,WAHa,EAKb;MADAjB,SACA,yDADY,IACZ;;MACMkB,aAAaF,UAAU/N,MAAV,CAAiB;WAAQgO,YAAYE,OAAZ,CAAoBC,IAApB,MAA8B,CAAC,CAAvC;GAAjB,CAAnB;;;;;;;;UAEWA,IAHX;;UAIQxO,OAAO,MAAb;UACMd,QAAQ,OAAd;;UAEMuP,QAAQ9P,YAAUqB,IAAV,UAAmBwO,IAAnB,QAAd;;;;;UAKME,SACJD,MAAMzM,GAAN,CAAU,UAACjC,KAAD,EAAQf,IAAR;eAAiBL,EAAEK,IAAF,EAAQG,IAAR,CAAaD,KAAb,CAAjB;OAAV,EACMyP,OADN,GAEMtO,MAFN,CAEa;eAAQ0G,SAAS,EAAjB;OAFb,CADF;;;;;;UASI2H,OAAOvN,MAAP,KAAkB,CAAtB,EAAyB;YACnByN,kBAAJ;;;YAGIxB,SAAJ,EAAe;sBACDyB,UAAUH,OAAO,CAAP,CAAV,EAAqB/P,CAArB,CAAZ;SADF,MAEO;sBACO+P,OAAO,CAAP,CAAZ;;;;aAGKE;;;;;yBA5BQN,UAAnB,8HAA+B;;;;;;;;;;;;;;;;;;;;;;SAiCxB,IAAP;;;AC3CF,SAASQ,UAAT,CAAoB7P,KAApB,EAA2B8P,WAA3B,EAAwC;;;MAGlC9P,MAAMiC,QAAN,GAAiBC,MAAjB,GAA0B4N,WAA9B,EAA2C;WAClC,KAAP;;;MAGEC,cAAc/P,KAAd,CAAJ,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;;;;AAMF,AAAe,SAASgQ,oBAAT,CACbtQ,CADa,EAEbuQ,SAFa,EAKb;MAFAH,WAEA,yDAFc,CAEd;MADAI,QACA,yDADW,IACX;;;;;;yBACuBD,SAAvB,8HAAkC;UAAvBjN,QAAuB;;UAC1BwM,QAAQ9P,EAAEsD,QAAF,CAAd;;;;UAIIwM,MAAMtN,MAAN,KAAiB,CAArB,EAAwB;YAChBlC,QAAQN,EAAE8P,MAAM,CAAN,CAAF,CAAd;;YAEIK,WAAW7P,KAAX,EAAkB8P,WAAlB,CAAJ,EAAoC;cAC9BlO,gBAAJ;cACIsO,QAAJ,EAAc;sBACFlQ,MAAM8H,IAAN,EAAV;WADF,MAEO;sBACK9H,MAAMmQ,IAAN,EAAV;;;cAGEvO,OAAJ,EAAa;mBACJA,OAAP;;;;;;;;;;;;;;;;;;;;SAMD,IAAP;;;AChDF;AACA,AAAe,SAASgO,SAAT,CAAmB9H,IAAnB,EAAyBpI,CAAzB,EAA4B;;;MAGnC0Q,YAAY1Q,aAAWoI,IAAX,cAA0BA,IAA1B,EAAlB;SACOsI,cAAc,EAAd,GAAmBtI,IAAnB,GAA0BsI,SAAjC;;;ACLa,SAASL,aAAT,CAAuB/P,KAAvB,EAA8B;MACrCwC,UAAUxC,MAAMwC,OAAN,GAAgBkN,OAAhB,EAAhB;MACMW,gBAAgB7N,QAAQtB,IAAR,CAAa,UAACyI,MAAD,EAAY;QACvCxE,aAAgBwE,OAAOhJ,OAAP,CAAe2P,KAA/B,SAAwC3G,OAAOhJ,OAAP,CAAeuE,EAA7D;WACOC,WAAWtD,QAAX,CAAoB,SAApB,CAAP;GAFoB,CAAtB;;SAKOwO,kBAAkBE,SAAzB;;;ACPF;;;;AAIA,AAAe,SAASC,gBAAT,CAA0BxQ,KAA1B,EAAiC;SACvCA,MAAM8H,IAAN,GAAaC,IAAb,GAAoB7F,MAApB,IAA8B,GAArC;;;ACHa,SAASuO,WAAT,CAAqB/Q,CAArB,EAAwB;SAC9BA,EAAE2E,cAAF,EAAkBnC,MAAlB,GAA2B,CAAlC;;;ACHF;AACA,AAAO,IAAMwO,kBAAkB,wCAAxB;;;;AAIP,AAAO,IAAMC,eAAe,IAAI5S,MAAJ,CAAW,aAAX,EAA0B,GAA1B,CAArB;AACP,AAYA,AASA;AACA,AAAO,IAAM6S,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,OAAO/S,IAAP,CAAY,GAAZ,CAAlB;AACA,IAAMiT,aAAa,qCAAnB;AACA,IAAMC,aAAa,wCAAnB;AACA,AAAO,IAAMC,oBACX,IAAIpT,MAAJ,OAAekT,UAAf,WAA+BC,UAA/B,wBAA4DF,SAA5D,QAA0E,IAA1E,CADK;;;;;AAMP,AAAO,IAAMI,qBAAqB,gBAA3B;;AAEP,AAAO,IAAMC,oBACX,IAAItT,MAAJ,CAAW,2BAAX,EAAwC,GAAxC,CADK;;ACtDP;;AAEA,AAAe,SAASuT,WAAT,CAAqBC,MAArB,EAA6B;SACnCA,OAAOhH,OAAP,CAAemG,eAAf,EAAgC,IAAhC,EAAsC3I,IAAtC,EAAP;;;ACHa,SAASzG,OAAT,CAAekQ,YAAf,EAA6B;iBAC3BA,aAAazJ,IAAb,EAAf;MACI0J,SAASC,QAAT,CAAkBF,YAAlB,CAAJ,EAAqC;WAC5BA,YAAP;;;SAGK,IAAP;;;ACJF;;AAEA,AAAe,SAASG,QAAT,CAAkBC,GAAlB,QAA8B;MAALlS,CAAK,QAALA,CAAK;;;MAEvCkS,IAAI1P,MAAJ,GAAa,IAAb,IAAqB0P,IAAI1P,MAAJ,GAAa,CAAtC,EAAyC,OAAO,IAAP;;MAEnC2P,UAAUjC,UAAUgC,GAAV,EAAelS,CAAf,CAAhB;;;;MAIIiR,aAAaxR,IAAb,CAAkB0S,OAAlB,CAAJ,EAAgC,OAAO,IAAP;;SAEzBA,QAAQ9J,IAAR,EAAP;;;ACfF;;;;AAIA,AAOA,AAAO,SAAS+J,eAAT,CAAyBC,UAAzB,EAAqC;SACnC,CAACA,WAAWrJ,KAAX,CAAiByI,iBAAjB,KAAuC,EAAxC,EACWnT,IADX,CACgB,GADhB,EAEWuM,OAFX,CAEmBuG,qBAFnB,EAE0C,GAF1C,EAGWvG,OAHX,CAGmBsG,sBAHnB,EAG2C,UAH3C,EAIWtG,OAJX,CAImBqG,oBAJnB,EAIyC,IAJzC,EAKW7I,IALX,EAAP;;;;;AAUF,AAAe,SAASiK,kBAAT,CAA4BD,UAA5B,EAAwC;MACjDE,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;;;ACnBF;;AACA,AAAe,SAASC,gBAAT,CACbjL,OADa,QAQb;MALE3H,CAKF,QALEA,CAKF;mCAJE6S,kBAIF;MAJEA,kBAIF,yCAJuB,IAIvB;wBAHElE,KAGF;MAHEA,KAGF,8BAHU,EAGV;sBAFEjP,GAEF;MAFEA,GAEF,4BAFQ,EAER;;;;kBAGgBiI,OAAhB,EAAyB3H,CAAzB;;;cAGY2H,OAAZ,EAAqB3H,CAArB;;;;gBAIc2H,OAAd,EAAuB3H,CAAvB;;;;;aAKW2H,OAAX,EAAoB3H,CAApB;;;eAGa2H,OAAb,EAAsB3H,CAAtB,EAAyB2O,KAAzB;;;oBAGkBhH,OAAlB,EAA2B3H,CAA3B,EAA8BN,GAA9B;;;kBAGgBiI,OAAhB;;;;;YAKUA,OAAV,EAAmB3H,CAAnB,EAAsB6S,kBAAtB;;;cAGYlL,OAAZ,EAAqB3H,CAArB;;SAEO2H,OAAP;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AClDa,SAASmL,UAAT,CAAoBnE,KAApB,QAAuC;MAAVjP,GAAU,QAAVA,GAAU;MAALM,CAAK,QAALA,CAAK;;;;MAGhD0R,mBAAmBjS,IAAnB,CAAwBkP,KAAxB,CAAJ,EAAoC;YAC1BoE,kBAAkBpE,KAAlB,EAAyBjP,GAAzB,CAAR;;;;;MAKEiP,MAAMnM,MAAN,GAAe,GAAnB,EAAwB;;QAEhBwQ,KAAKhT,EAAE,IAAF,CAAX;QACIgT,GAAGxQ,MAAH,KAAc,CAAlB,EAAqB;cACXwQ,GAAG5K,IAAH,EAAR;;;;;SAKG8H,UAAUvB,KAAV,EAAiB3O,CAAjB,EAAoBqI,IAApB,EAAP;;;ACdF,SAAS4K,sBAAT,CAAgCC,UAAhC,EAA4C9K,IAA5C,EAAkD;;;;MAI5C8K,WAAW1Q,MAAX,IAAqB,CAAzB,EAA4B;;;;;UAIpB2Q,aAAaD,WAAWnL,MAAX,CAAkB,UAACC,GAAD,EAAMoL,SAAN,EAAoB;YACnDA,SAAJ,IAAiBpL,IAAIoL,SAAJ,IAAiBpL,IAAIoL,SAAJ,IAAiB,CAAlC,GAAsC,CAAvD;eACOpL,GAAP;OAFiB,EAGhB,EAHgB,CAAnB;;kCAMEhB,QAAQhG,OAAR,CAAgBmS,UAAhB,EACQpL,MADR,CACe,UAACC,GAAD,EAAMf,GAAN,EAAc;YAChBe,IAAI,CAAJ,IAASmL,WAAWlM,GAAX,CAAb,EAA8B;iBACrB,CAACA,GAAD,EAAMkM,WAAWlM,GAAX,CAAN,CAAP;;;eAGKe,GAAP;OANT,EAOU,CAAC,CAAD,EAAI,CAAJ,CAPV,CAVwB;;;;UASnBqL,OATmB;UASVC,SATU;;;;;;;UAuBtBA,aAAa,CAAb,IAAkBD,QAAQ7Q,MAAR,IAAkB,CAAxC,EAA2C;qBAC5B4F,KAAKuD,KAAL,CAAW0H,OAAX,CAAb;;;UAGIE,YAAY,CAACL,WAAW,CAAX,CAAD,EAAgBA,WAAWxJ,KAAX,CAAiB,CAAC,CAAlB,CAAhB,CAAlB;UACM8J,aAAaD,UAAUxL,MAAV,CAAiB,UAACC,GAAD,EAAMyL,GAAN;eAAczL,IAAIxF,MAAJ,GAAaiR,IAAIjR,MAAjB,GAA0BwF,GAA1B,GAAgCyL,GAA9C;OAAjB,EAAoE,EAApE,CAAnB;;UAEID,WAAWhR,MAAX,GAAoB,EAAxB,EAA4B;;aACnBgR;;;;;WAGFpL;;;;;;;SAGF,IAAP;;;AAGF,SAASsL,oBAAT,CAA8BR,UAA9B,EAA0CxT,GAA1C,EAA+C;;;;;;;mBAO5BC,IAAIC,KAAJ,CAAUF,GAAV,CAP4B;;MAOrC0M,IAPqC,cAOrCA,IAPqC;;MAQvCuH,cAAcvH,KAAKvB,OAAL,CAAa8G,iBAAb,EAAgC,EAAhC,CAApB;;MAEMiC,YAAYV,WAAW,CAAX,EAAclH,WAAd,GAA4BnB,OAA5B,CAAoC,GAApC,EAAyC,EAAzC,CAAlB;MACMgJ,iBAAiBC,MAAMC,WAAN,CAAkBH,SAAlB,EAA6BD,WAA7B,CAAvB;;MAEIE,iBAAiB,GAAjB,IAAwBD,UAAUpR,MAAV,GAAmB,CAA/C,EAAkD;WACzC0Q,WAAWxJ,KAAX,CAAiB,CAAjB,EAAoBpL,IAApB,CAAyB,EAAzB,CAAP;;;MAGI0V,UAAUd,WAAWxJ,KAAX,CAAiB,CAAC,CAAlB,EAAqB,CAArB,EAAwBsC,WAAxB,GAAsCnB,OAAtC,CAA8C,GAA9C,EAAmD,EAAnD,CAAhB;MACMoJ,eAAeH,MAAMC,WAAN,CAAkBC,OAAlB,EAA2BL,WAA3B,CAArB;;MAEIM,eAAe,GAAf,IAAsBD,QAAQxR,MAAR,IAAkB,CAA5C,EAA+C;WACtC0Q,WAAWxJ,KAAX,CAAiB,CAAjB,EAAoB,CAAC,CAArB,EAAwBpL,IAAxB,CAA6B,EAA7B,CAAP;;;SAGK,IAAP;;;;;AAKF,AAAe,SAASyU,iBAAT,CAA2BpE,KAA3B,EAA4C;MAAVjP,GAAU,yDAAJ,EAAI;;;;MAGnDwT,aAAavE,MAAMhD,KAAN,CAAY+F,kBAAZ,CAAnB;MACIwB,WAAW1Q,MAAX,KAAsB,CAA1B,EAA6B;WACpBmM,KAAP;;;MAGEuF,WAAWjB,uBAAuBC,UAAvB,EAAmCvE,KAAnC,CAAf;MACIuF,QAAJ,EAAc,OAAOA,QAAP;;aAEHR,qBAAqBR,UAArB,EAAiCxT,GAAjC,CAAX;MACIwU,QAAJ,EAAc,OAAOA,QAAP;;;;SAIPvF,KAAP;;;AC3FF,IAAMwF,WAAW;UACPvC,WADO;gBAEDwC,OAFC;OAGVnC,QAHU;iBAIAK,kBAJA;WAKN+B,gBALM;SAMRvB;CANT,CAUA,AAEA,AACA,AACA,AACA,AACA,AACA,AACA;;ACfA;;;;;;;;;;;AAWA,AAAe,SAASwB,eAAT,CAAyBtU,CAAzB,EAA4BuU,IAA5B,EAAkC;;;;;;;MAO3CA,KAAKlP,uBAAT,EAAkC;QAC5BA,wBAAwBrF,CAAxB,CAAJ;;;MAGE6G,oBAAoB7G,CAApB,CAAJ;MACIwK,aAAaxK,CAAb,EAAgBuU,KAAKrK,WAArB,CAAJ;MACMsK,gBAAgB3G,iBAAiB7N,CAAjB,CAAtB;;SAEOwU,aAAP;;;AC3BF,IAAMC,0BAA0B;eACjB;6BACc,IADd;iBAEE,IAFF;wBAGS;GAJQ;;;;;;;;;;;;;;;;;;;;;SAAA,yBA0BGF,IA1BH,EA0BS;QAA7BvU,CAA6B,QAA7BA,CAA6B;QAA1ByQ,IAA0B,QAA1BA,IAA0B;QAApB9B,KAAoB,QAApBA,KAAoB;QAAbjP,GAAa,QAAbA,GAAa;;wBACzB,KAAKgV,WAAjB,EAAiCH,IAAjC;;QAEIvU,KAAKoC,QAAQC,IAAR,CAAaoO,IAAb,CAAT;;;;QAIIpQ,OAAO,KAAKsU,cAAL,CAAoB3U,CAApB,EAAuB2O,KAAvB,EAA8BjP,GAA9B,EAAmC6U,IAAnC,CAAX;;QAEIzD,iBAAiBzQ,IAAjB,CAAJ,EAA4B;aACnB,KAAKuU,kBAAL,CAAwBvU,IAAxB,EAA8BL,CAA9B,CAAP;;;;;;;;;;2BAKgBgH,QAAQhG,OAAR,CAAgBuT,IAAhB,EAAsB7S,MAAtB,CAA6B;eAAK6S,KAAKM,CAAL,MAAY,IAAjB;OAA7B,CAAlB,8HAAuE;YAA5D5N,GAA4D;;aAChEA,GAAL,IAAY,KAAZ;YACI7E,QAAQC,IAAR,CAAaoO,IAAb,CAAJ;;eAEO,KAAKkE,cAAL,CAAoB3U,CAApB,EAAuB2O,KAAvB,EAA8BjP,GAA9B,EAAmC6U,IAAnC,CAAP;;YAEIzD,iBAAiBzQ,IAAjB,CAAJ,EAA4B;;;;;;;;;;;;;;;;;;;WAKvB,KAAKuU,kBAAL,CAAwBvU,IAAxB,EAA8BL,CAA9B,CAAP;GApD4B;;;;gBAAA,0BAwDfA,CAxDe,EAwDZ2O,KAxDY,EAwDLjP,GAxDK,EAwDA6U,IAxDA,EAwDM;WAC3BF,iBACGC,gBAAgBtU,CAAhB,EAAmBuU,IAAnB,CADH,EAEL;UAAA;0BAEsBA,KAAK1B,kBAF3B;kBAAA;;KAFK,CAAP;GAzD4B;;;;;;oBAAA,8BAsEXxS,IAtEW,EAsELL,CAtEK,EAsEF;QACtB,CAACK,IAAL,EAAW;aACF,IAAP;;;WAGKuK,gBAAgB5K,EAAEyQ,IAAF,CAAOpQ,IAAP,CAAhB,CAAP;;;;;;;CA3EJ,CAqFA;;AC9FA;;;;;;;AAOA,AAAO,IAAMyU,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;QAArBlV,CAAqB,QAArBA,CAAqB;QAAlBN,GAAkB,QAAlBA,GAAkB;QAAbyV,SAAa,QAAbA,SAAa;;;;QAGzBxG,cAAJ;;YAEQa,gBAAgBxP,CAAhB,EAAmB8U,sBAAnB,EAA2CK,SAA3C,CAAR;QACIxG,KAAJ,EAAW,OAAOmE,WAAWnE,KAAX,EAAkB,EAAEjP,QAAF,EAAOM,IAAP,EAAlB,CAAP;;;;YAIHsQ,qBAAqBtQ,CAArB,EAAwBgV,sBAAxB,CAAR;QACIrG,KAAJ,EAAW,OAAOmE,WAAWnE,KAAX,EAAkB,EAAEjP,QAAF,EAAOM,IAAP,EAAlB,CAAP;;;YAGHwP,gBAAgBxP,CAAhB,EAAmB+U,oBAAnB,EAAyCI,SAAzC,CAAR;QACIxG,KAAJ,EAAW,OAAOmE,WAAWnE,KAAX,EAAkB,EAAEjP,QAAF,EAAOM,IAAP,EAAlB,CAAP;;;YAGHsQ,qBAAqBtQ,CAArB,EAAwBiV,oBAAxB,CAAR;QACItG,KAAJ,EAAW,OAAOmE,WAAWnE,KAAX,EAAkB,EAAEjP,QAAF,EAAOM,IAAP,EAAlB,CAAP;;;WAGJ,EAAP;;CAvBJ,CA2BA;;ACxCA;;;;;;AAMA,AAAO,IAAMoV,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;QAAhBzV,CAAgB,QAAhBA,CAAgB;QAAbmV,SAAa,QAAbA,SAAa;;QACpBtD,eAAJ;;;;aAISrC,gBAAgBxP,CAAhB,EAAmBoV,gBAAnB,EAAqCD,SAArC,CAAT;QACItD,UAAUA,OAAOrP,MAAP,GAAgB6S,iBAA9B,EAAiD;aACxCzD,YAAYC,MAAZ,CAAP;;;;aAIOvB,qBAAqBtQ,CAArB,EAAwBsV,gBAAxB,EAA0C,CAA1C,CAAT;QACIzD,UAAUA,OAAOrP,MAAP,GAAgB6S,iBAA9B,EAAiD;aACxCzD,YAAYC,MAAZ,CAAP;;;;;;;;;;2BAK8B2D,mBAAhC,8HAAqD;;;;;YAAzClS,QAAyC;YAA/BoS,KAA+B;;YAC7CrV,OAAOL,EAAEsD,QAAF,CAAb;YACIjD,KAAKmC,MAAL,KAAgB,CAApB,EAAuB;cACf4F,OAAO/H,KAAK+H,IAAL,EAAb;cACIsN,MAAMjW,IAAN,CAAW2I,IAAX,CAAJ,EAAsB;mBACbwJ,YAAYxJ,IAAZ,CAAP;;;;;;;;;;;;;;;;;;;WAKC,IAAP;;CA7BJ,CAiCA;;AC9CA;;;;AAIA,AAAO,IAAMuN,2BAA2B,CACtC,wBADsC,EAEtC,aAFsC,EAGtC,SAHsC,EAItC,gBAJsC,EAKtC,WALsC,EAMtC,cANsC,EAOtC,UAPsC,EAQtC,UARsC,EAStC,SATsC,EAUtC,eAVsC,EAWtC,UAXsC,EAYtC,cAZsC,EAatC,qBAbsC,EActC,cAdsC,EAetC,SAfsC,EAgBtC,MAhBsC,CAAjC;;;;;AAsBP,AAAO,IAAMC,2BAA2B,CACtC,4BADsC,EAEtC,oBAFsC,EAGtC,0BAHsC,EAItC,kBAJsC,EAKtC,oBALsC,EAMtC,kBANsC,EAOtC,iBAPsC,EAQtC,aARsC,EAStC,eATsC,EAUtC,qBAVsC,EAWtC,mBAXsC,EAYtC,cAZsC,EAatC,aAbsC,EActC,YAdsC,EAetC,kBAfsC,EAgBtC,WAhBsC,EAiBtC,UAjBsC,CAAjC;;;;;AAuBP,IAAMC,kBAAkB,mDAAxB;AACA,AAAO,IAAMC,yBAAyB;;AAEpC,IAAIzX,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAFoC;;;;AAMpC,IAAIA,MAAJ,CAAW,6BAAX,EAA0C,GAA1C,CANoC;;AAQpC,IAAIA,MAAJ,iBAAyBwX,eAAzB,kBAAuD,GAAvD,CARoC,CAA/B;;ACrCP,IAAME,gCAAgC;SAAA,yBACL;QAArB/V,CAAqB,QAArBA,CAAqB;QAAlBN,GAAkB,QAAlBA,GAAkB;QAAbyV,SAAa,QAAbA,SAAa;;QACzBa,sBAAJ;;;;oBAIgBxG,gBAAgBxP,CAAhB,EAAmB2V,wBAAnB,EAA6CR,SAA7C,EAAwD,KAAxD,CAAhB;QACIa,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;;oBAIH1F,qBAAqBtQ,CAArB,EAAwB4V,wBAAxB,CAAhB;QACII,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;oBAGHlL,eAAepL,GAAf,EAAoBoW,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,IAAIhY,MAAJ,CAAW+X,8BAA8B9X,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAMgY,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,IAAIlY,MAAJ,CAAWiY,8BAA8BhY,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAMkY,SAAS,gBAAf;AACP,AAAO,IAAMC,SAAS,kBAAf;;AC3CP,SAASC,MAAT,CAAgBpW,KAAhB,EAAuB;UACXA,MAAME,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCF,MAAME,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;;AAIF,AAAO,SAASmW,aAAT,CAAuBjX,GAAvB,EAA4B;QAC3BA,IAAI2I,IAAJ,EAAN;MACIO,QAAQ,CAAZ;;MAEIyN,iCAAiC5W,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;MAGE6W,iCAAiC9W,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;;;MAKE8W,OAAO/W,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;MAGE+W,OAAOhX,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;;;SAKKkJ,KAAP;;;;AAIF,AAAO,SAASgO,SAAT,CAAmBzP,IAAnB,EAAyB;MAC1BA,KAAK3G,IAAL,CAAU,KAAV,CAAJ,EAAsB;WACb,CAAP;;;SAGK,CAAP;;;;;AAKF,AAAO,SAASqW,cAAT,CAAwB1P,IAAxB,EAA8B;MAC/ByB,QAAQ,CAAZ;MACMkO,aAAa3P,KAAKrE,OAAL,CAAa,QAAb,EAAuBgL,KAAvB,EAAnB;;MAEIgJ,WAAWtU,MAAX,KAAsB,CAA1B,EAA6B;aAClB,EAAT;;;MAGI8H,UAAUnD,KAAK8C,MAAL,EAAhB;MACI8M,iBAAJ;MACIzM,QAAQ9H,MAAR,KAAmB,CAAvB,EAA0B;eACb8H,QAAQL,MAAR,EAAX;;;GAGDK,OAAD,EAAUyM,QAAV,EAAoB7V,OAApB,CAA4B,UAACZ,KAAD,EAAW;QACjCgE,iBAAe7E,IAAf,CAAoBiX,OAAOpW,KAAP,CAApB,CAAJ,EAAwC;eAC7B,EAAT;;GAFJ;;SAMOsI,KAAP;;;;;AAKF,AAAO,SAASoO,cAAT,CAAwB7P,IAAxB,EAA8B;MAC/ByB,QAAQ,CAAZ;MACMqO,WAAW9P,KAAKrB,IAAL,EAAjB;MACMG,UAAUgR,SAASzY,GAAT,CAAa,CAAb,CAAhB;;MAEIyH,WAAWA,QAAQtD,OAAR,KAAoB,YAAnC,EAAiD;aACtC,EAAT;;;MAGE2B,iBAAe7E,IAAf,CAAoBiX,OAAOO,QAAP,CAApB,CAAJ,EAA2C;aAChC,EAAT;;;SAGKrO,KAAP;;;AAGF,AAAO,SAASsO,iBAAT,CAA2B/P,IAA3B,EAAiC;MAClCyB,QAAQ,CAAZ;;MAEMtB,QAAQwB,WAAW3B,KAAK3G,IAAL,CAAU,OAAV,CAAX,CAAd;MACM4G,SAAS0B,WAAW3B,KAAK3G,IAAL,CAAU,QAAV,CAAX,CAAf;MACM2W,MAAMhQ,KAAK3G,IAAL,CAAU,KAAV,CAAZ;;;MAGI8G,SAASA,SAAS,EAAtB,EAA0B;aACf,EAAT;;;;MAIEF,UAAUA,UAAU,EAAxB,EAA4B;aACjB,EAAT;;;MAGEE,SAASF,MAAT,IAAmB,CAAC+P,IAAIhV,QAAJ,CAAa,QAAb,CAAxB,EAAgD;QACxCiV,OAAO9P,QAAQF,MAArB;QACIgQ,OAAO,IAAX,EAAiB;;eACN,GAAT;KADF,MAEO;eACI9N,KAAK+N,KAAL,CAAWD,OAAO,IAAlB,CAAT;;;;SAIGxO,KAAP;;;AAGF,AAAO,SAAS0O,eAAT,CAAyBC,KAAzB,EAAgCnW,KAAhC,EAAuC;SACpCmW,MAAM/U,MAAN,GAAe,CAAhB,GAAqBpB,KAA5B;;;ACxGF;;;;;;;;AAQA,IAAMoW,+BAA+B;SAAA,yBACA;QAAzBxX,CAAyB,QAAzBA,CAAyB;QAAtBkC,OAAsB,QAAtBA,OAAsB;QAAbiT,SAAa,QAAbA,SAAa;;QAC7BsC,iBAAJ;;;;;;QAMMC,WACJlI,gBACExP,CADF,EAEEkW,wBAFF,EAGEf,SAHF,EAIE,KAJF,CADF;;QAQIuC,QAAJ,EAAc;iBACDtD,QAAWsD,QAAX,CAAX;;UAEID,QAAJ,EAAc,OAAOA,QAAP;;;;;;QAMVE,OAAO3X,EAAE,KAAF,EAASkC,OAAT,EAAkB8N,OAAlB,EAAb;QACM4H,YAAY,EAAlB;;SAEK1W,OAAL,CAAa,UAACH,GAAD,EAAMK,KAAN,EAAgB;UACrB+F,OAAOnH,EAAEe,GAAF,CAAb;UACMoW,MAAMhQ,KAAK3G,IAAL,CAAU,KAAV,CAAZ;;UAEI,CAAC2W,GAAL,EAAU;;UAENvO,QAAQ+N,cAAcQ,GAAd,CAAZ;eACSP,UAAUzP,IAAV,CAAT;eACS0P,eAAe1P,IAAf,CAAT;eACS6P,eAAe7P,IAAf,CAAT;eACS+P,kBAAkB/P,IAAlB,CAAT;eACSmQ,gBAAgBK,IAAhB,EAAsBvW,KAAtB,CAAT;;gBAEU+V,GAAV,IAAiBvO,KAAjB;KAbF;;gCAiBE5B,QAAQhG,OAAR,CAAgB4W,SAAhB,EAA2B7P,MAA3B,CAAkC,UAACC,GAAD,EAAMf,GAAN;aAChC2Q,UAAU3Q,GAAV,IAAiBe,IAAI,CAAJ,CAAjB,GAA0B,CAACf,GAAD,EAAM2Q,UAAU3Q,GAAV,CAAN,CAA1B,GAAkDe,GADlB;KAAlC,EAEE,CAAC,IAAD,EAAO,CAAP,CAFF,CA5C+B;;;;QA2C1B6P,MA3C0B;QA2ClB7K,QA3CkB;;;QAgD7BA,WAAW,CAAf,EAAkB;iBACLoH,QAAWyD,MAAX,CAAX;;UAEIJ,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;2BAKOtB,wBAAvB,8HAAiD;YAAtC7S,QAAsC;;YACzChD,QAAQN,EAAEsD,QAAF,EAAYwK,KAAZ,EAAd;YACMqJ,MAAM7W,MAAME,IAAN,CAAW,KAAX,CAAZ;YACI2W,GAAJ,EAAS;qBACI/C,QAAW+C,GAAX,CAAX;cACIM,QAAJ,EAAc,OAAOA,QAAP;;;YAGVK,OAAOxX,MAAME,IAAN,CAAW,MAAX,CAAb;YACIsX,IAAJ,EAAU;qBACG1D,QAAW0D,IAAX,CAAX;cACIL,QAAJ,EAAc,OAAOA,QAAP;;;YAGVlX,QAAQD,MAAME,IAAN,CAAW,OAAX,CAAd;YACID,KAAJ,EAAW;qBACE6T,QAAW7T,KAAX,CAAX;cACIkX,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;;;;;;;;;WAIX,IAAP;;CA9EJ;;AAkFA,AAEA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AC7Ge,SAASM,eAAT,CAAyBnP,KAAzB,EAAgCoP,UAAhC,EAA4CF,IAA5C,EAAkD;;;;;;MAM3DlP,QAAQ,CAAZ,EAAe;QACPqP,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;WACOzP,QAAQ0P,YAAf;;;SAGK,CAAP;;;ACnBa,SAASC,aAAT,CAAuBjJ,QAAvB,EAAiC7D,OAAjC,EAA0C;;;;;MAKnD7C,QAAQ,CAAZ;;MAEI0C,YAAY7L,IAAZ,CAAiB6P,SAASjH,IAAT,EAAjB,CAAJ,EAAuC;QAC/BmQ,gBAAgBnR,SAASiI,QAAT,EAAmB,EAAnB,CAAtB;;;;QAIIkJ,gBAAgB,CAApB,EAAuB;cACb,CAAC,EAAT;KADF,MAEO;cACGlP,KAAKE,GAAL,CAAS,CAAT,EAAY,KAAKgP,aAAjB,CAAR;;;;;;QAME/M,WAAWA,WAAW+M,aAA1B,EAAyC;eAC9B,EAAT;;;;SAIG5P,KAAP;;;AC5Ba,SAAS6P,eAAT,CAAyBhN,OAAzB,EAAkCiN,IAAlC,EAAwC;;;;MAIjDjN,WAAW,CAACiN,IAAhB,EAAsB;WACb,EAAP;;;SAGK,CAAP;;;ACRK,IAAMC,aAAW,IAAjB;;;;AAIP,AAAO,IAAM/T,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,IAAIxG,MAAJ,CAAWuG,wBAAsBtG,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;;;AAKP,AAAO,IAAMsa,sBAAoB,IAAIva,MAAJ,CAAW,4CAAX,EAAyD,GAAzD,CAA1B;;;;AAIP,AAAO,IAAMwa,qBAAmB,IAAIxa,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAzB;;;;AAIP,AAAO,IAAMya,sBAAoB,IAAIza,MAAJ,CAAW,yBAAX,EAAsC,GAAtC,CAA1B,CAEP;;ACjCe,SAAS0a,oBAAT,CAA8BjB,IAA9B,EAAoC;;MAE7CjT,2BAAyBpF,IAAzB,CAA8BqY,IAA9B,CAAJ,EAAyC;WAChC,CAAC,EAAR;;;SAGK,CAAP;;;eCRuBkB;;AAAzB,AAAe,SAAUA,KAAV;MAAgBC,KAAhB,yDAAwB,CAAxB;MAA2BxF,GAA3B,yDAAiC,CAAjC;;;;;gBACNwF,SAASxF,GADH;;;;;;iBAELwF,SAAS,CAFJ;;;;;;;;;;;;;;ACQf,SAASC,SAAT,CAAiBC,KAAjB,EAAwB;UACZA,MAAM3Y,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuC2Y,MAAM3Y,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;AAGF,AAAe,SAASqW,gBAAT,CAAwBsC,KAAxB,EAA+B;;;;MAIxC7O,UAAU6O,MAAMlP,MAAN,EAAd;MACImP,gBAAgB,KAApB;MACIC,gBAAgB,KAApB;MACIzQ,QAAQ,CAAZ;;QAEM3I,IAAN,CAAW+Y,MAAM,CAAN,EAAS,CAAT,CAAX,EAAwB9X,OAAxB,CAAgC,YAAM;QAChCoJ,QAAQ9H,MAAR,KAAmB,CAAvB,EAA0B;;;;QAIpB8W,aAAaJ,UAAQ5O,OAAR,EAAiB,GAAjB,CAAnB;;;;QAII,CAAC8O,aAAD,IAAkBtU,QAAQrF,IAAR,CAAa6Z,UAAb,CAAtB,EAAgD;sBAC9B,IAAhB;eACS,EAAT;;;;;;QAME,CAACD,aAAD,IAAkB3U,kBAAkBjF,IAAlB,CAAuB6Z,UAAvB,CAAlB,IACEzU,2BAAyBpF,IAAzB,CAA8B6Z,UAA9B,CADN,EACiD;UAC3C,CAAC9U,kBAAkB/E,IAAlB,CAAuB6Z,UAAvB,CAAL,EAAyC;wBACvB,IAAhB;iBACS,EAAT;;;;cAIMhP,QAAQL,MAAR,EAAV;GAzBF;;SA4BOrB,KAAP;;;AC/Ca,SAAS2Q,aAAT,CAAuBC,QAAvB,EAAiC;;;MAG1CV,oBAAkBrZ,IAAlB,CAAuB+Z,QAAvB,CAAJ,EAAsC;WAC7B,CAAC,GAAR;;;SAGK,CAAP;;;ACFa,SAASC,WAAT,CACb3B,IADa,EAEbE,UAFa,EAGb0B,OAHa,EAIb7Z,SAJa,EAKbyP,QALa,EAMbqK,YANa,EAOb;;MAEIA,aAAanY,IAAb,CAAkB;WAAOsW,SAASpY,GAAhB;GAAlB,MAA2CmR,SAA/C,EAA0D;WACjD,KAAP;;;;;MAKE,CAACiH,IAAD,IAASA,SAASE,UAAlB,IAAgCF,SAAS4B,OAA7C,EAAsD;WAC7C,KAAP;;;MAGME,QAZR,GAYqB/Z,SAZrB,CAYQ+Z,QAZR;;mBAa+Bja,IAAIC,KAAJ,CAAUkY,IAAV,CAb/B;;MAakB+B,QAblB,cAaQD,QAbR;;;;MAgBIC,aAAaD,QAAjB,EAA2B;WAClB,KAAP;;;;;MAKIE,WAAWhC,KAAKjN,OAAL,CAAa6O,OAAb,EAAsB,EAAtB,CAAjB;MACI,CAACf,WAASlZ,IAAT,CAAcqa,QAAd,CAAL,EAA8B;WACrB,KAAP;;;;;MAKEjV,2BAAyBpF,IAAzB,CAA8B6P,QAA9B,CAAJ,EAA6C;WACpC,KAAP;;;;MAIEA,SAAS9M,MAAT,GAAkB,EAAtB,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;ACpDa,SAASuX,YAAT,CAAsBjC,IAAtB,EAA4BkC,SAA5B,EAAuC;;;;;MAKhD,CAACA,UAAUva,IAAV,CAAeqY,IAAf,CAAL,EAA2B;WAClB,CAAC,EAAR;;;SAGK,CAAP;;;ACPa,SAASmC,iBAAT,CAA2BT,QAA3B,EAAqC;;MAE9CZ,oBAAkBnZ,IAAlB,CAAuB+Z,QAAvB,CAAJ,EAAsC;WAC7B,EAAP;;;SAGK,CAAP;;;ACHa,SAASU,aAAT,CAAuBV,QAAvB,EAAiC;;MAE1CX,mBAAiBpZ,IAAjB,CAAsB+Z,QAAtB,CAAJ,EAAqC;;;;;QAK/BZ,oBAAkBnZ,IAAlB,CAAuB+Z,QAAvB,CAAJ,EAAsC;aAC7B,CAAC,EAAR;;;;SAIG,CAAP;;;ACKK,SAASW,aAAT,CAAuBT,OAAvB,EAAgC;SAC9B,IAAIrb,MAAJ,OAAeqb,OAAf,EAA0B,GAA1B,CAAP;;;AAGF,SAASR,OAAT,CAAiBC,KAAjB,EAAwB7J,QAAxB,EAAkC;UACtBA,YAAY6J,MAAM/Q,IAAN,EAAtB,WAAsC+Q,MAAM3Y,IAAN,CAAW,OAAX,KAAuB,EAA7D,WAAmE2Y,MAAM3Y,IAAN,CAAW,IAAX,KAAoB,EAAvF;;;AAGF,AAAe,SAAS4Z,UAAT,OAOZ;MANDC,KAMC,QANDA,KAMC;MALDrC,UAKC,QALDA,UAKC;MAJD0B,OAIC,QAJDA,OAIC;MAHD7Z,SAGC,QAHDA,SAGC;MAFDG,CAEC,QAFDA,CAEC;+BADD2Z,YACC;MADDA,YACC,qCADc,EACd;;cACW9Z,aAAaF,IAAIC,KAAJ,CAAUoY,UAAV,CAAzB;MACMgC,YAAYG,cAAcT,OAAd,CAAlB;MACMhB,OAAO3H,YAAY/Q,CAAZ,CAAb;;;;;;;;;MASMsa,cAAcD,MAAMtS,MAAN,CAAa,UAACwS,aAAD,EAAgBC,IAAhB,EAAyB;;;;QAIlD1C,OAAOpM,aAAa8O,KAAKvZ,OAAL,CAAa6W,IAA1B,CAAb;QACMqB,QAAQnZ,EAAEwa,IAAF,CAAd;QACMlL,WAAW6J,MAAM/Q,IAAN,EAAjB;;QAEI,CAACqR,YAAY3B,IAAZ,EAAkBE,UAAlB,EAA8B0B,OAA9B,EAAuC7Z,SAAvC,EAAkDyP,QAAlD,EAA4DqK,YAA5D,CAAL,EAAgF;aACvEY,aAAP;;;;QAIE,CAACA,cAAczC,IAAd,CAAL,EAA0B;oBACVA,IAAd,IAAsB;eACb,CADa;0BAAA;;OAAtB;KADF,MAMO;oBACSA,IAAd,EAAoBxI,QAApB,GAAkCiL,cAAczC,IAAd,EAAoBxI,QAAtD,SAAkEA,QAAlE;;;QAGImL,eAAeF,cAAczC,IAAd,CAArB;QACM0B,WAAWN,QAAQC,KAAR,EAAe7J,QAAf,CAAjB;QACM7D,UAAUF,eAAeuM,IAAf,CAAhB;;QAEIlP,QAAQmR,aAAajC,IAAb,EAAmBkC,SAAnB,CAAZ;aACSC,kBAAkBT,QAAlB,CAAT;aACSU,cAAcV,QAAd,CAAT;aACSD,cAAcC,QAAd,CAAT;aACS3C,iBAAesC,KAAf,CAAT;aACSJ,qBAAqBjB,IAArB,CAAT;aACSW,gBAAgBhN,OAAhB,EAAyBiN,IAAzB,CAAT;aACSH,cAAcjJ,QAAd,EAAwB7D,OAAxB,CAAT;aACSsM,gBAAgBnP,KAAhB,EAAuBoP,UAAvB,EAAmCF,IAAnC,CAAT;;iBAEalP,KAAb,GAAqBA,KAArB;;WAEO2R,aAAP;GAvCkB,EAwCjB,EAxCiB,CAApB;;SA0COvT,QAAQhG,OAAR,CAAgBsZ,WAAhB,EAA6B9X,MAA7B,KAAwC,CAAxC,GAA4C,IAA5C,GAAmD8X,WAA1D;;;AClFF;;AAEA,IAAMI,8BAA8B;SAAA,yBACgB;QAAxC1a,CAAwC,QAAxCA,CAAwC;QAArCN,GAAqC,QAArCA,GAAqC;QAAhCG,SAAgC,QAAhCA,SAAgC;iCAArB8Z,YAAqB;QAArBA,YAAqB,qCAAN,EAAM;;gBACpC9Z,aAAaF,IAAIC,KAAJ,CAAUF,GAAV,CAAzB;;QAEMsY,aAAatM,aAAahM,GAAb,CAAnB;QACMga,UAAUzN,eAAevM,GAAf,EAAoBG,SAApB,CAAhB;;QAEMwa,QAAQra,EAAE,SAAF,EAAagQ,OAAb,EAAd;;QAEM2K,cAAcP,WAAW;kBAAA;4BAAA;sBAAA;0BAAA;UAAA;;KAAX,CAApB;;;QAUI,CAACO,WAAL,EAAkB,OAAO,IAAP;;;;QAIZC,UAAU5T,QAAQhG,OAAR,CAAgB2Z,WAAhB,EAA6B5S,MAA7B,CAAoC,UAACC,GAAD,EAAMwS,IAAN,EAAe;UAC3DK,aAAaF,YAAYH,IAAZ,CAAnB;aACOK,WAAWjS,KAAX,GAAmBZ,IAAIY,KAAvB,GAA+BiS,UAA/B,GAA4C7S,GAAnD;KAFc,EAGb,EAAEY,OAAO,CAAC,GAAV,EAHa,CAAhB;;;;QAOIgS,QAAQhS,KAAR,IAAiB,EAArB,EAAyB;aAChBgS,QAAQ9C,IAAf;;;WAGK,IAAP;;CAlCJ,CAuCA;;AClDO,IAAMgD,2BAA2B,CACtC,QADsC,CAAjC;;ACKP,SAASC,WAAT,CAAqBrb,GAArB,EAA0B;MAClBG,YAAYF,IAAIC,KAAJ,CAAUF,GAAV,CAAlB;MACQka,QAFgB,GAEH/Z,SAFG,CAEhB+Z,QAFgB;;SAGjBA,QAAP;;;AAGF,SAAS3X,MAAT,CAAgBvC,GAAhB,EAAqB;SACZ;YAAA;YAEGqb,YAAYrb,GAAZ;GAFV;;;AAMF,IAAMsb,sBAAsB;SAAA,yBACK;QAArBhb,CAAqB,QAArBA,CAAqB;QAAlBN,GAAkB,QAAlBA,GAAkB;QAAbyV,SAAa,QAAbA,SAAa;;QACvB8F,aAAajb,EAAE,qBAAF,CAAnB;QACIib,WAAWzY,MAAX,KAAsB,CAA1B,EAA6B;UACrBsV,OAAOmD,WAAWza,IAAX,CAAgB,MAAhB,CAAb;UACIsX,IAAJ,EAAU;eACD7V,OAAO6V,IAAP,CAAP;;;;QAIEoD,UAAU1L,gBAAgBxP,CAAhB,EAAmB8a,wBAAnB,EAA6C3F,SAA7C,CAAhB;QACI+F,OAAJ,EAAa;aACJjZ,OAAOiZ,OAAP,CAAP;;;WAGKjZ,OAAOvC,GAAP,CAAP;;CAfJ,CAoBA;;ACtCO,IAAMyb,yBAAyB,CACpC,gBADoC,EAEpC,qBAFoC,CAA/B;;ACSA,SAASvZ,OAAT,CAAeM,OAAf,EAAwBlC,CAAxB,EAA4C;MAAjBob,SAAiB,yDAAL,GAAK;;YACvClZ,QAAQ2I,OAAR,CAAgB,UAAhB,EAA4B,GAA5B,EAAiCxC,IAAjC,EAAV;SACOgT,UAAUnZ,OAAV,EAAmBkZ,SAAnB,EAA8B,EAAEE,SAAS,UAAX,EAA9B,CAAP;;;AAGF,IAAMC,0BAA0B;SAAA,yBACK;QAAzBvb,CAAyB,QAAzBA,CAAyB;QAAtBkC,OAAsB,QAAtBA,OAAsB;QAAbiT,SAAa,QAAbA,SAAa;;QAC3BqG,UAAUhM,gBAAgBxP,CAAhB,EAAmBmb,sBAAnB,EAA2ChG,SAA3C,CAAhB;QACIqG,OAAJ,EAAa;aACJ5Z,QAAMsO,UAAUsL,OAAV,EAAmBxb,CAAnB,CAAN,CAAP;;;QAGIob,YAAY,GAAlB;QACMK,eAAevZ,QAAQwH,KAAR,CAAc,CAAd,EAAiB0R,YAAY,CAA7B,CAArB;WACOxZ,QAAM5B,EAAEyb,YAAF,EAAgBrT,IAAhB,EAAN,EAA8BpI,CAA9B,EAAiCob,SAAjC,CAAP;;CATJ,CAaA;;ACvBA,IAAMM,4BAA4B;SAAA,yBACX;QAAXxZ,OAAW,QAAXA,OAAW;;QACblC,IAAIoC,QAAQC,IAAR,CAAaH,OAAb,CAAV;;QAEMkG,OAAOwC,gBAAgB5K,EAAE,KAAF,EAAS8N,KAAT,GAAiB1F,IAAjB,EAAhB,CAAb;WACOA,KAAKuD,KAAL,CAAW,IAAX,EAAiBnJ,MAAxB;;CALJ,CASA;;ACAA,IAAMmZ,mBAAmB;;UAEf,GAFe;SAGhBzG,sBAAsB0G,OAHN;iBAIR7F,8BAA8B6F,OAJtB;UAKfnG,uBAAuBmG,OALR;WAMdnH,wBAAwBmH,OAAxB,CAAgCC,IAAhC,CAAqCpH,uBAArC,CANc;gBAOT+C,6BAA6BoE,OAPpB;OAQlB3F,oBAAoB2F,OARF;eASVlB,4BAA4BkB,OATlB;gBAUTZ,oBAAoBY,OAVX;WAWdL,wBAAwBK,OAXV;aAYZF,0BAA0BE,OAZd;;SAAA,mBAcfnd,OAde,EAcN;QACPgS,IADO,GACEhS,OADF,CACPgS,IADO;;;QAGXA,IAAJ,EAAU;UACFzQ,IAAIoC,QAAQC,IAAR,CAAaoO,IAAb,CAAV;cACQzQ,CAAR,GAAYA,CAAZ;;;QAGI2O,QAAQ,KAAKA,KAAL,CAAWlQ,OAAX,CAAd;QACMuX,gBAAgB,KAAKA,aAAL,CAAmBvX,OAAnB,CAAtB;QACMoT,SAAS,KAAKA,MAAL,CAAYpT,OAAZ,CAAf;QACMyD,UAAU,KAAKA,OAAL,cAAkBzD,OAAlB,IAA2BkQ,YAA3B,IAAhB;QACMmD,eAAe,KAAKA,YAAL,cAAuBrT,OAAvB,IAAgCyD,gBAAhC,IAArB;QACMgQ,MAAM,KAAKA,GAAL,cAAczT,OAAd,IAAuByD,gBAAvB,IAAZ;QACM4Z,cAAc,KAAKA,WAAL,CAAiBrd,OAAjB,CAApB;QACM+c,UAAU,KAAKA,OAAL,cAAkB/c,OAAlB,IAA2ByD,gBAA3B,IAAhB;QACM6Z,YAAY,KAAKP,OAAL,cAAkB/c,OAAlB,IAA2ByD,gBAA3B,IAAlB;;wBACwB,KAAK8Z,YAAL,CAAkBvd,OAAlB,CAjBT;;QAiBPiB,GAjBO,iBAiBPA,GAjBO;QAiBFuc,MAjBE,iBAiBFA,MAjBE;;;WAmBR;kBAAA;oBAAA;qBAGUjG,iBAAiB,IAH3B;cAAA;gCAAA;sBAAA;8BAAA;cAAA;oBAAA;sBAAA;;KAAP;;CAjCJ,CAiDA;;ACzDe,SAASkG,YAAT,CAAsBxc,GAAtB,EAA2B;MAClCG,YAAYF,IAAIC,KAAJ,CAAUF,GAAV,CAAlB;MACQka,QAFgC,GAEnB/Z,SAFmB,CAEhC+Z,QAFgC;;MAGlCuC,aAAavC,SAASjO,KAAT,CAAe,GAAf,EAAoBjC,KAApB,CAA0B,CAAC,CAA3B,EAA8BpL,IAA9B,CAAmC,GAAnC,CAAnB;;SAEO0E,WAAW4W,QAAX,KAAwB5W,WAAWmZ,UAAX,CAAxB,IAAkDR,gBAAzD;;;ACVK,IAAMS,UAAU,cAAhB;;ACOP;AACA,AAAO,SAASC,gBAAT,CAA0BnN,QAA1B,EAAoClP,CAApC,QAAkD;MAAT4B,KAAS,QAATA,KAAS;;MACnD,CAACA,KAAL,EAAY,OAAO,IAAP;;IAEVA,MAAMtD,IAAN,CAAW,GAAX,CAAF,EAAmB4Q,QAAnB,EAA6BvN,MAA7B;;SAEOuN,QAAP;;;;AAIF,AAAO,SAASoN,iBAAT,CAA2BpN,QAA3B,EAAqClP,CAArC,SAAwD;MAAduc,UAAc,SAAdA,UAAc;;MACzD,CAACA,UAAL,EAAiB,OAAO,IAAP;;UAETvb,OAAR,CAAgBub,UAAhB,EAA4Brb,OAA5B,CAAoC,UAAC+F,GAAD,EAAS;QACrCuV,WAAWxc,EAAEiH,GAAF,EAAOiI,QAAP,CAAjB;QACM3O,QAAQgc,WAAWtV,GAAX,CAAd;;;QAGI,OAAO1G,KAAP,KAAiB,QAArB,EAA+B;eACpBJ,IAAT,CAAc,UAACiB,KAAD,EAAQf,IAAR,EAAiB;sBACfL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0Buc,WAAWtV,GAAX,CAA1B;OADF;KADF,MAIO,IAAI,OAAO1G,KAAP,KAAiB,UAArB,EAAiC;;eAE7BJ,IAAT,CAAc,UAACiB,KAAD,EAAQf,IAAR,EAAiB;YACvB4B,SAAS1B,MAAMP,EAAEK,IAAF,CAAN,EAAeL,CAAf,CAAf;;YAEI,OAAOiC,MAAP,KAAkB,QAAtB,EAAgC;wBAChBjC,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0BiC,MAA1B;;OAJJ;;GAXJ;;SAqBOiN,QAAP;;;AAGF,AAAO,SAASuN,MAAT,CAAgBlI,IAAhB,EAAsB;MACnBvU,CADmB,GAC8BuU,IAD9B,CACnBvU,CADmB;MAChBqB,IADgB,GAC8BkT,IAD9B,CAChBlT,IADgB;MACVqb,cADU,GAC8BnI,IAD9B,CACVmI,cADU;0BAC8BnI,IAD9B,CACMoI,WADN;MACMA,WADN,qCACoB,KADpB;;;MAGvB,CAACD,cAAL,EAAqB,OAAO,IAAP;;;;MAIjB,OAAOA,cAAP,KAA0B,QAA9B,EAAwC,OAAOA,cAAP;;MAEhCnM,SATmB,GASLmM,cATK,CASnBnM,SATmB;;;MAWrBqM,mBAAmBrM,UAAU/O,IAAV,CAAe;WAAYxB,EAAEsD,QAAF,EAAYd,MAAZ,KAAuB,CAAvB,IAA4BxC,EAAEsD,QAAF,EAAY8E,IAAZ,GAAmBC,IAAnB,OAA8B,EAAtE;GAAf,CAAzB;;MAEI,CAACuU,gBAAL,EAAuB,OAAO,IAAP;;;;;;;;MAQnBD,WAAJ,EAAiB;QACXzN,WAAWlP,EAAE4c,gBAAF,CAAf;;;aAGSC,IAAT,CAAc7c,EAAE,aAAF,CAAd;eACWkP,SAASjF,MAAT,EAAX;;eAEWqS,kBAAkBpN,QAAlB,EAA4BlP,CAA5B,EAA+B0c,cAA/B,CAAX;eACWL,iBAAiBnN,QAAjB,EAA2BlP,CAA3B,EAA8B0c,cAA9B,CAAX;;eAEWvI,SAAS9S,IAAT,EAAe6N,QAAf,EAAyBqF,IAAzB,CAAX;;WAEOvU,EAAEyQ,IAAF,CAAOvB,QAAP,CAAP;;;;MAII1O,OAAOoc,iBAAiB5T,KAAjB,CAAuBoT,OAAvB,CAAb;MACIna,eAAJ;;MAEIzB,IAAJ,EAAU;aACCR,EAAE4c,gBAAF,EAAoBpc,IAApB,CAAyBA,KAAK,CAAL,CAAzB,CAAT;GADF,MAEO;;aAEIR,EAAE4c,gBAAF,EAAoBxU,IAApB,EAAT;;SAEK+L,SAAS9S,IAAT,EAAeY,MAAf,EAAuBsS,IAAvB,CAAP;;;AAGF,SAASuI,aAAT,CAAuBvI,IAAvB,EAA6B;MACnBlT,IADmB,GACCkT,IADD,CACnBlT,IADmB;MACb0b,SADa,GACCxI,IADD,CACbwI,SADa;;;;;SAKpBN,oBAAYlI,IAAZ,IAAkBmI,gBAAgBK,UAAU1b,IAAV,CAAlC,QACLsa,iBAAiBta,IAAjB,EAAuBkT,IAAvB,CADF;;;AAIF,IAAMyI,gBAAgB;SAAA,qBACwB;QAApCD,SAAoC,yDAAxBpB,gBAAwB;QAANpH,IAAM;gBACFA,IADE;QAClC0I,WADkC,SAClCA,WADkC;QACrBC,cADqB,SACrBA,cADqB;;;QAGtCH,UAAUd,MAAV,KAAqB,GAAzB,EAA8B,OAAOc,UAAUnB,OAAV,CAAkBrH,IAAlB,CAAP;;wBAGzBA,IADL;;;;QAKI0I,WAAJ,EAAiB;UACT/a,WAAU4a,2BACXvI,IADW,IACLlT,MAAM,SADD,EACYsb,aAAa,IADzB,EAC+BhO,OAAOuO;SADtD;aAGO;;OAAP;;QAIIvO,QAAQmO,2BAAmBvI,IAAnB,IAAyBlT,MAAM,OAA/B,IAAd;QACM2U,gBAAgB8G,2BAAmBvI,IAAnB,IAAyBlT,MAAM,eAA/B,IAAtB;QACMwQ,SAASiL,2BAAmBvI,IAAnB,IAAyBlT,MAAM,QAA/B,IAAf;QACMya,cAAcgB,2BAAmBvI,IAAnB,IAAyBlT,MAAM,aAA/B,IAApB;QACMa,UAAU4a,2BACXvI,IADW,IACLlT,MAAM,SADD,EACYsb,aAAa,IADzB,EAC+BhO;OAD/C;QAGMmD,eAAegL,2BAAmBvI,IAAnB,IAAyBlT,MAAM,cAA/B,EAA+Ca,gBAA/C,IAArB;QACMgQ,MAAM4K,2BAAmBvI,IAAnB,IAAyBlT,MAAM,KAA/B,EAAsCa,gBAAtC,IAAZ;QACMsZ,UAAUsB,2BAAmBvI,IAAnB,IAAyBlT,MAAM,SAA/B,EAA0Ca,gBAA1C,IAAhB;QACM6Z,YAAYe,2BAAmBvI,IAAnB,IAAyBlT,MAAM,WAA/B,EAA4Ca,gBAA5C,IAAlB;;yBACwB4a,2BAAmBvI,IAAnB,IAAyBlT,MAAM,cAA/B,IA7BkB;;QA6BlC3B,GA7BkC,kBA6BlCA,GA7BkC;QA6B7Buc,MA7B6B,kBA6B7BA,MA7B6B;;;WA+BnC;kBAAA;sBAAA;oBAAA;kCAAA;gCAAA;cAAA;8BAAA;cAAA;oBAAA;sBAAA;;KAAP;;CAhCJ,CAgDA;;ACjJA;sDAAe;QAEXH,WAFW,SAEXA,WAFW;QAGXrL,IAHW,SAGXA,IAHW;QAIXzQ,CAJW,SAIXA,CAJW;QAKXmV,SALW,SAKXA,SALW;QAMXlT,MANW,SAMXA,MANW;QAOXkb,SAPW,SAOXA,SAPW;QAQXxO,KARW,SAQXA,KARW;QASXjP,GATW,SASXA,GATW;;;;;;;iBAAA,GAaD,CAbC;wBAAA,GAcQ,CAACgM,aAAahM,GAAb,CAAD,CAdR;;;;;;kBAkBNoc,eAAesB,QAAQ,EAlBjB;;;;;qBAmBF,CAAT;;mBACUvb,SAASwb,MAAT,CAAgBvB,WAAhB,CApBC;;;aAAA;;mBAqBJ9b,EAAEyQ,IAAF,EAAP;;yBArBW,GAuBW;mBACfqL,WADe;wBAAA;kBAAA;kCAAA;2BAKP,IALO;8BAMJnN,KANI;;aAvBX;0BAAA,GAiCYqO,cAAcpB,OAAd,CAAsBuB,SAAtB,EAAiCG,aAAjC,CAjCZ;;;yBAmCE3Q,IAAb,CAAkBmP,WAAlB;kCAEK7Z,MADL;sCAGMA,OAAOC,OADX,yCAGakb,KAHb,uBAIIG,eAAerb,OAJnB;;;0BAQYqb,eAAezB,WAA7B;;;;;0DAIG7Z,MAlDQ;0BAmDCmb,KAnDD;6BAoDIA;;;;;;;;;GApDnB;;WAA8BI,eAA9B;;;;SAA8BA,eAA9B;;;ACAA,IAAMC,OAAO;OAAA,iBACC/d,GADD,EACM+Q,IADN,EACuB;;;QAAX8D,IAAW,yDAAJ,EAAI;;;;;;;;qBACCA,QAAQ,IADT;wCACxBmJ,aADwB;2BAAA,sCACR,IADQ;uBAAA,GAEdxB,aAAaxc,GAAb,CAFc;;;;qBAKhBmC,SAASwb,MAAT,CAAgB3d,GAAhB,EAAqB+Q,IAArB,CALgB;;;eAAA;;qBAMzBzQ,EAAEyQ,IAAF,EAAP;;;;uBANgC,GAUdzQ,EAAE,MAAF,EAAUqD,GAAV,CAAc,UAACjD,CAAD,EAAIC,IAAJ;uBAAaL,EAAEK,IAAF,EAAQG,IAAR,CAAa,MAAb,CAAb;eAAd,EAAiDwP,OAAjD,EAVc;oBAAA,GAYnBgN,cAAcpB,OAAd,CAAsBuB,SAAtB,EAAiC,EAAEzd,QAAF,EAAO+Q,UAAP,EAAazQ,IAAb,EAAgBmV,oBAAhB,EAAjC,CAZmB;wBAaDlT,MAbC;mBAAA,WAaxB0M,KAbwB;yBAAA,WAajBmN,WAbiB;;;;oBAgB5B4B,iBAAiB5B,WAhBW;;;;;;qBAiBf0B,gBACb;oCAAA;wCAAA;0BAAA;oBAAA;oCAAA;8BAAA;4BAAA;;eADa,CAjBe;;;oBAAA;;;;;oCA+BzBvb,MADL;4BAEc,CAFd;+BAGiB;;;;+CAIZA,MArCyB;;;;;;;;;;CADpC,CA2CA;;"}
\ No newline at end of file
+{"version":3,"file":null,"sources":["../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/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/utils/range.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/constants.js","../src/extractors/root-extractor.js","../src/extractors/collect-all-pages.js","../src/iris.js"],"sourcesContent":["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';\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 ${MAX_CONTENT_LENGTH}.`\n );\n }\n\n return true;\n}\n\n// Grabs the last two pieces of the URL and joins them back together\n// This is to get the 'livejournal.com' from 'erotictrains.livejournal.com'\nexport function baseDomain({ host }) {\n return host.split('.').slice(-2).join('.');\n}\n\n// Set our response attribute to the result of fetching our URL.\n// TODO: This should gracefully handle timeouts and raise the\n// proper exceptions on the many failure cases of HTTP.\n// TODO: Ensure we are not fetching something enormous. Always return\n// unicode content for HTML, with charset conversion.\n\nexport default async function fetchResource(url) {\n const parsedUrl = URL.parse(url);\n\n const options = {\n url: parsedUrl,\n headers: { ...REQUEST_HEADERS },\n timeout: FETCH_TIMEOUT,\n // Don't set encoding; this fixes issues\n // w/gzipped responses\n encoding: null,\n // Accept cookies\n jar: true,\n };\n\n const { response, body } = await get(options);\n\n try {\n validateResponse(response);\n return { body, response };\n } catch (e) {\n return e;\n }\n}\n","function convertMetaProp($, 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) {\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);\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// def __init__(self, url, parse_non_2xx=False, response=None):\n// \"\"\" Create a Resource.\n//\n// :param url: The URL for the document we should retrieve.\n// :param parse_non_2xx: If True, attempt to parse non-200 level\n// resources. If False, raise a RetrievalFailed\n// based exception. Default is False.\n// :param response: If not None, use as the response rather than\n// attempting to fetch it ourselves. Expects a\n// requests.models.Response object.\n// \"\"\"\n// self.url = url\n// self.parse_non_2xx = parse_non_2xx\n//\n// if response:\n// self.response = response\n// else:\n// self.response = self._fetch_resource()\n\n// Iris: Human-friendly content extraction.\n\n// import logging\n// import lxml\n// import re\n// import requests\n// import socket\n//\n// from django.conf import settings\n// from lxml.etree import XPathEvaluator\n// from lxml.html.clean import Cleaner\n// from urlparse import urlparse\n//\n// from utils.dom import extract_by_selector as ebs, convert_lazy_loaded_images\n// from utils.dom.attribmap import AttribMap\n// from utils.statsd import stats\n// from utils.text import is_text\n// from utils.html import get_charset_from_html, strip_content_encodings\n//\n// from . import exceptions\n//\n// logger = logging.getLogger(__name__)\n//\n// # Hosts that are allowed to use embeds and iframes. We should be very\n// # restrictive with this and only include top-tier video sites.\n// host_whitelist = ['www.youtube.com', 'www.vimeo.com']\n//\n// # The number of seconds to attempt to fetch a resource before timing out.\n// FETCH_TIMEOUT = 10\n//\n// cleaner = Cleaner(\n// style=True,\n// page_structure=False,\n// meta=False,\n// add_nofollow=False, # done by hand\n// remove_unknown_tags=False,\n// links=False,\n// host_whitelist=host_whitelist)\n//\n//\n//\n// class Resource(object):\n// \"\"\" A Resource is a wrapper class for an HTTP resource. Provides\n// functionality to fetch a resource as well as a handful of shortcut\n// methods to run xpath efficiently on HTML, etc.\n//\n// Uses requests and lxml internally for fetching and querying.\n// \"\"\"\n//\n//\n// def __init__(self, url, parse_non_2xx=False, response=None):\n// \"\"\" Create a Resource.\n//\n// :param url: The URL for the document we should retrieve.\n// :param parse_non_2xx: If True, attempt to parse non-200 level\n// resources. If False, raise a RetrievalFailed\n// based exception. Default is False.\n// :param response: If not None, use as the response rather than\n// attempting to fetch it ourselves. Expects a\n// requests.models.Response object.\n// \"\"\"\n// self.url = url\n// self.parse_non_2xx = parse_non_2xx\n//\n// if response:\n// self.response = response\n// else:\n// self.response = self._fetch_resource()\n//\n// def __unicode__(self):\n// return u'<Resource ({0})>'.format(self.url)\n//\n// def __repr__(self):\n// return \"<Resource ({0})>\".format(self.url)\n//\n// @classmethod\n// def fabricate(kls, url, content, headers=None):\n// \"\"\" Given a URL and some content, create a fake Resource that looks\n// as though it has already fetched the content. Useful for using\n// Resource objects without having to do a GET.\n// \"\"\"\n//\n// if type(content) != unicode:\n// raise TypeError(\"Provided content must be unicode.\")\n//\n// if headers is None:\n// headers = {}\n//\n// try:\n// utf8_content = content.encode('utf-8', 'strict')\n// except UnicodeDecodeError:\n// logger.warning(\"Unable to encode content for url %s. Content \"\n// \"should be unicode and encodeable at this point.\")\n// utf8_content = content.encode('utf-8', 'replace')\n//\n// mocked_response_dict = {\n// \"cookies\": {},\n// \"_content\": utf8_content,\n// \"headers\": dict({\n// \"content-length\": len(content),\n// \"accept-ranges\": \"bytes\",\n// \"vary\": \"Accept-Encoding,Cookie\",\n// \"server\": \"Apache/2.2.21\",\n// \"content-type\": \"text/html; charset=UTF-8\"\n// }, **headers),\n// \"url\": url,\n// \"status_code\": 200,\n// \"_content_consumed\": False,\n// \"request\": None,\n// \"raw\": None,\n// \"error\": None,\n// \"config\": {\n// \"decode_unicode\": True,\n// \"pool_connections\": 10,\n// \"verbose\": None,\n// \"keep_alive\": True,\n// \"max_retries\": 0,\n// \"base_headers\": {\n// \"Accept-Encoding\": \"identity, deflate, compress, gzip\",\n// \"Accept\": \"|)}>#*\",\n// \"User-Agent\": \"python-requests/0.8.1\"\n// },\n// \"pool_maxsize\": 10,\n// \"safe_mode\": False,\n// \"max_redirects\": 30\n// },\n// \"history\": []\n// }\n// mocked_response = requests.Response()\n// for k, v in mocked_response_dict.items():\n// setattr(mocked_response, k, v)\n//\n// return Resource(\n// url = url,\n// response = mocked_response\n// )\n//\n//\n// @property\n// def url(self):\n// return self._url\n//\n//\n// @url.setter\n// def url(self, value):\n// parsed_url = urlparse(value)\n// if parsed_url.scheme not in ('http', 'https'):\n// raise ValueError(\"Resource only allows HTTP and HTTPS urls.\")\n//\n// if not parsed_url.netloc:\n// raise ValueError(\"Relative URLs are not allowed.\")\n//\n// self._url = value\n//\n// _parsed_url = None\n// @property\n// def parsed_url(self):\n// if self._parsed_url is None:\n// self._parsed_url = urlparse(self.url)\n// return self._parsed_url\n//\n// @property\n// def status_code(self):\n// return self.response.status_code\n//\n//\n// _content = None\n// @property\n// def content(self):\n// \"\"\"Return the content for a resource. Always returns unicode.\n//\n// \"\"\"\n// if self._content is None:\n// # Requests that come in without content-type encoding headers will\n// # default to iso-8859-1, which could be wrong\n// if (self.response.encoding and\n// self.response.encoding.lower() == 'iso-8859-1'):\n// # Dont send unicode, because it could have been decoded wrong\n// # by an incorrect content-type guess.\n// encoding = get_charset_from_html(self.response.content) or 'iso-8859-1'\n//\n// if encoding != self.response.encoding:\n// # First, try to use the encoding we found in the markup\n// try:\n// self._content = self.response.content.decode(encoding)\n// except (LookupError, UnicodeDecodeError):\n// stats.increment(\n// 'iris.resource.encoding.encoding_mismatch')\n// # That encoding might be wrong though, so if it is, use\n// # the one it reported since they could have the wrong\n// # one set in the markup. eg. sending the content over\n// # as iso but declaring it to be utf-8 like gq.com does.\n// # We may also end up with an invalid encoding type, at\n// # which point we should also just use the request\n// # encoding and replace silently.\n// self._content = self.response.content.decode(\n// self.response.encoding, 'replace')\n// else:\n// # If the encoding guess was right, just use the unicode\n// self._content = self.response.text\n//\n// else:\n// # Otherwise we trust the encoding\n// self._content = self.response.text\n//\n// return self._content\n//\n//\n// @property\n// def content_type(self):\n// return self.response.headers.get('content-type', '')\n//\n//\n// @property\n// def is_html(self):\n// if 'html' in self.content_type:\n// return True\n//\n// # Otherwise, just try parsing it and see if it succeeds\n// try:\n// return (self.doc is not None)\n// except:\n// return False\n//\n// @property\n// def is_plaintext(self):\n// if 'text/plain' in self.content_type:\n// return True\n//\n// return False\n//\n// @property\n// def is_image(self):\n// if 'image' in self.content_type:\n// return True\n//\n// return False\n//\n// @property\n// def is_pdf(self):\n// if 'pdf' in self.content_type:\n// return True\n//\n// return False\n//\n// _lxml_doc = None\n// @property\n// def doc(self):\n// if self._lxml_doc is None:\n// self._generate_lxml_doc()\n//\n// return self._lxml_doc\n//\n// _docxp = None\n// @property\n// def docxp(self):\n// \"\"\" Generate an XPath Evaluator for this doc. \"\"\"\n// if self._docxp is None:\n// self._docxp = XPathEvaluator(self.doc)\n//\n// return self._docxp\n//\n// _redocxp = None\n// @property\n// def redocxp(self):\n// \"\"\" Generate an XPath Evaluator for this doc, that includes the RE\n// namespace for regular expression matching.\n//\n// \"\"\"\n// if self._redocxp is None:\n// _rens = {'re':'http://exslt.org/regular-expressions'}\n// self._redocxp = XPathEvaluator(self.doc, namespaces=_rens)\n//\n// return self._redocxp\n//\n// def _generate_lxml_doc(self):\n// # First check if we have a text based resource\n// if (not 'html' in self.content_type and\n// not 'text' in self.content_type and\n// not is_text(self.content[:512])):\n// raise ValueError(\"Content does not appear to be text.\")\n//\n//\n// # Remove useless carriage returns which get parsed as
otherwise\n// content = re.sub(r'(\\n\\r|\\r\\n)', '\\n', self.content)\n//\n// # Dont pass any content encodings into lxml, it is dumb about them\n// content = strip_content_encodings(content)\n//\n// self._lxml_doc = lxml.html.fromstring(content)\n//\n//\n//\n//\n// if len(self._lxml_doc.getchildren()) == 0:\n// stats.increment('iris.resource.encoding.no_children')\n// raise ValueError(\"No children, likely a bad parse.\")\n//\n//\n// # Sometimes, lxml (or BeautifulSoup) will wrap the whole document\n// # in an extra html tag. This screws up a whole bunch of things in\n// # the parsing process. If this is the case, reset the doc to the\n// # ACTUAL root of the doc.\n// # Sample cases:\n// # * Strange Doctype causing issues: http://bit.ly/IATz0B\n// # * Messy markup causing double HTML tags: http://bit.ly/IGOq4o\n// # Also check for a body inside of our internal HTML tag, to determine\n// # that it's not just a junk HTML tag sibling at the bottom of the\n// # doc or something.\n// internal_html_tag = self._lxml_doc.find('html')\n// if (internal_html_tag is not None and\n// len(internal_html_tag.xpath('.//body')) > 0):\n// self._lxml_doc = internal_html_tag\n//\n// self._normalize_meta_tags()\n//\n// self._lxml_doc.make_links_absolute(self.url)\n//\n// # Convert any lazy loaded images into normal images before clean_html\n// # which will strip all other attributes\n// self._lxml_doc = convert_lazy_loaded_images(self._lxml_doc)\n//\n// # Clean the doc of anything malicious.\n// self._lxml_doc = cleaner.clean_html(self._lxml_doc)\n//\n// # Manually nofollow links so that we don't clobber rel author\n// # Workaround for https://bugs.launchpad.net/lxml/+bug/971754\n// for a in self.docxp('//a'):\n// if a.attrib.get('rel', None):\n// rel_attribs = set(a.attrib['rel'].split())\n// rel_attribs.add('nofollow')\n// a.attrib['rel'] = ' '.join(rel_attribs)\n// else:\n// a.attrib['rel'] = 'nofollow'\n//\n// # Re-relativize anchor links\n// anchor_link_xpath = (\"//a[starts-with(@href, '%s#')]\" %\n// self.url.replace(\"'\", \"%27\"))\n// for link in self.docxp(anchor_link_xpath):\n// link.attrib['href'] = link.attrib['href'].replace(self.url, '')\n//\n//\n// _attrib_map = None\n// @property\n// def attrib_map(self):\n// \"\"\" Create an AttribMap object for fast checking of class/id existence\n// in the document. Used in association with extract_by_selector.\n//\n// \"\"\"\n// if self._attrib_map is None:\n// self._attrib_map = AttribMap(self.doc)\n//\n// return self._attrib_map\n//\n//\n// def extract_by_selector(self, selector):\n// \" Shortcut to run extract_by_selector on our doc with our AttribMap. \"\n// return ebs(self.doc, selector, self.attrib_map, self.docxp)\n//\n//\n","const NYMagExtractor = {\n domain: 'nymag.com',\n content: {\n // Order by most likely. Extractor will stop on first occurence\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 && $children.get(0).tagName === 'img') {\n return 'figure';\n }\n\n return null;\n },\n },\n },\n\n title: {\n selectors: [\n 'h1.lede-feature-title',\n 'h1.headline-primary',\n 'h1',\n ],\n },\n\n author: {\n selectors: [\n '.by-authors',\n '.lede-feature-author',\n ],\n },\n\n dek: {\n selectors: [\n '.lede-feature-teaser',\n ],\n },\n\n datePublished: {\n selectors: [\n 'time.article-timestamp[datetime]',\n 'time.article-timestamp',\n ],\n },\n};\n\nexport default NYMagExtractor;\n","const BloggerExtractor = {\n domain: 'blogspot.com',\n content: {\n // Blogger is insane and does not load its content\n // initially in the page, but it's all there\n // in noscript\n selectors: [\n '.post-content noscript',\n ],\n\n // Selectors to remove from the extracted content\n clean: [\n ],\n\n // Convert the noscript tag to a div\n transforms: {\n noscript: 'div',\n },\n },\n\n author: {\n selectors: [\n '.post-author-name',\n ],\n },\n\n title: {\n selectors: [\n 'h2.title',\n ],\n },\n\n datePublished: {\n selectors: [\n 'span.publishdate',\n ],\n },\n};\n\nexport default BloggerExtractor;\n","const WikipediaExtractor = {\n domain: 'wikipedia.org',\n content: {\n selectors: [\n '#mw-content-text',\n ],\n\n // transform top infobox to an image with caption\n transforms: {\n '.infobox img': ($node) => {\n $node.parents('.infobox').prepend($node);\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 ],\n\n },\n\n author: 'Wikipedia Contributors',\n\n title: {\n selectors: [\n 'h2.title',\n ],\n },\n\n datePublished: {\n selectors: [\n '#footer-info-lastmod',\n ],\n },\n\n};\n\nexport default WikipediaExtractor;\n","import NYMagExtractor from './custom/nymag.com';\nimport BloggerExtractor from './custom/blogspot.com';\nimport WikipediaExtractor from './custom/wikipedia.org';\n\nconst Extractors = {\n 'nymag.com': NYMagExtractor,\n 'blogspot.com': BloggerExtractor,\n 'wikipedia.org': WikipediaExtractor,\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 => `[${selector}]`);\nexport const REMOVE_ATTR_LIST = REMOVE_ATTRS.join(',');\nexport const WHITELIST_ATTRS = ['src', 'href', 'class', 'id'];\nexport const WHITELIST_ATTRS_RE = new RegExp(`^(${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'].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',\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// 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\n// ## NOTES:\n// This is a working first pass, but if/when we start optimizing\n// this is a good candidate. - AP\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 = $div.children(DIV_TO_P_BLOCK_TAGS).length === 0;\n // .not(DIV_TO_P_BLOCK_TAGS).length === 0;\n if (convertable) {\n convertNodeTo($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 { 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, $) {\n $(STRIP_OUTPUT_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 = $.find('h1')\n const $hOnes = $('h1', article);\n if ($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 * 0.2);\n const wrappingDiv = $('<div></div>');\n\n $candidate.parent().children().each((index, child) => {\n const $child = $(child);\n // Ignore tags like BR, HR, etc\n if (NON_TOP_CANDIDATE_TAGS_RE.test(child.tagName)) {\n return null;\n }\n\n const childScore = getScore($child);\n if (childScore) {\n if ($child === $candidate) {\n wrappingDiv.append($child);\n } else {\n let contentBonus = 0;\n // extract to scoreLinkDensity() TODO\n const density = linkDensity($child);\n\n // If sibling has a very low link density,\n // give it a small bonus\n if (density < 0.05) {\n contentBonus += 20;\n }\n\n // If sibling has a high link density,\n // give it a penalty\n if (density >= 0.5) {\n contentBonus -= 20;\n }\n\n // If sibling node has the same class as\n // candidate, give it a bonus\n if ($child.attr('class') === $candidate.attr('class')) {\n contentBonus += topScore * 0.2;\n }\n\n const newScore = getScore($child) + contentBonus;\n\n if (newScore >= siblingScoreThreshold) {\n return wrappingDiv.append($child);\n } else if (child.tagName === 'p') {\n const childContent = $child.text();\n const childContentLength = textLength(childContent);\n\n if (childContentLength > 80 && density < 0.25) {\n return wrappingDiv.append($child);\n } else if (childContentLength <= 80 && density === 0 &&\n hasSentenceEnd(childContent)) {\n return wrappingDiv.append($child);\n }\n }\n }\n }\n\n return null;\n });\n\n return wrappingDiv;\n}\n","import { NON_TOP_CANDIDATE_TAGS_RE } from './constants';\nimport { getScore } from './index';\nimport mergeSiblings from './merge-siblings';\n\n// After we've calculated scores, loop through all of the possible\n// candidate nodes we found and find the one with the highest score.\nexport default function findTopCandidate($) {\n let $candidate;\n let topScore = 0;\n\n $('[score]').each((index, node) => {\n const $node = $(node);\n const score = getScore($node);\n // Ignore tags like BR, HR, etc\n if (NON_TOP_CANDIDATE_TAGS_RE.test(node.tagName)) {\n return;\n }\n\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, $content));\n\n // console.log($content.html())\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 $(IS_WP_SELECTOR).length > 0;\n}\n","// CLEAN AUTHOR CONSTANTS\nexport const CLEAN_AUTHOR_RE = /^\\s*(posted |written )?by\\s*:?\\s*(.*)/i;\n // author = re.sub(r'^\\s*(posted |written )?by\\s*:?\\s*(.*)(?i)',\n\n// CLEAN DEK CONSTANTS\nexport const TEXT_LINK_RE = new RegExp('http(s)?://', 'i');\n// An ordered list of meta tag names that denote likely article deks.\n// From most distinct to least distinct.\n//\n// NOTE: There are currently no meta tags that seem to provide the right\n// content consistenty enough. Two options were:\n// - og:description\n// - dc.description\n// However, these tags often have SEO-specific junk in them that's not\n// header-worthy like a dek is. Excerpt material at best.\nexport const DEK_META_TAGS = [\n];\n\n// An ordered list of Selectors to find likely article deks. From\n// most explicit to least explicit.\n//\n// Should be more restrictive than not, as a failed dek can be pretty\n// detrimental to the aesthetics of an article.\nexport const DEK_SELECTORS = [\n '.entry-summary',\n];\n\n// CLEAN DATE PUBLISHED CONSTANTS\nexport const CLEAN_DATE_STRING_RE = /^\\s*published\\s*:?\\s*(.*)/i;\nexport const TIME_MERIDIAN_SPACE_RE = /(.*\\d)(am|pm)(.*)/i;\nexport const TIME_MERIDIAN_DOTS_RE = /\\.m\\./i;\nconst months = [\n 'jan',\n 'feb',\n 'mar',\n 'apr',\n 'may',\n 'jun',\n 'jul',\n 'aug',\n 'sep',\n 'oct',\n 'nov',\n 'dec',\n];\nconst allMonths = months.join('|');\nconst timestamp1 = '[0-9]{1,2}:[0-9]{2,2}( ?[ap].?m.?)?';\nconst timestamp2 = '[0-9]{1,2}[/-][0-9]{1,2}[/-][0-9]{2,4}';\nexport const SPLIT_DATE_STRING =\n new RegExp(`(${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 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 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 }\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 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 style or align 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 cleanTags(article, $, cleanConditionally);\n\n // Remove empty paragraph nodes\n removeEmpty(article, $);\n\n return article;\n}\n // headers = doc.xpath('.//h2 | .//h3 | .//h4 | .//h5 | .//h6')\n // for header in headers:\n // drop_header = False\n //\n // # Remove any headers that are before any 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 // no_previous_ps = int(header.xpath(\"count(preceding::p[1])\")) == 0\n // if no_previous_ps:\n // similar_header_count = int(doc.xpath('count(.//%s)' % header.tag))\n // if similar_header_count < 3:\n // drop_header = True\n //\n // # Remove any headers that match the title exactly.\n // if inner_text(header) == self.title:\n // drop_header = True\n //\n // # If this header has a negative weight, it's probably junk.\n // # Get rid of it.\n // if self.get_weight(header) < 0:\n // drop_header = True\n //\n // if drop_header:\n // try:\n // header.drop_tree()\n // except AssertionError:\n // # No parent exists for this node, so just blank it out.\n // header.text = ''\n //\n // if clean_conditionally:\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 // self._clean_conditionally(doc, ['ul', 'ol', 'table', 'div'])\n //\n // return doc\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 leadImageUrl: cleanImage,\n dek: cleanDek,\n datePublished: 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|>([^|]|$)|»([^|]|$))', 'i');\n\n// Match any link text/classname/id that looks like it is an end link: things\n// like \"first\", \"last\", \"end\", etc.\nexport const CAP_LINK_TEXT_RE = new RegExp('(first|last|end)', 'i');\n\n// Match any link text/classname/id that looks like it means the previous\n// page.\nexport const PREV_LINK_TEXT_RE = new RegExp('(prev|earl|old|new|<|«)', 'i');\n\n// Match any phrase that looks like it could be page, or paging, or pagination\nexport const PAGE_RE = new RegExp('pag(e|ing|inat)', 'i');\n\n","import { EXTRANEOUS_LINK_HINTS_RE } from '../constants';\n\nexport default function scoreExtraneousLinks(href) {\n // If the URL itself contains extraneous values, give a penalty.\n if (EXTRANEOUS_LINK_HINTS_RE.test(href)) {\n return -25;\n }\n\n return 0;\n}\n","export default function* range(start = 1, end = 1) {\n while (start <= end) {\n yield start += 1;\n }\n}\n","import { range } from 'utils';\nimport {\n NEGATIVE_SCORE_RE,\n POSITIVE_SCORE_RE,\n PAGE_RE,\n} from 'utils/dom/constants';\nimport { EXTRANEOUS_LINK_HINTS_RE } from '../constants';\n\nfunction makeSig($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';\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\n extract(options) {\n const { html } = options;\n\n if (html) {\n const $ = cheerio.load(html);\n options.$ = $;\n }\n\n const title = this.title(options);\n const date_published = this.date_published(options);\n const author = this.author(options);\n const content = this.content({ ...options, title });\n const lead_image_url = this.lead_image_url({ ...options, content });\n const dek = this.dek({ ...options, content });\n const next_page_url = this.next_page_url(options);\n const excerpt = this.excerpt({ ...options, content });\n const word_count = this.word_count({ ...options, content });\n const { url, domain } = this.url_and_domain(options);\n\n return {\n title,\n author,\n date_published: date_published || null,\n dek,\n lead_image_url,\n content,\n next_page_url,\n url,\n domain,\n excerpt,\n word_count,\n };\n },\n};\n\nexport default GenericExtractor;\n","import URL from 'url';\n\nimport Extractors from './all';\nimport GenericExtractor from './generic';\n\nexport default function getExtractor(url) {\n const parsedUrl = URL.parse(url);\n const { hostname } = parsedUrl;\n const baseDomain = hostname.split('.').slice(-2).join('.');\n\n return Extractors[hostname] || Extractors[baseDomain] || GenericExtractor;\n}\n","export const ATTR_RE = /\\[([\\w-]+)\\]/;\n","import 'babel-polyfill';\n\nimport Cleaners from 'cleaners';\nimport { convertNodeTo } from 'utils/dom';\nimport GenericExtractor from './generic';\nimport { ATTR_RE } from './constants';\n\n// Remove elements by an array of selectors\nexport function cleanBySelectors($content, $, { clean }) {\n if (!clean) return null;\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 null;\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\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 } = extractionOpts;\n\n const matchingSelector = selectors.find(selector => $(selector).length === 1 && $(selector).text().trim() !== '');\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);\n\n return $.html($content);\n }\n // if selector includes an attr (e.g., img[src]),\n // extract the attr\n const attr = matchingSelector.match(ATTR_RE);\n let result;\n\n if (attr) {\n result = $(matchingSelector).attr(attr[1]);\n } else {\n // otherwise use the text of the node\n result = $(matchingSelector).text();\n }\n return Cleaners[type](result, opts);\n}\n\nfunction extractResult(opts) {\n const { type, extractor } = opts;\n\n // If nothing matches the selector,\n // run the Generic extraction\n return select({ ...opts, extractionOpts: extractor[type] }) ||\n GenericExtractor[type](opts);\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 { url, domain } = extractResult({ ...opts, type: 'url_and_domain' });\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 };\n },\n};\n\nexport default RootExtractor;\n","import 'babel-polyfill';\nimport { removeAnchor } from 'utils/text';\nimport RootExtractor from 'extractors/root-extractor';\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 return {\n ...result,\n total_pages: pages,\n pages_rendered: pages,\n };\n}\n","import Resource from 'resource';\nimport getExtractor from 'extractors/get-extractor';\nimport RootExtractor from 'extractors/root-extractor';\nimport collectAllPages from 'extractors/collect-all-pages';\n\nconst Iris = {\n async parse(url, html, opts = {}) {\n const { fetchAllPages = true } = opts || true;\n const Extractor = getExtractor(url);\n // console.log(`Using extractor for ${Extractor.domain}`);\n\n const $ = await Resource.create(url, html);\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 });\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 $,\n metaCache,\n result,\n title,\n url,\n }\n );\n } else {\n result = {\n ...result,\n total_pages: 1,\n rendered_pages: 1,\n };\n }\n\n return result;\n },\n\n};\n\nexport default Iris;\n"],"names":["REQUEST_HEADERS","FETCH_TIMEOUT","BAD_CONTENT_TYPES","BAD_CONTENT_TYPES_RE","RegExp","join","MAX_CONTENT_LENGTH","get","options","Promise","resolve","reject","err","response","body","validateResponse","parseNon2xx","statusMessage","statusCode","Error","error","headers","contentType","contentLength","test","url","URL","parse","parsedUrl","fetchResource","convertMetaProp","$","from","to","each","_","node","$node","value","attr","removeAttr","normalizeMetaTags","IS_LINK","IS_IMAGE","TAGS_TO_REMOVE","convertLazyLoadedImages","img","ownKeys","attribs","forEach","isComment","index","type","cleanComments","root","find","contents","filter","remove","clean","Resource","preparedResponse","validResponse","generateDoc","result","content","includes","cheerio","load","normalizeWhitespace","children","length","NYMagExtractor","$children","tagName","BloggerExtractor","WikipediaExtractor","parents","prepend","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","replaceWith","convertDivs","div","$div","convertable","convertSpans","span","$span","convertToParagraphs","convertNodeTo","attribString","Reflect","key","cleanForHeight","$img","height","parseInt","width","removeSpacers","cleanImages","$article","stripJunkTags","article","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","$parent","rawScore","scoreContent","parentSelector","childSelector","NORMALIZE_RE","normalizeSpaces","replace","extractFromUrl","regexList","matchRe","re","exec","PAGE_IN_HREF_RE","HAS_ALPHA_RE","IS_ALPHA_RE","IS_DIGIT_RE","pageNumFromUrl","matches","pageNum","removeAnchor","split","isGoodSegment","segment","firstSegmentHasLetters","goodSegment","toLowerCase","articleBaseUrl","parsed","protocol","host","path","cleanedSegments","reverse","rawSegment","possibleSegment","fileExt","push","SENTENCE_END_RE","hasSentenceEnd","mergeSiblings","$candidate","topScore","siblingScoreThreshold","wrappingDiv","child","$child","childScore","append","contentBonus","density","linkDensity","newScore","childContent","childContentLength","findTopCandidate","first","removeUnlessContent","weight","hasClass","pCount","inputCount","imgCount","nodeIsList","previousNode","prev","scriptCount","cleanTags","cleanHeaders","title","header","$header","prevAll","rewriteTopLevel","absolutize","rootUrl","$content","absoluteUrl","makeLinksAbsolute","totalTextLength","linkText","linkLength","extractFromMeta","metaNames","cachedNames","foundNames","indexOf","name","nodes","values","toArray","metaValue","stripTags","isGoodNode","maxChildren","withinComment","extractFromSelectors","selectors","textOnly","html","cleanText","commentParent","class","undefined","nodeIsSufficient","isWordpress","CLEAN_AUTHOR_RE","TEXT_LINK_RE","CLEAN_DATE_STRING_RE","TIME_MERIDIAN_SPACE_RE","TIME_MERIDIAN_DOTS_RE","months","allMonths","timestamp1","timestamp2","SPLIT_DATE_STRING","TITLE_SPLITTERS_RE","DOMAIN_ENDINGS_RE","cleanAuthor","author","leadImageUrl","validUrl","isWebUri","cleanDek","dek","dekText","cleanDateString","dateString","cleanDatePublished","date","moment","Date","isValid","toISOString","extractCleanNode","cleanConditionally","cleanTitle","resolveSplitTitle","h1","extractBreadcrumbTitle","splitTitle","termCounts","titleText","maxTerm","termCount","splitEnds","longestEnd","end","cleanDomainFromTitle","nakedDomain","startSlug","startSlugRatio","wuzzy","levenshtein","endSlug","endSlugRatio","newTitle","Cleaners","cleanImage","cleanContent","extractBestNode","opts","$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","$sibling","scoreByDimensions","src","area","round","scoreByPosition","$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","range","start","makeSig","$link","positiveMatch","negativeMatch","parentData","scorePrevLink","linkData","shouldScore","baseUrl","previousUrls","hostname","linkHost","fragment","scoreBaseUrl","baseRegex","scoreNextLinkText","scoreCapLinks","makeBaseRegex","scoreLinks","links","scoredPages","possiblePages","link","possiblePage","GenericNextPageUrlExtractor","scoredLinks","topPage","scoredLink","CANONICAL_META_SELECTORS","parseDomain","GenericUrlExtractor","$canonical","metaUrl","EXCERPT_META_SELECTORS","maxLength","ellipsize","ellipse","GenericExcerptExtractor","excerpt","shortContent","GenericWordCountExtractor","GenericExtractor","extract","bind","date_published","lead_image_url","next_page_url","word_count","url_and_domain","domain","getExtractor","baseDomain","ATTR_RE","cleanBySelectors","transformElements","transforms","$matches","select","extractionOpts","extractHtml","matchingSelector","wrap","extractResult","extractor","RootExtractor","contentOnly","extractedTitle","Extractor","pages","create","extractorOpts","nextPageResult","collectAllPages","Iris","fetchAllPages"],"mappings":";;;;;;;;;;;;;;AAAO,IAAMA,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;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;ACnBA,SAASC,GAAT,CAAaC,OAAb,EAAsB;SACb,IAAIC,OAAJ,CAAY,UAACC,OAAD,EAAUC,MAAV,EAAqB;YAC9BH,OAAR,EAAiB,UAACI,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,yDAAP,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;;;;MAoB1DpB,qBAAqBqB,IAArB,CAA0BF,WAA1B,CAAJ,EAA4C;UACpC,IAAIH,KAAJ,yCACkCG,WADlC,0BAAN;;;;MAMEC,gBAAgBjB,kBAApB,EAAwC;UAChC,IAAIa,KAAJ,yEACkEb,kBADlE,OAAN;;;SAKK,IAAP;;;AAGF,AAMA;;;;;;AAMA;uDAAe,iBAA6BmB,GAA7B;;;;;;;qBAAA,GACKC,IAAIC,KAAJ,CAAUF,GAAV,CADL;mBAAA,GAGG;mBACTG,SADS;oCAEA5B,eAAd,CAFc;uBAGLC,aAHK;;;wBAMJ,IANI;;mBAQT;aAXM;;mBAcoBM,IAAIC,OAAJ,CAdpB;;;;oBAAA,SAcLK,QAdK;gBAAA,SAcKC,IAdL;;;6BAiBMD,QAAjB;6CACO,EAAEC,UAAF,EAAQD,kBAAR,EAlBI;;;;;;;;;;;;;GAAf;;WAA8BgB,aAA9B;;;;SAA8BA,aAA9B;;;AC7EA,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,IAAItC,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAhB;AACP,AAAO,IAAMuC,WAAW,IAAIvC,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB;;AAEP,AAAO,IAAMwC,iBAAiB,CAC5B,QAD4B,EAE5B,OAF4B,EAG5B,MAH4B,EAI5BvC,IAJ4B,CAIvB,GAJuB,CAAvB;;ACIP;;;;;AAKA,AAAe,SAASwC,uBAAT,CAAiCd,CAAjC,EAAoC;IAC/C,KAAF,EAASG,IAAT,CAAc,UAACC,CAAD,EAAIW,GAAJ,EAAY;YAChBC,OAAR,CAAgBD,IAAIE,OAApB,EAA6BC,OAA7B,CAAqC,UAACV,IAAD,EAAU;UACvCD,QAAQQ,IAAIE,OAAJ,CAAYT,IAAZ,CAAd;;UAEIA,SAAS,KAAT,IAAkBG,QAAQlB,IAAR,CAAac,KAAb,CAAlB,IACAK,SAASnB,IAAT,CAAcc,KAAd,CADJ,EAC0B;UACtBQ,GAAF,EAAOP,IAAP,CAAY,KAAZ,EAAmBD,KAAnB;;KALJ;GADF;;SAWOP,CAAP;;;ACtBF,SAASmB,SAAT,CAAmBC,KAAnB,EAA0Bf,IAA1B,EAAgC;SACvBA,KAAKgB,IAAL,KAAc,SAArB;;;AAGF,SAASC,aAAT,CAAuBtB,CAAvB,EAA0B;IACtBuB,IAAF,GAASC,IAAT,CAAc,GAAd,EACSC,QADT,GAESC,MAFT,CAEgBP,SAFhB,EAGSQ,MAHT;;SAKO3B,CAAP;;;AAGF,AAAe,SAAS4B,KAAT,CAAe5B,CAAf,EAAkB;IAC7Ba,cAAF,EAAkBc,MAAlB;;MAEIL,cAActB,CAAd,CAAJ;SACOA,CAAP;;;ACRF,IAAM6B,WAAW;;;;;;;;QAAA,kBAQFnC,GARE,EAQGoC,gBARH,EAQqB;;;;;;;;;oBAAA;;mBAG9BA,gBAH8B;;;;;2BAAA,GAIV;+BACL,IADK;4BAER,GAFQ;yBAGX;kCACS,WADT;oCAEW;;eATU;;;uBAavB,EAAE/C,MAAM+C,gBAAR,EAA0BhD,UAAUiD,aAApC,EAAT;;;;;;qBAEejC,cAAcJ,GAAd,CAfiB;;;oBAAA;;;+CAiB3B,MAAKsC,WAAL,CAAiBC,MAAjB,CAjB2B;;;;;;;;;GARrB;aAAA,6BA4B0B;QAArBC,OAAqB,QAA3BnD,IAA2B;QAAZD,QAAY,QAAZA,QAAY;QACfS,WADe,GACCT,SAASQ,OADV,CAC/B,cAD+B;;;;;QAKnC,CAACC,YAAY4C,QAAZ,CAAqB,MAArB,CAAD,IACA,CAAC5C,YAAY4C,QAAZ,CAAqB,MAArB,CADL,EACmC;YAC3B,IAAI/C,KAAJ,CAAU,qCAAV,CAAN;;;QAGEY,IAAIoC,QAAQC,IAAR,CAAaH,OAAb,EAAsB,EAAEI,qBAAqB,IAAvB,EAAtB,CAAR;;QAEItC,EAAEuB,IAAF,GAASgB,QAAT,GAAoBC,MAApB,KAA+B,CAAnC,EAAsC;YAC9B,IAAIpD,KAAJ,CAAU,kCAAV,CAAN;;;QAGEsB,kBAAkBV,CAAlB,CAAJ;QACIc,wBAAwBd,CAAxB,CAAJ;QACI4B,MAAM5B,CAAN,CAAJ;;WAEOA,CAAP;;CAhDJ;;AAoDA,AACA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AChEA,IAAMyC,iBAAiB;UACb,WADa;WAEZ;;eAEI,CACT,qBADS,EAET,cAFS,EAGT,iBAHS,CAFJ;;;WASA,CACL,KADK,EAEL,uBAFK,CATA;;;;;;;;gBAoBK;;UAEN,IAFM;;;gBAKA,kBAACnC,KAAD,EAAW;YACboC,YAAYpC,MAAMiC,QAAN,EAAlB;YACIG,UAAUF,MAAV,KAAqB,CAArB,IAA0BE,UAAUlE,GAAV,CAAc,CAAd,EAAiBmE,OAAjB,KAA6B,KAA3D,EAAkE;iBACzD,QAAP;;;eAGK,IAAP;;;GAjCe;;SAsCd;eACM,CACT,uBADS,EAET,qBAFS,EAGT,IAHS;GAvCQ;;UA8Cb;eACK,CACT,aADS,EAET,sBAFS;GA/CQ;;OAqDhB;eACQ,CACT,sBADS;GAtDQ;;iBA2DN;eACF,CACT,kCADS,EAET,wBAFS;;CA5Df,CAmEA;;ACnEA,IAAMC,mBAAmB;UACf,cADe;WAEd;;;;eAII,CACT,wBADS,CAJJ;;;WASA,EATA;;;gBAaK;gBACA;;GAhBS;;UAoBf;eACK,CACT,mBADS;GArBU;;SA0BhB;eACM,CACT,UADS;GA3BU;;iBAgCR;eACF,CACT,kBADS;;CAjCf,CAuCA;;ACvCA,IAAMC,qBAAqB;UACjB,eADiB;WAEhB;eACI,CACT,kBADS,CADJ;;;gBAMK;sBACM,oBAACvC,KAAD,EAAW;cACnBwC,OAAN,CAAc,UAAd,EAA0BC,OAA1B,CAAkCzC,KAAlC;OAFQ;0BAIU,YAJV;kBAKE;KAXP;;;WAeA,CACL,iBADK,EAEL,oCAFK,EAGL,MAHK;;GAjBgB;;UAyBjB,wBAzBiB;;SA2BlB;eACM,CACT,UADS;GA5BY;;iBAiCV;eACF,CACT,sBADS;;;CAlCf,CAyCA;;ACrCA,IAAM0C,aAAa;eACJP,cADI;kBAEDG,gBAFC;mBAGAC;CAHnB,CAMA;;ACVA;AACA,AAAO,IAAMI,YAAY,IAAI5E,MAAJ,CAAW,gCAAX,EAA6C,GAA7C,CAAlB;;;AAGP,AAAO,IAAM6E,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,aAAa7E,IAAb,CAAkB,GAAlB,CAAzB;AACP,AAAO,IAAMkF,kBAAkB,CAAC,KAAD,EAAQ,MAAR,EAAgB,OAAhB,EAAyB,IAAzB,CAAxB;AACP,AAAO,IAAMC,qBAAqB,IAAIpF,MAAJ,QAAgBmF,gBAAgBlF,IAAhB,CAAqB,GAArB,CAAhB,SAA+C,GAA/C,CAA3B;;;AAGP,AAAO,IAAMoF,oBAAoB,CAAC,GAAD,CAA1B;AACP,AAAO,IAAMC,yBAAyBD,kBAAkBL,GAAlB,CAAsB;SAAUO,GAAV;CAAtB,EAA6CtF,IAA7C,CAAkD,GAAlD,CAA/B;;;AAGP,AAAO,IAAMuF,2BAA2B,CAAC,IAAD,EAAO,IAAP,EAAa,OAAb,EAAsB,KAAtB,EAA6BvF,IAA7B,CAAkC,GAAlC,CAAjC;;;AAGP,IAAMwF,cAAc,CAAC,IAAD,EAAO,IAAP,EAAa,IAAb,EAAmB,IAAnB,EAAyB,IAAzB,CAApB;AACA,AAAO,IAAMC,kBAAkBD,YAAYxF,IAAZ,CAAiB,GAAjB,CAAxB;;;;;;;;AASP,AAAO,IAAM0F,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,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,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,EASjC5F,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM6F,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,IAAI/F,MAAJ,QAAgB8F,uBAAuB7F,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;AAGP,AAYA,AAAO,IAAM+F,cAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,iBAAiB,IAAIjG,MAAJ,CAAWgG,YAAY/F,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAMiG,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,IAAInG,MAAJ,CAAWkG,qBAAqBjG,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA;;;;AAIA,AAAO,IAAMmG,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,IAAIrG,MAAJ,CAAWoG,qBAAqBnG,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMqG,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,IAAIxG,MAAJ,CAAWuG,sBAAsBtG,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;AAGP,AAAO,IAAMwG,UAAU,IAAIzG,MAAJ,CAAW,iBAAX,EAA8B,GAA9B,CAAhB;;AAEP,AAMA,AAIA,AAIA,AAGA,AAGA;;AAEA,AAAO,IAAM0G,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,IAAI3G,MAAJ,QAAgB0G,iBAAiBzG,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAM2G,sBAAsBjB,8BAA8B1F,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM4G,uBAAuB,IAAI7G,MAAJ,CAAW4G,mBAAX,EAAgC,GAAhC,CAA7B;;AAEP,IAAME,sBAAsBlB,8BAA8B3F,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAAO,IAAM8G,uBAAuB,IAAI/G,MAAJ,CAAW8G,mBAAX,EAAgC,GAAhC,CAA7B,CAEP,AAGA,AACA,AACA,AAEA;;ACzXA;;;;AAIA,AAAe,SAASE,uBAAT,CAAiCrF,CAAjC,EAAoC;;;;;;;;;;IAU/C,GAAF,EAAOsF,GAAP,CAAW,GAAX,EAAgBnF,IAAhB,CAAqB,UAACiB,KAAD,EAAQf,IAAR,EAAiB;QAC9BC,QAAQN,EAAEK,IAAF,CAAd;QACMkF,UAAUjF,MAAME,IAAN,CAAW,OAAX,CAAhB;QACMgF,KAAKlF,MAAME,IAAN,CAAW,IAAX,CAAX;QACI,CAACgF,EAAD,IAAO,CAACD,OAAZ,EAAqB;;QAEfE,cAAgBF,WAAW,EAA3B,WAAiCC,MAAM,EAAvC,CAAN;QACIJ,qBAAqB3F,IAArB,CAA0BgG,UAA1B,CAAJ,EAA2C;;KAA3C,MAEO,IAAIP,qBAAqBzF,IAArB,CAA0BgG,UAA1B,CAAJ,EAA2C;YAC1C9D,MAAN;;GAVJ;;SAcO3B,CAAP;;;AC/BF;;;;;;;;;;AAUA,AAAe,SAAS0F,OAAT,CAAiB1F,CAAjB,EAAoB;MAC7B2F,aAAa,KAAjB;IACE,IAAF,EAAQxF,IAAR,CAAa,UAACiB,KAAD,EAAQwE,OAAR,EAAoB;QACzBC,cAAc7F,EAAE4F,OAAF,EAAWE,IAAX,GAAkBtH,GAAlB,CAAsB,CAAtB,CAApB;;QAEIqH,eAAeA,YAAYlD,OAAZ,KAAwB,IAA3C,EAAiD;mBAClC,IAAb;QACEiD,OAAF,EAAWjE,MAAX;KAFF,MAGO,IAAIgE,UAAJ,EAAgB;mBACR,KAAb;;mBAEaC,OAAb,EAAsB5F,CAAtB,EAAyB,IAAzB;;GATJ;;SAaOA,CAAP;;;ACzBF;;;;;;;;;;;AAWA,AAAe,SAAS+F,YAAT,CAAsB1F,IAAtB,EAA4BL,CAA5B,EAA2C;MAAZgG,EAAY,yDAAP,KAAO;;MAClD1F,QAAQN,EAAEK,IAAF,CAAd;;MAEI2F,EAAJ,EAAQ;QACFC,UAAU5F,KAAK6F,WAAnB;QACMC,IAAInG,EAAE,SAAF,CAAV;;;;WAIOiG,WAAW,EAAEA,QAAQtD,OAAR,IAAmBqC,oBAAoBvF,IAApB,CAAyBwG,QAAQtD,OAAjC,CAArB,CAAlB,EAAmF;UAC3EuD,cAAcD,QAAQC,WAA5B;QACED,OAAF,EAAWG,QAAX,CAAoBD,CAApB;gBACUD,WAAV;;;UAGIG,WAAN,CAAkBF,CAAlB;UACMxE,MAAN;WACO3B,CAAP;;;SAGKA,CAAP;;;AC7BF,SAASsG,WAAT,CAAqBtG,CAArB,EAAwB;IACpB,KAAF,EAASG,IAAT,CAAc,UAACiB,KAAD,EAAQmF,GAAR,EAAgB;QACtBC,OAAOxG,EAAEuG,GAAF,CAAb;QACME,cAAcD,KAAKjE,QAAL,CAAc2B,mBAAd,EAAmC1B,MAAnC,KAA8C,CAAlE;;QAEIiE,WAAJ,EAAiB;oBACDD,IAAd,EAAoBxG,CAApB,EAAuB,GAAvB;;GALJ;;SASOA,CAAP;;;AAGF,SAAS0G,YAAT,CAAsB1G,CAAtB,EAAyB;IACrB,MAAF,EAAUG,IAAV,CAAe,UAACiB,KAAD,EAAQuF,IAAR,EAAiB;QACxBC,QAAQ5G,EAAE2G,IAAF,CAAd;QACMF,cAAcG,MAAM9D,OAAN,CAAc,QAAd,EAAwBN,MAAxB,KAAmC,CAAvD;QACIiE,WAAJ,EAAiB;oBACDG,KAAd,EAAqB5G,CAArB,EAAwB,GAAxB;;GAJJ;;SAQOA,CAAP;;;;;;;;;;;;;;;AAeF,AAAe,SAAS6G,mBAAT,CAA6B7G,CAA7B,EAAgC;MACzC0F,QAAQ1F,CAAR,CAAJ;MACIsG,YAAYtG,CAAZ,CAAJ;MACI0G,aAAa1G,CAAb,CAAJ;;SAEOA,CAAP;;;AC5Ca,SAAS8G,aAAT,CAAuBxG,KAAvB,EAA8BN,CAA9B,EAA4C;MAAX4D,GAAW,yDAAL,GAAK;;mBACrCtD,MAAM9B,GAAN,CAAU,CAAV,CADqC;;MACjDyC,OADiD,cACjDA,OADiD;;MAEnD8F,eAAeC,QAAQhG,OAAR,CAAgBC,OAAhB,EACQoC,GADR,CACY;WAAU4D,GAAV,SAAiBhG,QAAQgG,GAAR,CAAjB;GADZ,EAEQ3I,IAFR,CAEa,GAFb,CAArB;;QAIM+H,WAAN,OAAsBzC,GAAtB,SAA6BmD,YAA7B,SAA6CzG,MAAMmB,QAAN,EAA7C,UAAkEmC,GAAlE;SACO5D,CAAP;;;ACPF,SAASkH,cAAT,CAAwBC,IAAxB,EAA8BnH,CAA9B,EAAiC;MACzBoH,SAASC,SAASF,KAAK3G,IAAL,CAAU,QAAV,CAAT,EAA8B,EAA9B,CAAf;MACM8G,QAAQD,SAASF,KAAK3G,IAAL,CAAU,OAAV,CAAT,EAA6B,EAA7B,KAAoC,EAAlD;;;;;MAKI,CAAC4G,UAAU,EAAX,IAAiB,EAAjB,IAAuBE,QAAQ,EAAnC,EAAuC;SAChC3F,MAAL;GADF,MAEO,IAAIyF,MAAJ,EAAY;;;;SAIZ3G,UAAL,CAAgB,QAAhB;;;SAGKT,CAAP;;;;;AAKF,SAASuH,aAAT,CAAuBJ,IAAvB,EAA6BnH,CAA7B,EAAgC;MAC1BiD,UAAUxD,IAAV,CAAe0H,KAAK3G,IAAL,CAAU,KAAV,CAAf,CAAJ,EAAsC;SAC/BmB,MAAL;;;SAGK3B,CAAP;;;AAGF,AAAe,SAASwH,WAAT,CAAqBC,QAArB,EAA+BzH,CAA/B,EAAkC;WACtCwB,IAAT,CAAc,KAAd,EAAqBrB,IAArB,CAA0B,UAACiB,KAAD,EAAQL,GAAR,EAAgB;QAClCoG,OAAOnH,EAAEe,GAAF,CAAb;;mBAEeoG,IAAf,EAAqBnH,CAArB;kBACcmH,IAAd,EAAoBnH,CAApB;GAJF;;SAOOA,CAAP;;;ACnCa,SAAS0H,aAAT,CAAuBC,OAAvB,EAAgC3H,CAAhC,EAAmC;IAC9CkD,kBAAkB5E,IAAlB,CAAuB,GAAvB,CAAF,EAA+BqJ,OAA/B,EAAwChG,MAAxC;;SAEO3B,CAAP;;;ACLF;;;;AAGA,AAAe,SAAS4H,UAAT,CAAoBD,OAApB,EAA6B3H,CAA7B,EAAgC;;MAEvC6H,SAAS7H,EAAE,IAAF,EAAQ2H,OAAR,CAAf;MACIE,OAAOrF,MAAP,GAAgB,CAApB,EAAuB;WACdrC,IAAP,CAAY,UAACiB,KAAD,EAAQf,IAAR;aAAiBL,EAAEK,IAAF,EAAQsB,MAAR,EAAjB;KAAZ;GADF,MAEO;WACExB,IAAP,CAAY,UAACiB,KAAD,EAAQf,IAAR,EAAiB;oBACbL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0B,IAA1B;KADF;;;SAKKA,CAAP;;;ACZF,SAAS8H,qBAAT,CAA+BL,QAA/B,EAAyC;;WAE9BjG,IAAT,CAAc,GAAd,EAAmBrB,IAAnB,CAAwB,UAACiB,KAAD,EAAQf,IAAR,EAAiB;SAClCY,OAAL,GAAe+F,QAAQhG,OAAR,CAAgBX,KAAKY,OAArB,EAA8B8G,MAA9B,CAAqC,UAACC,GAAD,EAAMxH,IAAN,EAAe;UAC7DiD,mBAAmBhE,IAAnB,CAAwBe,IAAxB,CAAJ,EAAmC;4BACrBwH,GAAZ,qBAAkBxH,IAAlB,EAAyBH,KAAKY,OAAL,CAAaT,IAAb,CAAzB;;;aAGKwH,GAAP;KALa,EAMZ,EANY,CAAf;GADF;;;;;;;;;;AAkBF,AAAe,SAASC,eAAT,CAAyBR,QAAzB,EAAmC;wBAC1BA,QAAtB;;SAEOA,QAAP;;;AC3Ba,SAASS,WAAT,CAAqBT,QAArB,EAA+BzH,CAA/B,EAAkC;WACtCwB,IAAT,CAAc,GAAd,EAAmBrB,IAAnB,CAAwB,UAACiB,KAAD,EAAQ+E,CAAR,EAAc;QAC9BgC,KAAKnI,EAAEmG,CAAF,CAAX;QACIgC,GAAGC,IAAH,GAAUC,IAAV,OAAqB,EAAzB,EAA6BF,GAAGxG,MAAH;GAF/B;;SAKO3B,CAAP;;;ACNF;;;;;;AAMA,AAAO,IAAMgE,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,EASjC5F,IATiC,CAS5B,GAT4B,CAA5B;;;;AAaP,AAAO,IAAM6F,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,IAAI/F,MAAJ,QAAgB8F,yBAAuB7F,IAAvB,CAA4B,GAA5B,CAAhB,SAAsD,GAAtD,CADK;;;;;AAMP,AAAO,IAAMgK,4BAA0B,CACrC,CAAC,SAAD,EAAY,gBAAZ,CADqC,EAErC,CAAC,OAAD,EAAU,gBAAV,CAFqC,EAGrC,CAAC,QAAD,EAAW,gBAAX,CAHqC,EAIrC,CAAC,OAAD,EAAU,WAAV,CAJqC,EAKrC,CAAC,OAAD,EAAU,YAAV,CALqC,EAMrC,CAAC,OAAD,EAAU,YAAV,CANqC,CAAhC;;AASP,AAAO,IAAMjE,gBAAc,CACzB,QADyB,EAEzB,OAFyB,EAGzB,OAHyB,EAIzB,SAJyB,CAApB;AAMP,AAAO,IAAMC,mBAAiB,IAAIjG,MAAJ,CAAWgG,cAAY/F,IAAZ,CAAiB,GAAjB,CAAX,EAAkC,GAAlC,CAAvB;;;;;;AAOP,AAAO,IAAMiG,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,IAAInG,MAAJ,CAAWkG,uBAAqBjG,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;;AAGP,AAAO,IAAMiK,sBAAoB,IAAIlK,MAAJ,CAAW,qBAAX,EAAkC,GAAlC,CAA1B;;;;;;AAMP,AAAO,IAAMoG,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,IAAIrG,MAAJ,CAAWoG,uBAAqBnG,IAArB,CAA0B,GAA1B,CAAX,EAA2C,GAA3C,CAA1B;;AAEP,AAGA,AAGA,AAGA;;AAEA,AAAO,IAAMyG,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,IAAI3G,MAAJ,QAAgB0G,mBAAiBzG,IAAjB,CAAsB,GAAtB,CAAhB,SAAgD,GAAhD,CAA5B;;;;;;AAOP,IAAM2G,wBAAsBjB,gCAA8B1F,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,IAAM6G,wBAAsBlB,gCAA8B3F,IAA9B,CAAmC,GAAnC,CAA5B;AACA,AAEA,AAGA,AAAO,IAAMkK,yBAAuB,IAAInK,MAAJ,CAAW,mBAAX,EAAgC,GAAhC,CAA7B;AACP,AAAO,IAAMoK,uBAAqB,IAAIpK,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAA3B;AACP,AAAO,IAAMqK,aAAW,IAAIrK,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAjB,CAEP;;AC3SA;AACA,AAAe,SAASsK,SAAT,CAAmBtI,IAAnB,EAAyB;MAChCkF,UAAUlF,KAAKG,IAAL,CAAU,OAAV,CAAhB;MACMgF,KAAKnF,KAAKG,IAAL,CAAU,IAAV,CAAX;MACIoI,QAAQ,CAAZ;;MAEIpD,EAAJ,EAAQ;;QAEFhB,oBAAkB/E,IAAlB,CAAuB+F,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;QAEEd,oBAAkBjF,IAAlB,CAAuB+F,EAAvB,CAAJ,EAAgC;eACrB,EAAT;;;;MAIAD,OAAJ,EAAa;QACPqD,UAAU,CAAd,EAAiB;;;UAGXpE,oBAAkB/E,IAAlB,CAAuB8F,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;UAEEb,oBAAkBjF,IAAlB,CAAuB8F,OAAvB,CAAJ,EAAqC;iBAC1B,EAAT;;;;;;;QAOAjB,iBAAe7E,IAAf,CAAoB8F,OAApB,CAAJ,EAAkC;eACvB,EAAT;;;;;;;QAOEgD,oBAAkB9I,IAAlB,CAAuB8F,OAAvB,CAAJ,EAAqC;eAC1B,EAAT;;;;SAIGqD,KAAP;;;ACpDF;;;AAGA,AAAe,SAASC,QAAT,CAAkBvI,KAAlB,EAAyB;SAC/BwI,WAAWxI,MAAME,IAAN,CAAW,OAAX,CAAX,KAAmC,IAA1C;;;ACJF;AACA,AAAe,SAASuI,WAAT,CAAqBX,IAArB,EAA2B;SACjC,CAACA,KAAKY,KAAL,CAAW,IAAX,KAAoB,EAArB,EAAyBxG,MAAhC;;;ACFF,IAAMyG,QAAQ,IAAI5K,MAAJ,CAAW,WAAX,EAAwB,GAAxB,CAAd;;AAEA,AAAe,SAAS6K,WAAT,CAAqBC,UAArB,EAAgD;MAAfxG,OAAe,yDAAL,GAAK;;MACvDyG,SAASD,aAAa,EAA5B;;MAEIC,SAAS,CAAb,EAAgB;QACVC,oBAAJ;;;;;;;QAOIJ,MAAMxJ,IAAN,CAAWkD,OAAX,CAAJ,EAAyB;oBACTyG,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,CAAwBpJ,IAAxB,EAA8B;MACvCuI,QAAQ,CAAZ;MACMR,OAAO/H,KAAK+H,IAAL,GAAYC,IAAZ,EAAb;MACMc,aAAaf,KAAK5F,MAAxB;;;MAGI2G,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,CAAkBrJ,KAAlB,EAAyBN,CAAzB,EAA4B4I,KAA5B,EAAmC;QAC1CpI,IAAN,CAAW,OAAX,EAAoBoI,KAApB;SACOtI,KAAP;;;ACEa,SAASsJ,QAAT,CAAkBtJ,KAAlB,EAAyBN,CAAzB,EAA4B6J,MAA5B,EAAoC;MAC7C;QACIjB,QAAQkB,eAAexJ,KAAf,EAAsBN,CAAtB,IAA2B6J,MAAzC;aACSvJ,KAAT,EAAgBN,CAAhB,EAAmB4I,KAAnB;GAFF,CAGE,OAAOmB,CAAP,EAAU;;;;SAILzJ,KAAP;;;ACXF;AACA,AAAe,SAAS0J,WAAT,CAAqB3J,IAArB,EAA2BL,CAA3B,EAA8B4I,KAA9B,EAAqC;MAC5CqB,SAAS5J,KAAK4J,MAAL,EAAf;MACIA,MAAJ,EAAY;aACDA,MAAT,EAAiBjK,CAAjB,EAAoB4I,QAAQ,IAA5B;;;SAGKvI,IAAP;;;ACFF;;;AAGA,AAAe,SAASyJ,cAAT,CAAwBxJ,KAAxB,EAA+BN,CAA/B,EAAsD;MAApBkK,WAAoB,yDAAN,IAAM;;MAC/DtB,QAAQC,SAASvI,KAAT,CAAZ;;MAEIsI,KAAJ,EAAW;WACFA,KAAP;;;UAGMuB,UAAU7J,KAAV,CAAR;;MAEI4J,WAAJ,EAAiB;aACNvB,UAAUrI,KAAV,CAAT;;;cAGUA,KAAZ,EAAmBN,CAAnB,EAAsB4I,KAAtB;;SAEOA,KAAP;;;AClBF;;AAEA,AAAe,SAASuB,SAAT,CAAmB7J,KAAnB,EAA0B;mBACnBA,MAAM9B,GAAN,CAAU,CAAV,CADmB;;MAC/BmE,OAD+B,cAC/BA,OAD+B;;;;;;MAMnC6F,uBAAqB/I,IAArB,CAA0BkD,OAA1B,CAAJ,EAAwC;WAC/B8G,eAAenJ,KAAf,CAAP;GADF,MAEO,IAAIqC,YAAY,KAAhB,EAAuB;WACrB,CAAP;GADK,MAEA,IAAI8F,qBAAmBhJ,IAAnB,CAAwBkD,OAAxB,CAAJ,EAAsC;WACpC,CAAP;GADK,MAEA,IAAI+F,WAASjJ,IAAT,CAAckD,OAAd,CAAJ,EAA4B;WAC1B,CAAC,CAAR;GADK,MAEA,IAAIA,YAAY,IAAhB,EAAsB;WACpB,CAAC,CAAR;;;SAGK,CAAP;;;ACjBF,SAAS+D,cAAT,CAAsBpG,KAAtB,EAA6BN,CAA7B,EAAgC;MAC1BM,MAAM9B,GAAN,CAAU,CAAV,CAAJ,EAAkB;qBACI8B,MAAM9B,GAAN,CAAU,CAAV,CADJ;;QACRmE,OADQ,cACRA,OADQ;;;QAGZA,YAAY,MAAhB,EAAwB;;oBAERrC,KAAd,EAAqBN,CAArB,EAAwB,KAAxB;;;;;AAKN,SAASoK,UAAT,CAAoB9J,KAApB,EAA2BN,CAA3B,EAA8B4I,KAA9B,EAAqC;MAC/BtI,KAAJ,EAAW;mBACIA,KAAb,EAAoBN,CAApB;aACSM,KAAT,EAAgBN,CAAhB,EAAmB4I,KAAnB;;;;AAIJ,SAASyB,OAAT,CAAiBrK,CAAjB,EAAoBkK,WAApB,EAAiC;IAC7B,QAAF,EAAY5E,GAAZ,CAAgB,SAAhB,EAA2BnF,IAA3B,CAAgC,UAACiB,KAAD,EAAQf,IAAR,EAAiB;;;QAG3CC,QAAQN,EAAEK,IAAF,CAAZ;YACQsJ,SAASrJ,KAAT,EAAgBN,CAAhB,EAAmB8J,eAAexJ,KAAf,EAAsBN,CAAtB,EAAyBkK,WAAzB,CAAnB,CAAR;;QAEMI,UAAUhK,MAAM2J,MAAN,EAAhB;QACMM,WAAWJ,UAAU7J,KAAV,CAAjB;;eAEWgK,OAAX,EAAoBtK,CAApB,EAAuBuK,QAAvB,EAAiCL,WAAjC;QACII,OAAJ,EAAa;;;iBAGAA,QAAQL,MAAR,EAAX,EAA6BjK,CAA7B,EAAgCuK,WAAW,CAA3C,EAA8CL,WAA9C;;GAbJ;;SAiBOlK,CAAP;;;;;AAKF,AAAe,SAASwK,YAAT,CAAsBxK,CAAtB,EAA6C;MAApBkK,WAAoB,yDAAN,IAAM;;;;4BAGlChJ,OAAxB,CAAgC,gBAAqC;;;QAAnCuJ,cAAmC;QAAnBC,aAAmB;;MAC9DD,cAAL,SAAuBC,aAAvB,EAAwCvK,IAAxC,CAA6C,UAACiB,KAAD,EAAQf,IAAR,EAAiB;eACnDL,EAAEK,IAAF,EAAQ4J,MAAR,CAAeQ,cAAf,CAAT,EAAyCzK,CAAzC,EAA4C,EAA5C;KADF;GADF;;;;;;;MAWIqK,QAAQrK,CAAR,EAAWkK,WAAX,CAAJ;MACIG,QAAQrK,CAAR,EAAWkK,WAAX,CAAJ;;SAEOlK,CAAP;;;ACpEF,IAAM2K,eAAe,SAArB;;AAEA,AAAe,SAASC,eAAT,CAAyBxC,IAAzB,EAA+B;SACrCA,KAAKyC,OAAL,CAAaF,YAAb,EAA2B,GAA3B,EAAgCtC,IAAhC,EAAP;;;ACHF;;;;;AAKA,AAAe,SAASyC,cAAT,CAAwBpL,GAAxB,EAA6BqL,SAA7B,EAAwC;MAC/CC,UAAUD,UAAUvJ,IAAV,CAAe;WAAMyJ,GAAGxL,IAAH,CAAQC,GAAR,CAAN;GAAf,CAAhB;MACIsL,OAAJ,EAAa;WACJA,QAAQE,IAAR,CAAaxL,GAAb,EAAkB,CAAlB,CAAP;;;SAGK,IAAP;;;ACXF;;;;;;;;;;;;;;;;AAgBA,AAAO,IAAMyL,kBAAkB,IAAI9M,MAAJ,CAAW,0EAAX,EAAuF,GAAvF,CAAxB;;AAEP,AAAO,IAAM+M,eAAe,QAArB;;AAEP,AAAO,IAAMC,cAAc,WAApB;AACP,AAAO,IAAMC,cAAc,WAApB;;ACnBQ,SAASC,cAAT,CAAwB7L,GAAxB,EAA6B;MACpC8L,UAAU9L,IAAIsJ,KAAJ,CAAUmC,eAAV,CAAhB;MACI,CAACK,OAAL,EAAc,OAAO,IAAP;;MAERC,UAAUpE,SAASmE,QAAQ,CAAR,CAAT,EAAqB,EAArB,CAAhB;;;;SAIOC,UAAU,GAAV,GAAgBA,OAAhB,GAA0B,IAAjC;;;ACVa,SAASC,YAAT,CAAsBhM,GAAtB,EAA2B;SACjCA,IAAIiM,KAAJ,CAAU,GAAV,EAAe,CAAf,EAAkBd,OAAlB,CAA0B,KAA1B,EAAiC,EAAjC,CAAP;;;ACOF,SAASe,aAAT,CAAuBC,OAAvB,EAAgCzK,KAAhC,EAAuC0K,sBAAvC,EAA+D;MACzDC,cAAc,IAAlB;;;;MAII3K,QAAQ,CAAR,IAAakK,YAAY7L,IAAZ,CAAiBoM,OAAjB,CAAb,IAA0CA,QAAQrJ,MAAR,GAAiB,CAA/D,EAAkE;kBAClD,IAAd;;;;;MAKEpB,UAAU,CAAV,IAAeyK,QAAQG,WAAR,OAA0B,OAA7C,EAAsD;kBACtC,KAAd;;;;;MAKE5K,QAAQ,CAAR,IAAayK,QAAQrJ,MAAR,GAAiB,CAA9B,IAAmC,CAACsJ,sBAAxC,EAAgE;kBAChD,KAAd;;;SAGKC,WAAP;;;;;;AAMF,AAAe,SAASE,cAAT,CAAwBvM,GAAxB,EAA6BwM,MAA7B,EAAqC;MAC5CrM,YAAYqM,UAAUvM,IAAIC,KAAJ,CAAUF,GAAV,CAA5B;MACQyM,QAF0C,GAEjBtM,SAFiB,CAE1CsM,QAF0C;MAEhCC,IAFgC,GAEjBvM,SAFiB,CAEhCuM,IAFgC;MAE1BC,IAF0B,GAEjBxM,SAFiB,CAE1BwM,IAF0B;;;MAI9CP,yBAAyB,KAA7B;MACMQ,kBAAkBD,KAAKV,KAAL,CAAW,GAAX,EACvBY,OADuB,GAEvBxE,MAFuB,CAEhB,UAACC,GAAD,EAAMwE,UAAN,EAAkBpL,KAAlB,EAA4B;QAC9ByK,UAAUW,UAAd;;;QAGIX,QAAQ1J,QAAR,CAAiB,GAAjB,CAAJ,EAA2B;2BACU0J,QAAQF,KAAR,CAAc,GAAd,CADV;;;;UAClBc,eADkB;UACDC,OADC;;UAErBrB,YAAY5L,IAAZ,CAAiBiN,OAAjB,CAAJ,EAA+B;kBACnBD,eAAV;;;;;;QAMAtB,gBAAgB1L,IAAhB,CAAqBoM,OAArB,KAAiCzK,QAAQ,CAA7C,EAAgD;gBACpCyK,QAAQhB,OAAR,CAAgBM,eAAhB,EAAiC,EAAjC,CAAV;;;;;;;QAOE/J,UAAU,CAAd,EAAiB;+BACUgK,aAAa3L,IAAb,CAAkBoM,OAAlB,CAAzB;;;;QAIED,cAAcC,OAAd,EAAuBzK,KAAvB,EAA8B0K,sBAA9B,CAAJ,EAA2D;UACrDa,IAAJ,CAASd,OAAT;;;WAGK7D,GAAP;GAhCsB,EAiCrB,EAjCqB,CAAxB;;SAmCUmE,QAAV,UAAuBC,IAAvB,GAA8BE,gBAAgBC,OAAhB,GAA0BjO,IAA1B,CAA+B,GAA/B,CAA9B;;;AC3EF;;AAEA,IAAMsO,kBAAkB,IAAIvO,MAAJ,CAAW,QAAX,CAAxB;AACA,AAAe,SAASwO,cAAT,CAAwBzE,IAAxB,EAA8B;SACpCwE,gBAAgBnN,IAAhB,CAAqB2I,IAArB,CAAP;;;ACKF;;;;;AAKA,AAAe,SAAS0E,aAAT,CAAuBC,UAAvB,EAAmCC,QAAnC,EAA6ChN,CAA7C,EAAgD;MACzD,CAAC+M,WAAW9C,MAAX,GAAoBzH,MAAzB,EAAiC;WACxBuK,UAAP;;;MAGIE,wBAAwB3D,KAAKE,GAAL,CAAS,EAAT,EAAawD,WAAW,GAAxB,CAA9B;MACME,cAAclN,EAAE,aAAF,CAApB;;aAEWiK,MAAX,GAAoB1H,QAApB,GAA+BpC,IAA/B,CAAoC,UAACiB,KAAD,EAAQ+L,KAAR,EAAkB;QAC9CC,SAASpN,EAAEmN,KAAF,CAAf;;QAEI/I,4BAA0B3E,IAA1B,CAA+B0N,MAAMxK,OAArC,CAAJ,EAAmD;aAC1C,IAAP;;;QAGI0K,aAAaxE,SAASuE,MAAT,CAAnB;QACIC,UAAJ,EAAgB;UACVD,WAAWL,UAAf,EAA2B;oBACbO,MAAZ,CAAmBF,MAAnB;OADF,MAEO;YACDG,eAAe,CAAnB;;YAEMC,UAAUC,YAAYL,MAAZ,CAAhB;;;;YAIII,UAAU,IAAd,EAAoB;0BACF,EAAhB;;;;;YAKEA,WAAW,GAAf,EAAoB;0BACF,EAAhB;;;;;YAKEJ,OAAO5M,IAAP,CAAY,OAAZ,MAAyBuM,WAAWvM,IAAX,CAAgB,OAAhB,CAA7B,EAAuD;0BACrCwM,WAAW,GAA3B;;;YAGIU,WAAW7E,SAASuE,MAAT,IAAmBG,YAApC;;YAEIG,YAAYT,qBAAhB,EAAuC;iBAC9BC,YAAYI,MAAZ,CAAmBF,MAAnB,CAAP;SADF,MAEO,IAAID,MAAMxK,OAAN,KAAkB,GAAtB,EAA2B;cAC1BgL,eAAeP,OAAOhF,IAAP,EAArB;cACMwF,qBAAqBzE,WAAWwE,YAAX,CAA3B;;cAEIC,qBAAqB,EAArB,IAA2BJ,UAAU,IAAzC,EAA+C;mBACtCN,YAAYI,MAAZ,CAAmBF,MAAnB,CAAP;WADF,MAEO,IAAIQ,sBAAsB,EAAtB,IAA4BJ,YAAY,CAAxC,IACDX,eAAec,YAAf,CADH,EACiC;mBAC/BT,YAAYI,MAAZ,CAAmBF,MAAnB,CAAP;;;;;;WAMD,IAAP;GApDF;;SAuDOF,WAAP;;;ACzEF;;AAEA,AAAe,SAASW,gBAAT,CAA0B7N,CAA1B,EAA6B;MACtC+M,mBAAJ;MACIC,WAAW,CAAf;;IAEE,SAAF,EAAa7M,IAAb,CAAkB,UAACiB,KAAD,EAAQf,IAAR,EAAiB;QAC3BC,QAAQN,EAAEK,IAAF,CAAd;QACMuI,QAAQC,SAASvI,KAAT,CAAd;;QAEI8D,4BAA0B3E,IAA1B,CAA+BY,KAAKsC,OAApC,CAAJ,EAAkD;;;;;;QAM9CiG,QAAQoE,QAAZ,EAAsB;iBACTpE,KAAX;mBACatI,KAAb;;GAZJ;;;;MAkBI,CAACyM,UAAL,EAAiB;WACR/M,EAAE,MAAF,KAAaA,EAAE,GAAF,EAAO8N,KAAP,EAApB;;;eAGWhB,cAAcC,UAAd,EAA0BC,QAA1B,EAAoChN,CAApC,CAAb;;SAEO+M,UAAP;;;ACvBF,SAASgB,mBAAT,CAA6BzN,KAA7B,EAAoCN,CAApC,EAAuCgO,MAAvC,EAA+C;;;;;MAKzC1N,MAAM2N,QAAN,CAAe,qBAAf,CAAJ,EAA2C;;;;MAIrC/L,UAAU0I,gBAAgBtK,MAAM8H,IAAN,EAAhB,CAAhB;;MAEIW,YAAY7G,OAAZ,IAAuB,EAA3B,EAA+B;QACvBgM,SAASlO,EAAE,GAAF,EAAOM,KAAP,EAAckC,MAA7B;QACM2L,aAAanO,EAAE,OAAF,EAAWM,KAAX,EAAkBkC,MAArC;;;QAGI2L,aAAcD,SAAS,CAA3B,EAA+B;YACvBvM,MAAN;;;;QAIInC,gBAAgB0C,QAAQM,MAA9B;QACM4L,WAAWpO,EAAE,KAAF,EAASM,KAAT,EAAgBkC,MAAjC;;;;QAIIhD,gBAAgB,EAAhB,IAAsB4O,aAAa,CAAvC,EAA0C;YAClCzM,MAAN;;;;QAII6L,UAAUC,YAAYnN,KAAZ,CAAhB;;;;;QAKI0N,SAAS,EAAT,IAAeR,UAAU,GAAzB,IAAgChO,gBAAgB,EAApD,EAAwD;YAChDmC,MAAN;;;;;;QAMEqM,UAAU,EAAV,IAAgBR,UAAU,GAA9B,EAAmC;;;;UAI3B7K,UAAUrC,MAAM9B,GAAN,CAAU,CAAV,EAAamE,OAA7B;UACM0L,aAAa1L,YAAY,IAAZ,IAAoBA,YAAY,IAAnD;UACI0L,UAAJ,EAAgB;YACRC,eAAehO,MAAMiO,IAAN,EAArB;YACID,gBAAgB1D,gBAAgB0D,aAAalG,IAAb,EAAhB,EAAqCsB,KAArC,CAA2C,CAAC,CAA5C,MAAmD,GAAvE,EAA4E;;;;;YAKxE/H,MAAN;;;;QAII6M,cAAcxO,EAAE,QAAF,EAAYM,KAAZ,EAAmBkC,MAAvC;;;QAGIgM,cAAc,CAAd,IAAmBhP,gBAAgB,GAAvC,EAA4C;YACpCmC,MAAN;;;;;;;;;;;;;AAaN,AAAe,SAAS8M,SAAT,CAAmBhH,QAAnB,EAA6BzH,CAA7B,EAAgC;IAC3C6D,wBAAF,EAA4B4D,QAA5B,EAAsCtH,IAAtC,CAA2C,UAACiB,KAAD,EAAQf,IAAR,EAAiB;QACpDC,QAAQN,EAAEK,IAAF,CAAd;QACI2N,SAASnF,SAASvI,KAAT,CAAb;QACI,CAAC0N,MAAL,EAAa;eACFlE,eAAexJ,KAAf,EAAsBN,CAAtB,CAAT;eACSM,KAAT,EAAgBN,CAAhB,EAAmBgO,MAAnB;;;;QAIEA,SAAS,CAAb,EAAgB;YACRrM,MAAN;KADF,MAEO;;0BAEerB,KAApB,EAA2BN,CAA3B,EAA8BgO,MAA9B;;GAbJ;;SAiBOhO,CAAP;;;ACrGa,SAAS0O,YAAT,CAAsBjH,QAAtB,EAAgCzH,CAAhC,EAA+C;MAAZ2O,KAAY,yDAAJ,EAAI;;IAC1D5K,eAAF,EAAmB0D,QAAnB,EAA6BtH,IAA7B,CAAkC,UAACiB,KAAD,EAAQwN,MAAR,EAAmB;QAC7CC,UAAU7O,EAAE4O,MAAF,CAAhB;;;;;QAKI5O,EAAE6O,OAAF,EAAWpH,QAAX,EAAqBqH,OAArB,CAA6B,GAA7B,EAAkCtM,MAAlC,KAA6C,CAAjD,EAAoD;aAC3CqM,QAAQlN,MAAR,EAAP;;;;QAIEiJ,gBAAgB5K,EAAE4O,MAAF,EAAUxG,IAAV,EAAhB,MAAsCuG,KAA1C,EAAiD;aACxCE,QAAQlN,MAAR,EAAP;;;;;QAKEgH,UAAU3I,EAAE4O,MAAF,CAAV,IAAuB,CAA3B,EAA8B;aACrBC,QAAQlN,MAAR,EAAP;;;WAGKkN,OAAP;GArBF;;SAwBO7O,CAAP;;;AC5BF;;;AAEA,AAAe,SAAS+O,eAAT,CAAyBpH,OAAzB,EAAkC3H,CAAlC,EAAqC;;;;MAI9C8G,cAAc9G,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;MACI8G,cAAc9G,EAAE,MAAF,CAAd,EAAyBA,CAAzB,EAA4B,KAA5B,CAAJ;;SAEOA,CAAP;;;ACTF,SAASgP,UAAT,CAAoBhP,CAApB,EAAuBiP,OAAvB,EAAgCzO,IAAhC,EAAsC0O,QAAtC,EAAgD;UACxC1O,IAAN,QAAe0O,QAAf,EAAyB/O,IAAzB,CAA8B,UAACC,CAAD,EAAIC,IAAJ,EAAa;QACnCX,MAAMW,KAAKY,OAAL,CAAaT,IAAb,CAAZ;QACM2O,cAAcxP,IAAIhB,OAAJ,CAAYsQ,OAAZ,EAAqBvP,GAArB,CAApB;;SAEKuB,OAAL,CAAaT,IAAb,IAAqB2O,WAArB;GAJF;;;AAQF,AAAe,SAASC,iBAAT,CAA2BF,QAA3B,EAAqClP,CAArC,EAAwCN,GAAxC,EAA6C;GACzD,MAAD,EAAS,KAAT,EAAgBwB,OAAhB,CAAwB;WAAQ8N,WAAWhP,CAAX,EAAcN,GAAd,EAAmBc,IAAnB,EAAyB0O,QAAzB,CAAR;GAAxB;;;SAGOA,QAAP;;;ACdK,SAAS/F,UAAT,CAAoBf,IAApB,EAA0B;SACxBA,KAAKC,IAAL,GACKwC,OADL,CACa,MADb,EACqB,GADrB,EAEKrI,MAFZ;;;;;;AAQF,AAAO,SAASiL,WAAT,CAAqBnN,KAArB,EAA4B;MAC3B+O,kBAAkBlG,WAAW7I,MAAM8H,IAAN,EAAX,CAAxB;;MAEMkH,WAAWhP,MAAMkB,IAAN,CAAW,GAAX,EAAgB4G,IAAhB,EAAjB;MACMmH,aAAapG,WAAWmG,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,CACbxP,CADa,EAEbyP,SAFa,EAGbC,WAHa,EAKb;MADAjB,SACA,yDADY,IACZ;;MACMkB,aAAaF,UAAU/N,MAAV,CAAiB;WAAQgO,YAAYE,OAAZ,CAAoBC,IAApB,MAA8B,CAAC,CAAvC;GAAjB,CAAnB;;;;;;;;UAEWA,IAHX;;UAIQxO,OAAO,MAAb;UACMd,QAAQ,OAAd;;UAEMuP,QAAQ9P,YAAUqB,IAAV,UAAmBwO,IAAnB,QAAd;;;;;UAKME,SACJD,MAAMzM,GAAN,CAAU,UAACjC,KAAD,EAAQf,IAAR;eAAiBL,EAAEK,IAAF,EAAQG,IAAR,CAAaD,KAAb,CAAjB;OAAV,EACMyP,OADN,GAEMtO,MAFN,CAEa;eAAQ0G,SAAS,EAAjB;OAFb,CADF;;;;;;UASI2H,OAAOvN,MAAP,KAAkB,CAAtB,EAAyB;YACnByN,kBAAJ;;;YAGIxB,SAAJ,EAAe;sBACDyB,UAAUH,OAAO,CAAP,CAAV,EAAqB/P,CAArB,CAAZ;SADF,MAEO;sBACO+P,OAAO,CAAP,CAAZ;;;;aAGKE;;;;;yBA5BQN,UAAnB,8HAA+B;;;;;;;;;;;;;;;;;;;;;;SAiCxB,IAAP;;;AC3CF,SAASQ,UAAT,CAAoB7P,KAApB,EAA2B8P,WAA3B,EAAwC;;;MAGlC9P,MAAMiC,QAAN,GAAiBC,MAAjB,GAA0B4N,WAA9B,EAA2C;WAClC,KAAP;;;MAGEC,cAAc/P,KAAd,CAAJ,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;;;;AAMF,AAAe,SAASgQ,oBAAT,CACbtQ,CADa,EAEbuQ,SAFa,EAKb;MAFAH,WAEA,yDAFc,CAEd;MADAI,QACA,yDADW,IACX;;;;;;yBACuBD,SAAvB,8HAAkC;UAAvBjN,QAAuB;;UAC1BwM,QAAQ9P,EAAEsD,QAAF,CAAd;;;;UAIIwM,MAAMtN,MAAN,KAAiB,CAArB,EAAwB;YAChBlC,QAAQN,EAAE8P,MAAM,CAAN,CAAF,CAAd;;YAEIK,WAAW7P,KAAX,EAAkB8P,WAAlB,CAAJ,EAAoC;cAC9BlO,gBAAJ;cACIsO,QAAJ,EAAc;sBACFlQ,MAAM8H,IAAN,EAAV;WADF,MAEO;sBACK9H,MAAMmQ,IAAN,EAAV;;;cAGEvO,OAAJ,EAAa;mBACJA,OAAP;;;;;;;;;;;;;;;;;;;;SAMD,IAAP;;;AChDF;AACA,AAAe,SAASgO,SAAT,CAAmB9H,IAAnB,EAAyBpI,CAAzB,EAA4B;;;MAGnC0Q,YAAY1Q,aAAWoI,IAAX,cAA0BA,IAA1B,EAAlB;SACOsI,cAAc,EAAd,GAAmBtI,IAAnB,GAA0BsI,SAAjC;;;ACLa,SAASL,aAAT,CAAuB/P,KAAvB,EAA8B;MACrCwC,UAAUxC,MAAMwC,OAAN,GAAgBkN,OAAhB,EAAhB;MACMW,gBAAgB7N,QAAQtB,IAAR,CAAa,UAACyI,MAAD,EAAY;QACvCxE,aAAgBwE,OAAOhJ,OAAP,CAAe2P,KAA/B,SAAwC3G,OAAOhJ,OAAP,CAAeuE,EAA7D;WACOC,WAAWtD,QAAX,CAAoB,SAApB,CAAP;GAFoB,CAAtB;;SAKOwO,kBAAkBE,SAAzB;;;ACPF;;;;AAIA,AAAe,SAASC,gBAAT,CAA0BxQ,KAA1B,EAAiC;SACvCA,MAAM8H,IAAN,GAAaC,IAAb,GAAoB7F,MAApB,IAA8B,GAArC;;;ACHa,SAASuO,WAAT,CAAqB/Q,CAArB,EAAwB;SAC9BA,EAAE2E,cAAF,EAAkBnC,MAAlB,GAA2B,CAAlC;;;ACHF;AACA,AAAO,IAAMwO,kBAAkB,wCAAxB;;;;AAIP,AAAO,IAAMC,eAAe,IAAI5S,MAAJ,CAAW,aAAX,EAA0B,GAA1B,CAArB;AACP,AAYA,AASA;AACA,AAAO,IAAM6S,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,OAAO/S,IAAP,CAAY,GAAZ,CAAlB;AACA,IAAMiT,aAAa,qCAAnB;AACA,IAAMC,aAAa,wCAAnB;AACA,AAAO,IAAMC,oBACX,IAAIpT,MAAJ,OAAekT,UAAf,WAA+BC,UAA/B,wBAA4DF,SAA5D,QAA0E,IAA1E,CADK;;;;;AAMP,AAAO,IAAMI,qBAAqB,gBAA3B;;AAEP,AAAO,IAAMC,oBACX,IAAItT,MAAJ,CAAW,2BAAX,EAAwC,GAAxC,CADK;;ACtDP;;AAEA,AAAe,SAASuT,WAAT,CAAqBC,MAArB,EAA6B;SACnCA,OAAOhH,OAAP,CAAemG,eAAf,EAAgC,IAAhC,EAAsC3I,IAAtC,EAAP;;;ACHa,SAASzG,OAAT,CAAekQ,YAAf,EAA6B;iBAC3BA,aAAazJ,IAAb,EAAf;MACI0J,SAASC,QAAT,CAAkBF,YAAlB,CAAJ,EAAqC;WAC5BA,YAAP;;;SAGK,IAAP;;;ACJF;;AAEA,AAAe,SAASG,QAAT,CAAkBC,GAAlB,QAA8B;MAALlS,CAAK,QAALA,CAAK;;;MAEvCkS,IAAI1P,MAAJ,GAAa,IAAb,IAAqB0P,IAAI1P,MAAJ,GAAa,CAAtC,EAAyC,OAAO,IAAP;;MAEnC2P,UAAUjC,UAAUgC,GAAV,EAAelS,CAAf,CAAhB;;;;MAIIiR,aAAaxR,IAAb,CAAkB0S,OAAlB,CAAJ,EAAgC,OAAO,IAAP;;SAEzBA,QAAQ9J,IAAR,EAAP;;;ACfF;;;;AAIA,AAOA,AAAO,SAAS+J,eAAT,CAAyBC,UAAzB,EAAqC;SACnC,CAACA,WAAWrJ,KAAX,CAAiByI,iBAAjB,KAAuC,EAAxC,EACWnT,IADX,CACgB,GADhB,EAEWuM,OAFX,CAEmBuG,qBAFnB,EAE0C,GAF1C,EAGWvG,OAHX,CAGmBsG,sBAHnB,EAG2C,UAH3C,EAIWtG,OAJX,CAImBqG,oBAJnB,EAIyC,IAJzC,EAKW7I,IALX,EAAP;;;;;AAUF,AAAe,SAASiK,kBAAT,CAA4BD,UAA5B,EAAwC;MACjDE,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;;;ACnBF;;AACA,AAAe,SAASC,gBAAT,CACbjL,OADa,QAQb;MALE3H,CAKF,QALEA,CAKF;mCAJE6S,kBAIF;MAJEA,kBAIF,yCAJuB,IAIvB;wBAHElE,KAGF;MAHEA,KAGF,8BAHU,EAGV;sBAFEjP,GAEF;MAFEA,GAEF,4BAFQ,EAER;;;;kBAGgBiI,OAAhB,EAAyB3H,CAAzB;;;cAGY2H,OAAZ,EAAqB3H,CAArB;;;;gBAIc2H,OAAd,EAAuB3H,CAAvB;;;;;aAKW2H,OAAX,EAAoB3H,CAApB;;;eAGa2H,OAAb,EAAsB3H,CAAtB,EAAyB2O,KAAzB;;;oBAGkBhH,OAAlB,EAA2B3H,CAA3B,EAA8BN,GAA9B;;;kBAGgBiI,OAAhB;;;;;YAKUA,OAAV,EAAmB3H,CAAnB,EAAsB6S,kBAAtB;;;cAGYlL,OAAZ,EAAqB3H,CAArB;;SAEO2H,OAAP;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AClDa,SAASmL,UAAT,CAAoBnE,KAApB,QAAuC;MAAVjP,GAAU,QAAVA,GAAU;MAALM,CAAK,QAALA,CAAK;;;;MAGhD0R,mBAAmBjS,IAAnB,CAAwBkP,KAAxB,CAAJ,EAAoC;YAC1BoE,kBAAkBpE,KAAlB,EAAyBjP,GAAzB,CAAR;;;;;MAKEiP,MAAMnM,MAAN,GAAe,GAAnB,EAAwB;;QAEhBwQ,KAAKhT,EAAE,IAAF,CAAX;QACIgT,GAAGxQ,MAAH,KAAc,CAAlB,EAAqB;cACXwQ,GAAG5K,IAAH,EAAR;;;;;SAKG8H,UAAUvB,KAAV,EAAiB3O,CAAjB,EAAoBqI,IAApB,EAAP;;;ACdF,SAAS4K,sBAAT,CAAgCC,UAAhC,EAA4C9K,IAA5C,EAAkD;;;;MAI5C8K,WAAW1Q,MAAX,IAAqB,CAAzB,EAA4B;;;;;UAIpB2Q,aAAaD,WAAWnL,MAAX,CAAkB,UAACC,GAAD,EAAMoL,SAAN,EAAoB;YACnDA,SAAJ,IAAiBpL,IAAIoL,SAAJ,IAAiBpL,IAAIoL,SAAJ,IAAiB,CAAlC,GAAsC,CAAvD;eACOpL,GAAP;OAFiB,EAGhB,EAHgB,CAAnB;;kCAMEhB,QAAQhG,OAAR,CAAgBmS,UAAhB,EACQpL,MADR,CACe,UAACC,GAAD,EAAMf,GAAN,EAAc;YAChBe,IAAI,CAAJ,IAASmL,WAAWlM,GAAX,CAAb,EAA8B;iBACrB,CAACA,GAAD,EAAMkM,WAAWlM,GAAX,CAAN,CAAP;;;eAGKe,GAAP;OANT,EAOU,CAAC,CAAD,EAAI,CAAJ,CAPV,CAVwB;;;;UASnBqL,OATmB;UASVC,SATU;;;;;;;UAuBtBA,aAAa,CAAb,IAAkBD,QAAQ7Q,MAAR,IAAkB,CAAxC,EAA2C;qBAC5B4F,KAAKuD,KAAL,CAAW0H,OAAX,CAAb;;;UAGIE,YAAY,CAACL,WAAW,CAAX,CAAD,EAAgBA,WAAWxJ,KAAX,CAAiB,CAAC,CAAlB,CAAhB,CAAlB;UACM8J,aAAaD,UAAUxL,MAAV,CAAiB,UAACC,GAAD,EAAMyL,GAAN;eAAczL,IAAIxF,MAAJ,GAAaiR,IAAIjR,MAAjB,GAA0BwF,GAA1B,GAAgCyL,GAA9C;OAAjB,EAAoE,EAApE,CAAnB;;UAEID,WAAWhR,MAAX,GAAoB,EAAxB,EAA4B;;aACnBgR;;;;;WAGFpL;;;;;;;SAGF,IAAP;;;AAGF,SAASsL,oBAAT,CAA8BR,UAA9B,EAA0CxT,GAA1C,EAA+C;;;;;;;mBAO5BC,IAAIC,KAAJ,CAAUF,GAAV,CAP4B;;MAOrC0M,IAPqC,cAOrCA,IAPqC;;MAQvCuH,cAAcvH,KAAKvB,OAAL,CAAa8G,iBAAb,EAAgC,EAAhC,CAApB;;MAEMiC,YAAYV,WAAW,CAAX,EAAclH,WAAd,GAA4BnB,OAA5B,CAAoC,GAApC,EAAyC,EAAzC,CAAlB;MACMgJ,iBAAiBC,MAAMC,WAAN,CAAkBH,SAAlB,EAA6BD,WAA7B,CAAvB;;MAEIE,iBAAiB,GAAjB,IAAwBD,UAAUpR,MAAV,GAAmB,CAA/C,EAAkD;WACzC0Q,WAAWxJ,KAAX,CAAiB,CAAjB,EAAoBpL,IAApB,CAAyB,EAAzB,CAAP;;;MAGI0V,UAAUd,WAAWxJ,KAAX,CAAiB,CAAC,CAAlB,EAAqB,CAArB,EAAwBsC,WAAxB,GAAsCnB,OAAtC,CAA8C,GAA9C,EAAmD,EAAnD,CAAhB;MACMoJ,eAAeH,MAAMC,WAAN,CAAkBC,OAAlB,EAA2BL,WAA3B,CAArB;;MAEIM,eAAe,GAAf,IAAsBD,QAAQxR,MAAR,IAAkB,CAA5C,EAA+C;WACtC0Q,WAAWxJ,KAAX,CAAiB,CAAjB,EAAoB,CAAC,CAArB,EAAwBpL,IAAxB,CAA6B,EAA7B,CAAP;;;SAGK,IAAP;;;;;AAKF,AAAe,SAASyU,iBAAT,CAA2BpE,KAA3B,EAA4C;MAAVjP,GAAU,yDAAJ,EAAI;;;;MAGnDwT,aAAavE,MAAMhD,KAAN,CAAY+F,kBAAZ,CAAnB;MACIwB,WAAW1Q,MAAX,KAAsB,CAA1B,EAA6B;WACpBmM,KAAP;;;MAGEuF,WAAWjB,uBAAuBC,UAAvB,EAAmCvE,KAAnC,CAAf;MACIuF,QAAJ,EAAc,OAAOA,QAAP;;aAEHR,qBAAqBR,UAArB,EAAiCxT,GAAjC,CAAX;MACIwU,QAAJ,EAAc,OAAOA,QAAP;;;;SAIPvF,KAAP;;;AC3FF,IAAMwF,WAAW;UACPvC,WADO;gBAEDwC,OAFC;OAGVnC,QAHU;iBAIAK,kBAJA;WAKN+B,gBALM;SAMRvB;CANT,CAUA,AAEA,AACA,AACA,AACA,AACA,AACA,AACA;;ACfA;;;;;;;;;;;AAWA,AAAe,SAASwB,eAAT,CAAyBtU,CAAzB,EAA4BuU,IAA5B,EAAkC;;;;;;;MAO3CA,KAAKlP,uBAAT,EAAkC;QAC5BA,wBAAwBrF,CAAxB,CAAJ;;;MAGE6G,oBAAoB7G,CAApB,CAAJ;MACIwK,aAAaxK,CAAb,EAAgBuU,KAAKrK,WAArB,CAAJ;MACMsK,gBAAgB3G,iBAAiB7N,CAAjB,CAAtB;;SAEOwU,aAAP;;;AC3BF,IAAMC,0BAA0B;eACjB;6BACc,IADd;iBAEE,IAFF;wBAGS;GAJQ;;;;;;;;;;;;;;;;;;;;;SAAA,yBA0BGF,IA1BH,EA0BS;QAA7BvU,CAA6B,QAA7BA,CAA6B;QAA1ByQ,IAA0B,QAA1BA,IAA0B;QAApB9B,KAAoB,QAApBA,KAAoB;QAAbjP,GAAa,QAAbA,GAAa;;wBACzB,KAAKgV,WAAjB,EAAiCH,IAAjC;;QAEIvU,KAAKoC,QAAQC,IAAR,CAAaoO,IAAb,CAAT;;;;QAIIpQ,OAAO,KAAKsU,cAAL,CAAoB3U,CAApB,EAAuB2O,KAAvB,EAA8BjP,GAA9B,EAAmC6U,IAAnC,CAAX;;QAEIzD,iBAAiBzQ,IAAjB,CAAJ,EAA4B;aACnB,KAAKuU,kBAAL,CAAwBvU,IAAxB,EAA8BL,CAA9B,CAAP;;;;;;;;;;2BAKgBgH,QAAQhG,OAAR,CAAgBuT,IAAhB,EAAsB7S,MAAtB,CAA6B;eAAK6S,KAAKM,CAAL,MAAY,IAAjB;OAA7B,CAAlB,8HAAuE;YAA5D5N,GAA4D;;aAChEA,GAAL,IAAY,KAAZ;YACI7E,QAAQC,IAAR,CAAaoO,IAAb,CAAJ;;eAEO,KAAKkE,cAAL,CAAoB3U,CAApB,EAAuB2O,KAAvB,EAA8BjP,GAA9B,EAAmC6U,IAAnC,CAAP;;YAEIzD,iBAAiBzQ,IAAjB,CAAJ,EAA4B;;;;;;;;;;;;;;;;;;;WAKvB,KAAKuU,kBAAL,CAAwBvU,IAAxB,EAA8BL,CAA9B,CAAP;GApD4B;;;;gBAAA,0BAwDfA,CAxDe,EAwDZ2O,KAxDY,EAwDLjP,GAxDK,EAwDA6U,IAxDA,EAwDM;WAC3BF,iBACGC,gBAAgBtU,CAAhB,EAAmBuU,IAAnB,CADH,EAEL;UAAA;0BAEsBA,KAAK1B,kBAF3B;kBAAA;;KAFK,CAAP;GAzD4B;;;;;;oBAAA,8BAsEXxS,IAtEW,EAsELL,CAtEK,EAsEF;QACtB,CAACK,IAAL,EAAW;aACF,IAAP;;;WAGKuK,gBAAgB5K,EAAEyQ,IAAF,CAAOpQ,IAAP,CAAhB,CAAP;;;;;;;CA3EJ,CAqFA;;AC9FA;;;;;;;AAOA,AAAO,IAAMyU,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;QAArBlV,CAAqB,QAArBA,CAAqB;QAAlBN,GAAkB,QAAlBA,GAAkB;QAAbyV,SAAa,QAAbA,SAAa;;;;QAGzBxG,cAAJ;;YAEQa,gBAAgBxP,CAAhB,EAAmB8U,sBAAnB,EAA2CK,SAA3C,CAAR;QACIxG,KAAJ,EAAW,OAAOmE,WAAWnE,KAAX,EAAkB,EAAEjP,QAAF,EAAOM,IAAP,EAAlB,CAAP;;;;YAIHsQ,qBAAqBtQ,CAArB,EAAwBgV,sBAAxB,CAAR;QACIrG,KAAJ,EAAW,OAAOmE,WAAWnE,KAAX,EAAkB,EAAEjP,QAAF,EAAOM,IAAP,EAAlB,CAAP;;;YAGHwP,gBAAgBxP,CAAhB,EAAmB+U,oBAAnB,EAAyCI,SAAzC,CAAR;QACIxG,KAAJ,EAAW,OAAOmE,WAAWnE,KAAX,EAAkB,EAAEjP,QAAF,EAAOM,IAAP,EAAlB,CAAP;;;YAGHsQ,qBAAqBtQ,CAArB,EAAwBiV,oBAAxB,CAAR;QACItG,KAAJ,EAAW,OAAOmE,WAAWnE,KAAX,EAAkB,EAAEjP,QAAF,EAAOM,IAAP,EAAlB,CAAP;;;WAGJ,EAAP;;CAvBJ,CA2BA;;ACxCA;;;;;;AAMA,AAAO,IAAMoV,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;QAAhBzV,CAAgB,QAAhBA,CAAgB;QAAbmV,SAAa,QAAbA,SAAa;;QACpBtD,eAAJ;;;;aAISrC,gBAAgBxP,CAAhB,EAAmBoV,gBAAnB,EAAqCD,SAArC,CAAT;QACItD,UAAUA,OAAOrP,MAAP,GAAgB6S,iBAA9B,EAAiD;aACxCzD,YAAYC,MAAZ,CAAP;;;;aAIOvB,qBAAqBtQ,CAArB,EAAwBsV,gBAAxB,EAA0C,CAA1C,CAAT;QACIzD,UAAUA,OAAOrP,MAAP,GAAgB6S,iBAA9B,EAAiD;aACxCzD,YAAYC,MAAZ,CAAP;;;;;;;;;;2BAK8B2D,mBAAhC,8HAAqD;;;;;YAAzClS,QAAyC;YAA/BoS,KAA+B;;YAC7CrV,OAAOL,EAAEsD,QAAF,CAAb;YACIjD,KAAKmC,MAAL,KAAgB,CAApB,EAAuB;cACf4F,OAAO/H,KAAK+H,IAAL,EAAb;cACIsN,MAAMjW,IAAN,CAAW2I,IAAX,CAAJ,EAAsB;mBACbwJ,YAAYxJ,IAAZ,CAAP;;;;;;;;;;;;;;;;;;;WAKC,IAAP;;CA7BJ,CAiCA;;AC9CA;;;;AAIA,AAAO,IAAMuN,2BAA2B,CACtC,wBADsC,EAEtC,aAFsC,EAGtC,SAHsC,EAItC,gBAJsC,EAKtC,WALsC,EAMtC,cANsC,EAOtC,UAPsC,EAQtC,UARsC,EAStC,SATsC,EAUtC,eAVsC,EAWtC,UAXsC,EAYtC,cAZsC,EAatC,qBAbsC,EActC,cAdsC,EAetC,SAfsC,EAgBtC,MAhBsC,CAAjC;;;;;AAsBP,AAAO,IAAMC,2BAA2B,CACtC,4BADsC,EAEtC,oBAFsC,EAGtC,0BAHsC,EAItC,kBAJsC,EAKtC,oBALsC,EAMtC,kBANsC,EAOtC,iBAPsC,EAQtC,aARsC,EAStC,eATsC,EAUtC,qBAVsC,EAWtC,mBAXsC,EAYtC,cAZsC,EAatC,aAbsC,EActC,YAdsC,EAetC,kBAfsC,EAgBtC,WAhBsC,EAiBtC,UAjBsC,CAAjC;;;;;AAuBP,IAAMC,kBAAkB,mDAAxB;AACA,AAAO,IAAMC,yBAAyB;;AAEpC,IAAIzX,MAAJ,CAAW,4BAAX,EAAyC,GAAzC,CAFoC;;;;AAMpC,IAAIA,MAAJ,CAAW,6BAAX,EAA0C,GAA1C,CANoC;;AAQpC,IAAIA,MAAJ,iBAAyBwX,eAAzB,kBAAuD,GAAvD,CARoC,CAA/B;;ACrCP,IAAME,gCAAgC;SAAA,yBACL;QAArB/V,CAAqB,QAArBA,CAAqB;QAAlBN,GAAkB,QAAlBA,GAAkB;QAAbyV,SAAa,QAAbA,SAAa;;QACzBa,sBAAJ;;;;oBAIgBxG,gBAAgBxP,CAAhB,EAAmB2V,wBAAnB,EAA6CR,SAA7C,EAAwD,KAAxD,CAAhB;QACIa,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;;oBAIH1F,qBAAqBtQ,CAArB,EAAwB4V,wBAAxB,CAAhB;QACII,aAAJ,EAAmB,OAAO1D,mBAAmB0D,aAAnB,CAAP;;;oBAGHlL,eAAepL,GAAf,EAAoBoW,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,IAAIhY,MAAJ,CAAW+X,8BAA8B9X,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAMgY,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,IAAIlY,MAAJ,CAAWiY,8BAA8BhY,IAA9B,CAAmC,GAAnC,CAAX,EAAoD,GAApD,CAAzC;;AAEP,AAAO,IAAMkY,SAAS,gBAAf;AACP,AAAO,IAAMC,SAAS,kBAAf;;AC3CP,SAASC,MAAT,CAAgBpW,KAAhB,EAAuB;UACXA,MAAME,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuCF,MAAME,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;;AAIF,AAAO,SAASmW,aAAT,CAAuBjX,GAAvB,EAA4B;QAC3BA,IAAI2I,IAAJ,EAAN;MACIO,QAAQ,CAAZ;;MAEIyN,iCAAiC5W,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;MAGE6W,iCAAiC9W,IAAjC,CAAsCC,GAAtC,CAAJ,EAAgD;aACrC,EAAT;;;;;MAKE8W,OAAO/W,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;MAGE+W,OAAOhX,IAAP,CAAYC,GAAZ,CAAJ,EAAsB;aACX,EAAT;;;;;SAKKkJ,KAAP;;;;AAIF,AAAO,SAASgO,SAAT,CAAmBzP,IAAnB,EAAyB;MAC1BA,KAAK3G,IAAL,CAAU,KAAV,CAAJ,EAAsB;WACb,CAAP;;;SAGK,CAAP;;;;;AAKF,AAAO,SAASqW,cAAT,CAAwB1P,IAAxB,EAA8B;MAC/ByB,QAAQ,CAAZ;MACMkO,aAAa3P,KAAKrE,OAAL,CAAa,QAAb,EAAuBgL,KAAvB,EAAnB;;MAEIgJ,WAAWtU,MAAX,KAAsB,CAA1B,EAA6B;aAClB,EAAT;;;MAGI8H,UAAUnD,KAAK8C,MAAL,EAAhB;MACI8M,iBAAJ;MACIzM,QAAQ9H,MAAR,KAAmB,CAAvB,EAA0B;eACb8H,QAAQL,MAAR,EAAX;;;GAGDK,OAAD,EAAUyM,QAAV,EAAoB7V,OAApB,CAA4B,UAACZ,KAAD,EAAW;QACjCgE,iBAAe7E,IAAf,CAAoBiX,OAAOpW,KAAP,CAApB,CAAJ,EAAwC;eAC7B,EAAT;;GAFJ;;SAMOsI,KAAP;;;;;AAKF,AAAO,SAASoO,cAAT,CAAwB7P,IAAxB,EAA8B;MAC/ByB,QAAQ,CAAZ;MACMqO,WAAW9P,KAAKrB,IAAL,EAAjB;MACMG,UAAUgR,SAASzY,GAAT,CAAa,CAAb,CAAhB;;MAEIyH,WAAWA,QAAQtD,OAAR,KAAoB,YAAnC,EAAiD;aACtC,EAAT;;;MAGE2B,iBAAe7E,IAAf,CAAoBiX,OAAOO,QAAP,CAApB,CAAJ,EAA2C;aAChC,EAAT;;;SAGKrO,KAAP;;;AAGF,AAAO,SAASsO,iBAAT,CAA2B/P,IAA3B,EAAiC;MAClCyB,QAAQ,CAAZ;;MAEMtB,QAAQwB,WAAW3B,KAAK3G,IAAL,CAAU,OAAV,CAAX,CAAd;MACM4G,SAAS0B,WAAW3B,KAAK3G,IAAL,CAAU,QAAV,CAAX,CAAf;MACM2W,MAAMhQ,KAAK3G,IAAL,CAAU,KAAV,CAAZ;;;MAGI8G,SAASA,SAAS,EAAtB,EAA0B;aACf,EAAT;;;;MAIEF,UAAUA,UAAU,EAAxB,EAA4B;aACjB,EAAT;;;MAGEE,SAASF,MAAT,IAAmB,CAAC+P,IAAIhV,QAAJ,CAAa,QAAb,CAAxB,EAAgD;QACxCiV,OAAO9P,QAAQF,MAArB;QACIgQ,OAAO,IAAX,EAAiB;;eACN,GAAT;KADF,MAEO;eACI9N,KAAK+N,KAAL,CAAWD,OAAO,IAAlB,CAAT;;;;SAIGxO,KAAP;;;AAGF,AAAO,SAAS0O,eAAT,CAAyBC,KAAzB,EAAgCnW,KAAhC,EAAuC;SACpCmW,MAAM/U,MAAN,GAAe,CAAhB,GAAqBpB,KAA5B;;;ACxGF;;;;;;;;AAQA,IAAMoW,+BAA+B;SAAA,yBACA;QAAzBxX,CAAyB,QAAzBA,CAAyB;QAAtBkC,OAAsB,QAAtBA,OAAsB;QAAbiT,SAAa,QAAbA,SAAa;;QAC7BsC,iBAAJ;;;;;;QAMMC,WACJlI,gBACExP,CADF,EAEEkW,wBAFF,EAGEf,SAHF,EAIE,KAJF,CADF;;QAQIuC,QAAJ,EAAc;iBACDtD,QAAWsD,QAAX,CAAX;;UAEID,QAAJ,EAAc,OAAOA,QAAP;;;;;;QAMVE,OAAO3X,EAAE,KAAF,EAASkC,OAAT,EAAkB8N,OAAlB,EAAb;QACM4H,YAAY,EAAlB;;SAEK1W,OAAL,CAAa,UAACH,GAAD,EAAMK,KAAN,EAAgB;UACrB+F,OAAOnH,EAAEe,GAAF,CAAb;UACMoW,MAAMhQ,KAAK3G,IAAL,CAAU,KAAV,CAAZ;;UAEI,CAAC2W,GAAL,EAAU;;UAENvO,QAAQ+N,cAAcQ,GAAd,CAAZ;eACSP,UAAUzP,IAAV,CAAT;eACS0P,eAAe1P,IAAf,CAAT;eACS6P,eAAe7P,IAAf,CAAT;eACS+P,kBAAkB/P,IAAlB,CAAT;eACSmQ,gBAAgBK,IAAhB,EAAsBvW,KAAtB,CAAT;;gBAEU+V,GAAV,IAAiBvO,KAAjB;KAbF;;gCAiBE5B,QAAQhG,OAAR,CAAgB4W,SAAhB,EAA2B7P,MAA3B,CAAkC,UAACC,GAAD,EAAMf,GAAN;aAChC2Q,UAAU3Q,GAAV,IAAiBe,IAAI,CAAJ,CAAjB,GAA0B,CAACf,GAAD,EAAM2Q,UAAU3Q,GAAV,CAAN,CAA1B,GAAkDe,GADlB;KAAlC,EAEE,CAAC,IAAD,EAAO,CAAP,CAFF,CA5C+B;;;;QA2C1B6P,MA3C0B;QA2ClB7K,QA3CkB;;;QAgD7BA,WAAW,CAAf,EAAkB;iBACLoH,QAAWyD,MAAX,CAAX;;UAEIJ,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;2BAKOtB,wBAAvB,8HAAiD;YAAtC7S,QAAsC;;YACzChD,QAAQN,EAAEsD,QAAF,EAAYwK,KAAZ,EAAd;YACMqJ,MAAM7W,MAAME,IAAN,CAAW,KAAX,CAAZ;YACI2W,GAAJ,EAAS;qBACI/C,QAAW+C,GAAX,CAAX;cACIM,QAAJ,EAAc,OAAOA,QAAP;;;YAGVK,OAAOxX,MAAME,IAAN,CAAW,MAAX,CAAb;YACIsX,IAAJ,EAAU;qBACG1D,QAAW0D,IAAX,CAAX;cACIL,QAAJ,EAAc,OAAOA,QAAP;;;YAGVlX,QAAQD,MAAME,IAAN,CAAW,OAAX,CAAd;YACID,KAAJ,EAAW;qBACE6T,QAAW7T,KAAX,CAAX;cACIkX,QAAJ,EAAc,OAAOA,QAAP;;;;;;;;;;;;;;;;;;WAIX,IAAP;;CA9EJ;;AAkFA,AAEA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AC7Ge,SAASM,eAAT,CAAyBnP,KAAzB,EAAgCoP,UAAhC,EAA4CF,IAA5C,EAAkD;;;;;;MAM3DlP,QAAQ,CAAZ,EAAe;QACPqP,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;WACOzP,QAAQ0P,YAAf;;;SAGK,CAAP;;;ACnBa,SAASC,aAAT,CAAuBjJ,QAAvB,EAAiC7D,OAAjC,EAA0C;;;;;MAKnD7C,QAAQ,CAAZ;;MAEI0C,YAAY7L,IAAZ,CAAiB6P,SAASjH,IAAT,EAAjB,CAAJ,EAAuC;QAC/BmQ,gBAAgBnR,SAASiI,QAAT,EAAmB,EAAnB,CAAtB;;;;QAIIkJ,gBAAgB,CAApB,EAAuB;cACb,CAAC,EAAT;KADF,MAEO;cACGlP,KAAKE,GAAL,CAAS,CAAT,EAAY,KAAKgP,aAAjB,CAAR;;;;;;QAME/M,WAAWA,WAAW+M,aAA1B,EAAyC;eAC9B,EAAT;;;;SAIG5P,KAAP;;;AC5Ba,SAAS6P,eAAT,CAAyBhN,OAAzB,EAAkCiN,IAAlC,EAAwC;;;;MAIjDjN,WAAW,CAACiN,IAAhB,EAAsB;WACb,EAAP;;;SAGK,CAAP;;;ACRK,IAAMC,aAAW,IAAjB;;;;AAIP,AAAO,IAAM/T,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,IAAIxG,MAAJ,CAAWuG,wBAAsBtG,IAAtB,CAA2B,GAA3B,CAAX,EAA4C,GAA5C,CAAjC;;;;;AAKP,AAAO,IAAMsa,sBAAoB,IAAIva,MAAJ,CAAW,4CAAX,EAAyD,GAAzD,CAA1B;;;;AAIP,AAAO,IAAMwa,qBAAmB,IAAIxa,MAAJ,CAAW,kBAAX,EAA+B,GAA/B,CAAzB;;;;AAIP,AAAO,IAAMya,sBAAoB,IAAIza,MAAJ,CAAW,yBAAX,EAAsC,GAAtC,CAA1B,CAEP;;ACjCe,SAAS0a,oBAAT,CAA8BjB,IAA9B,EAAoC;;MAE7CjT,2BAAyBpF,IAAzB,CAA8BqY,IAA9B,CAAJ,EAAyC;WAChC,CAAC,EAAR;;;SAGK,CAAP;;;eCRuBkB;;AAAzB,AAAe,SAAUA,KAAV;MAAgBC,KAAhB,yDAAwB,CAAxB;MAA2BxF,GAA3B,yDAAiC,CAAjC;;;;;gBACNwF,SAASxF,GADH;;;;;;iBAELwF,SAAS,CAFJ;;;;;;;;;;;;;;ACQf,SAASC,SAAT,CAAiBC,KAAjB,EAAwB;UACZA,MAAM3Y,IAAN,CAAW,OAAX,KAAuB,EAAjC,WAAuC2Y,MAAM3Y,IAAN,CAAW,IAAX,KAAoB,EAA3D;;;AAGF,AAAe,SAASqW,gBAAT,CAAwBsC,KAAxB,EAA+B;;;;MAIxC7O,UAAU6O,MAAMlP,MAAN,EAAd;MACImP,gBAAgB,KAApB;MACIC,gBAAgB,KAApB;MACIzQ,QAAQ,CAAZ;;QAEM3I,IAAN,CAAW+Y,MAAM,CAAN,EAAS,CAAT,CAAX,EAAwB9X,OAAxB,CAAgC,YAAM;QAChCoJ,QAAQ9H,MAAR,KAAmB,CAAvB,EAA0B;;;;QAIpB8W,aAAaJ,UAAQ5O,OAAR,EAAiB,GAAjB,CAAnB;;;;QAII,CAAC8O,aAAD,IAAkBtU,QAAQrF,IAAR,CAAa6Z,UAAb,CAAtB,EAAgD;sBAC9B,IAAhB;eACS,EAAT;;;;;;QAME,CAACD,aAAD,IAAkB3U,kBAAkBjF,IAAlB,CAAuB6Z,UAAvB,CAAlB,IACEzU,2BAAyBpF,IAAzB,CAA8B6Z,UAA9B,CADN,EACiD;UAC3C,CAAC9U,kBAAkB/E,IAAlB,CAAuB6Z,UAAvB,CAAL,EAAyC;wBACvB,IAAhB;iBACS,EAAT;;;;cAIMhP,QAAQL,MAAR,EAAV;GAzBF;;SA4BOrB,KAAP;;;AC/Ca,SAAS2Q,aAAT,CAAuBC,QAAvB,EAAiC;;;MAG1CV,oBAAkBrZ,IAAlB,CAAuB+Z,QAAvB,CAAJ,EAAsC;WAC7B,CAAC,GAAR;;;SAGK,CAAP;;;ACFa,SAASC,WAAT,CACb3B,IADa,EAEbE,UAFa,EAGb0B,OAHa,EAIb7Z,SAJa,EAKbyP,QALa,EAMbqK,YANa,EAOb;;MAEIA,aAAanY,IAAb,CAAkB;WAAOsW,SAASpY,GAAhB;GAAlB,MAA2CmR,SAA/C,EAA0D;WACjD,KAAP;;;;;MAKE,CAACiH,IAAD,IAASA,SAASE,UAAlB,IAAgCF,SAAS4B,OAA7C,EAAsD;WAC7C,KAAP;;;MAGME,QAZR,GAYqB/Z,SAZrB,CAYQ+Z,QAZR;;mBAa+Bja,IAAIC,KAAJ,CAAUkY,IAAV,CAb/B;;MAakB+B,QAblB,cAaQD,QAbR;;;;MAgBIC,aAAaD,QAAjB,EAA2B;WAClB,KAAP;;;;;MAKIE,WAAWhC,KAAKjN,OAAL,CAAa6O,OAAb,EAAsB,EAAtB,CAAjB;MACI,CAACf,WAASlZ,IAAT,CAAcqa,QAAd,CAAL,EAA8B;WACrB,KAAP;;;;;MAKEjV,2BAAyBpF,IAAzB,CAA8B6P,QAA9B,CAAJ,EAA6C;WACpC,KAAP;;;;MAIEA,SAAS9M,MAAT,GAAkB,EAAtB,EAA0B;WACjB,KAAP;;;SAGK,IAAP;;;ACpDa,SAASuX,YAAT,CAAsBjC,IAAtB,EAA4BkC,SAA5B,EAAuC;;;;;MAKhD,CAACA,UAAUva,IAAV,CAAeqY,IAAf,CAAL,EAA2B;WAClB,CAAC,EAAR;;;SAGK,CAAP;;;ACPa,SAASmC,iBAAT,CAA2BT,QAA3B,EAAqC;;MAE9CZ,oBAAkBnZ,IAAlB,CAAuB+Z,QAAvB,CAAJ,EAAsC;WAC7B,EAAP;;;SAGK,CAAP;;;ACHa,SAASU,aAAT,CAAuBV,QAAvB,EAAiC;;MAE1CX,mBAAiBpZ,IAAjB,CAAsB+Z,QAAtB,CAAJ,EAAqC;;;;;QAK/BZ,oBAAkBnZ,IAAlB,CAAuB+Z,QAAvB,CAAJ,EAAsC;aAC7B,CAAC,EAAR;;;;SAIG,CAAP;;;ACKK,SAASW,aAAT,CAAuBT,OAAvB,EAAgC;SAC9B,IAAIrb,MAAJ,OAAeqb,OAAf,EAA0B,GAA1B,CAAP;;;AAGF,SAASR,OAAT,CAAiBC,KAAjB,EAAwB7J,QAAxB,EAAkC;UACtBA,YAAY6J,MAAM/Q,IAAN,EAAtB,WAAsC+Q,MAAM3Y,IAAN,CAAW,OAAX,KAAuB,EAA7D,WAAmE2Y,MAAM3Y,IAAN,CAAW,IAAX,KAAoB,EAAvF;;;AAGF,AAAe,SAAS4Z,UAAT,OAOZ;MANDC,KAMC,QANDA,KAMC;MALDrC,UAKC,QALDA,UAKC;MAJD0B,OAIC,QAJDA,OAIC;MAHD7Z,SAGC,QAHDA,SAGC;MAFDG,CAEC,QAFDA,CAEC;+BADD2Z,YACC;MADDA,YACC,qCADc,EACd;;cACW9Z,aAAaF,IAAIC,KAAJ,CAAUoY,UAAV,CAAzB;MACMgC,YAAYG,cAAcT,OAAd,CAAlB;MACMhB,OAAO3H,YAAY/Q,CAAZ,CAAb;;;;;;;;;MASMsa,cAAcD,MAAMtS,MAAN,CAAa,UAACwS,aAAD,EAAgBC,IAAhB,EAAyB;;;;QAIlD1C,OAAOpM,aAAa8O,KAAKvZ,OAAL,CAAa6W,IAA1B,CAAb;QACMqB,QAAQnZ,EAAEwa,IAAF,CAAd;QACMlL,WAAW6J,MAAM/Q,IAAN,EAAjB;;QAEI,CAACqR,YAAY3B,IAAZ,EAAkBE,UAAlB,EAA8B0B,OAA9B,EAAuC7Z,SAAvC,EAAkDyP,QAAlD,EAA4DqK,YAA5D,CAAL,EAAgF;aACvEY,aAAP;;;;QAIE,CAACA,cAAczC,IAAd,CAAL,EAA0B;oBACVA,IAAd,IAAsB;eACb,CADa;0BAAA;;OAAtB;KADF,MAMO;oBACSA,IAAd,EAAoBxI,QAApB,GAAkCiL,cAAczC,IAAd,EAAoBxI,QAAtD,SAAkEA,QAAlE;;;QAGImL,eAAeF,cAAczC,IAAd,CAArB;QACM0B,WAAWN,QAAQC,KAAR,EAAe7J,QAAf,CAAjB;QACM7D,UAAUF,eAAeuM,IAAf,CAAhB;;QAEIlP,QAAQmR,aAAajC,IAAb,EAAmBkC,SAAnB,CAAZ;aACSC,kBAAkBT,QAAlB,CAAT;aACSU,cAAcV,QAAd,CAAT;aACSD,cAAcC,QAAd,CAAT;aACS3C,iBAAesC,KAAf,CAAT;aACSJ,qBAAqBjB,IAArB,CAAT;aACSW,gBAAgBhN,OAAhB,EAAyBiN,IAAzB,CAAT;aACSH,cAAcjJ,QAAd,EAAwB7D,OAAxB,CAAT;aACSsM,gBAAgBnP,KAAhB,EAAuBoP,UAAvB,EAAmCF,IAAnC,CAAT;;iBAEalP,KAAb,GAAqBA,KAArB;;WAEO2R,aAAP;GAvCkB,EAwCjB,EAxCiB,CAApB;;SA0COvT,QAAQhG,OAAR,CAAgBsZ,WAAhB,EAA6B9X,MAA7B,KAAwC,CAAxC,GAA4C,IAA5C,GAAmD8X,WAA1D;;;AClFF;;AAEA,IAAMI,8BAA8B;SAAA,yBACgB;QAAxC1a,CAAwC,QAAxCA,CAAwC;QAArCN,GAAqC,QAArCA,GAAqC;QAAhCG,SAAgC,QAAhCA,SAAgC;iCAArB8Z,YAAqB;QAArBA,YAAqB,qCAAN,EAAM;;gBACpC9Z,aAAaF,IAAIC,KAAJ,CAAUF,GAAV,CAAzB;;QAEMsY,aAAatM,aAAahM,GAAb,CAAnB;QACMga,UAAUzN,eAAevM,GAAf,EAAoBG,SAApB,CAAhB;;QAEMwa,QAAQra,EAAE,SAAF,EAAagQ,OAAb,EAAd;;QAEM2K,cAAcP,WAAW;kBAAA;4BAAA;sBAAA;0BAAA;UAAA;;KAAX,CAApB;;;QAUI,CAACO,WAAL,EAAkB,OAAO,IAAP;;;;QAIZC,UAAU5T,QAAQhG,OAAR,CAAgB2Z,WAAhB,EAA6B5S,MAA7B,CAAoC,UAACC,GAAD,EAAMwS,IAAN,EAAe;UAC3DK,aAAaF,YAAYH,IAAZ,CAAnB;aACOK,WAAWjS,KAAX,GAAmBZ,IAAIY,KAAvB,GAA+BiS,UAA/B,GAA4C7S,GAAnD;KAFc,EAGb,EAAEY,OAAO,CAAC,GAAV,EAHa,CAAhB;;;;QAOIgS,QAAQhS,KAAR,IAAiB,EAArB,EAAyB;aAChBgS,QAAQ9C,IAAf;;;WAGK,IAAP;;CAlCJ,CAuCA;;AClDO,IAAMgD,2BAA2B,CACtC,QADsC,CAAjC;;ACKP,SAASC,WAAT,CAAqBrb,GAArB,EAA0B;MAClBG,YAAYF,IAAIC,KAAJ,CAAUF,GAAV,CAAlB;MACQka,QAFgB,GAEH/Z,SAFG,CAEhB+Z,QAFgB;;SAGjBA,QAAP;;;AAGF,SAAS3X,MAAT,CAAgBvC,GAAhB,EAAqB;SACZ;YAAA;YAEGqb,YAAYrb,GAAZ;GAFV;;;AAMF,IAAMsb,sBAAsB;SAAA,yBACK;QAArBhb,CAAqB,QAArBA,CAAqB;QAAlBN,GAAkB,QAAlBA,GAAkB;QAAbyV,SAAa,QAAbA,SAAa;;QACvB8F,aAAajb,EAAE,qBAAF,CAAnB;QACIib,WAAWzY,MAAX,KAAsB,CAA1B,EAA6B;UACrBsV,OAAOmD,WAAWza,IAAX,CAAgB,MAAhB,CAAb;UACIsX,IAAJ,EAAU;eACD7V,OAAO6V,IAAP,CAAP;;;;QAIEoD,UAAU1L,gBAAgBxP,CAAhB,EAAmB8a,wBAAnB,EAA6C3F,SAA7C,CAAhB;QACI+F,OAAJ,EAAa;aACJjZ,OAAOiZ,OAAP,CAAP;;;WAGKjZ,OAAOvC,GAAP,CAAP;;CAfJ,CAoBA;;ACtCO,IAAMyb,yBAAyB,CACpC,gBADoC,EAEpC,qBAFoC,CAA/B;;ACSA,SAASvZ,OAAT,CAAeM,OAAf,EAAwBlC,CAAxB,EAA4C;MAAjBob,SAAiB,yDAAL,GAAK;;YACvClZ,QAAQ2I,OAAR,CAAgB,UAAhB,EAA4B,GAA5B,EAAiCxC,IAAjC,EAAV;SACOgT,UAAUnZ,OAAV,EAAmBkZ,SAAnB,EAA8B,EAAEE,SAAS,UAAX,EAA9B,CAAP;;;AAGF,IAAMC,0BAA0B;SAAA,yBACK;QAAzBvb,CAAyB,QAAzBA,CAAyB;QAAtBkC,OAAsB,QAAtBA,OAAsB;QAAbiT,SAAa,QAAbA,SAAa;;QAC3BqG,UAAUhM,gBAAgBxP,CAAhB,EAAmBmb,sBAAnB,EAA2ChG,SAA3C,CAAhB;QACIqG,OAAJ,EAAa;aACJ5Z,QAAMsO,UAAUsL,OAAV,EAAmBxb,CAAnB,CAAN,CAAP;;;QAGIob,YAAY,GAAlB;QACMK,eAAevZ,QAAQwH,KAAR,CAAc,CAAd,EAAiB0R,YAAY,CAA7B,CAArB;WACOxZ,QAAM5B,EAAEyb,YAAF,EAAgBrT,IAAhB,EAAN,EAA8BpI,CAA9B,EAAiCob,SAAjC,CAAP;;CATJ,CAaA;;ACvBA,IAAMM,4BAA4B;SAAA,yBACX;QAAXxZ,OAAW,QAAXA,OAAW;;QACblC,IAAIoC,QAAQC,IAAR,CAAaH,OAAb,CAAV;;QAEMkG,OAAOwC,gBAAgB5K,EAAE,KAAF,EAAS8N,KAAT,GAAiB1F,IAAjB,EAAhB,CAAb;WACOA,KAAKuD,KAAL,CAAW,IAAX,EAAiBnJ,MAAxB;;CALJ,CASA;;ACAA,IAAMmZ,mBAAmB;;UAEf,GAFe;SAGhBzG,sBAAsB0G,OAHN;kBAIP7F,8BAA8B6F,OAJvB;UAKfnG,uBAAuBmG,OALR;WAMdnH,wBAAwBmH,OAAxB,CAAgCC,IAAhC,CAAqCpH,uBAArC,CANc;kBAOP+C,6BAA6BoE,OAPtB;OAQlB3F,oBAAoB2F,OARF;iBASRlB,4BAA4BkB,OATpB;kBAUPZ,oBAAoBY,OAVb;WAWdL,wBAAwBK,OAXV;cAYXF,0BAA0BE,OAZf;;SAAA,mBAcfnd,OAde,EAcN;QACPgS,IADO,GACEhS,OADF,CACPgS,IADO;;;QAGXA,IAAJ,EAAU;UACFzQ,IAAIoC,QAAQC,IAAR,CAAaoO,IAAb,CAAV;cACQzQ,CAAR,GAAYA,CAAZ;;;QAGI2O,QAAQ,KAAKA,KAAL,CAAWlQ,OAAX,CAAd;QACMqd,iBAAiB,KAAKA,cAAL,CAAoBrd,OAApB,CAAvB;QACMoT,SAAS,KAAKA,MAAL,CAAYpT,OAAZ,CAAf;QACMyD,UAAU,KAAKA,OAAL,cAAkBzD,OAAlB,IAA2BkQ,YAA3B,IAAhB;QACMoN,iBAAiB,KAAKA,cAAL,cAAyBtd,OAAzB,IAAkCyD,gBAAlC,IAAvB;QACMgQ,MAAM,KAAKA,GAAL,cAAczT,OAAd,IAAuByD,gBAAvB,IAAZ;QACM8Z,gBAAgB,KAAKA,aAAL,CAAmBvd,OAAnB,CAAtB;QACM+c,UAAU,KAAKA,OAAL,cAAkB/c,OAAlB,IAA2ByD,gBAA3B,IAAhB;QACM+Z,aAAa,KAAKA,UAAL,cAAqBxd,OAArB,IAA8ByD,gBAA9B,IAAnB;;0BACwB,KAAKga,cAAL,CAAoBzd,OAApB,CAjBT;;QAiBPiB,GAjBO,mBAiBPA,GAjBO;QAiBFyc,MAjBE,mBAiBFA,MAjBE;;;WAmBR;kBAAA;oBAAA;sBAGWL,kBAAkB,IAH7B;cAAA;oCAAA;sBAAA;kCAAA;cAAA;oBAAA;sBAAA;;KAAP;;CAjCJ,CAiDA;;ACzDe,SAASM,YAAT,CAAsB1c,GAAtB,EAA2B;MAClCG,YAAYF,IAAIC,KAAJ,CAAUF,GAAV,CAAlB;MACQka,QAFgC,GAEnB/Z,SAFmB,CAEhC+Z,QAFgC;;MAGlCyC,aAAazC,SAASjO,KAAT,CAAe,GAAf,EAAoBjC,KAApB,CAA0B,CAAC,CAA3B,EAA8BpL,IAA9B,CAAmC,GAAnC,CAAnB;;SAEO0E,WAAW4W,QAAX,KAAwB5W,WAAWqZ,UAAX,CAAxB,IAAkDV,gBAAzD;;;ACVK,IAAMW,UAAU,cAAhB;;ACOP;AACA,AAAO,SAASC,gBAAT,CAA0BrN,QAA1B,EAAoClP,CAApC,QAAkD;MAAT4B,KAAS,QAATA,KAAS;;MACnD,CAACA,KAAL,EAAY,OAAO,IAAP;;IAEVA,MAAMtD,IAAN,CAAW,GAAX,CAAF,EAAmB4Q,QAAnB,EAA6BvN,MAA7B;;SAEOuN,QAAP;;;;AAIF,AAAO,SAASsN,iBAAT,CAA2BtN,QAA3B,EAAqClP,CAArC,SAAwD;MAAdyc,UAAc,SAAdA,UAAc;;MACzD,CAACA,UAAL,EAAiB,OAAO,IAAP;;UAETzb,OAAR,CAAgByb,UAAhB,EAA4Bvb,OAA5B,CAAoC,UAAC+F,GAAD,EAAS;QACrCyV,WAAW1c,EAAEiH,GAAF,EAAOiI,QAAP,CAAjB;QACM3O,QAAQkc,WAAWxV,GAAX,CAAd;;;QAGI,OAAO1G,KAAP,KAAiB,QAArB,EAA+B;eACpBJ,IAAT,CAAc,UAACiB,KAAD,EAAQf,IAAR,EAAiB;sBACfL,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0Byc,WAAWxV,GAAX,CAA1B;OADF;KADF,MAIO,IAAI,OAAO1G,KAAP,KAAiB,UAArB,EAAiC;;eAE7BJ,IAAT,CAAc,UAACiB,KAAD,EAAQf,IAAR,EAAiB;YACvB4B,SAAS1B,MAAMP,EAAEK,IAAF,CAAN,EAAeL,CAAf,CAAf;;YAEI,OAAOiC,MAAP,KAAkB,QAAtB,EAAgC;wBAChBjC,EAAEK,IAAF,CAAd,EAAuBL,CAAvB,EAA0BiC,MAA1B;;OAJJ;;GAXJ;;SAqBOiN,QAAP;;;AAGF,AAAO,SAASyN,MAAT,CAAgBpI,IAAhB,EAAsB;MACnBvU,CADmB,GAC8BuU,IAD9B,CACnBvU,CADmB;MAChBqB,IADgB,GAC8BkT,IAD9B,CAChBlT,IADgB;MACVub,cADU,GAC8BrI,IAD9B,CACVqI,cADU;0BAC8BrI,IAD9B,CACMsI,WADN;MACMA,WADN,qCACoB,KADpB;;;MAGvB,CAACD,cAAL,EAAqB,OAAO,IAAP;;;;MAIjB,OAAOA,cAAP,KAA0B,QAA9B,EAAwC,OAAOA,cAAP;;MAEhCrM,SATmB,GASLqM,cATK,CASnBrM,SATmB;;;MAWrBuM,mBAAmBvM,UAAU/O,IAAV,CAAe;WAAYxB,EAAEsD,QAAF,EAAYd,MAAZ,KAAuB,CAAvB,IAA4BxC,EAAEsD,QAAF,EAAY8E,IAAZ,GAAmBC,IAAnB,OAA8B,EAAtE;GAAf,CAAzB;;MAEI,CAACyU,gBAAL,EAAuB,OAAO,IAAP;;;;;;;;MAQnBD,WAAJ,EAAiB;QACX3N,WAAWlP,EAAE8c,gBAAF,CAAf;;;aAGSC,IAAT,CAAc/c,EAAE,aAAF,CAAd;eACWkP,SAASjF,MAAT,EAAX;;eAEWuS,kBAAkBtN,QAAlB,EAA4BlP,CAA5B,EAA+B4c,cAA/B,CAAX;eACWL,iBAAiBrN,QAAjB,EAA2BlP,CAA3B,EAA8B4c,cAA9B,CAAX;;eAEWzI,SAAS9S,IAAT,EAAe6N,QAAf,EAAyBqF,IAAzB,CAAX;;WAEOvU,EAAEyQ,IAAF,CAAOvB,QAAP,CAAP;;;;MAII1O,OAAOsc,iBAAiB9T,KAAjB,CAAuBsT,OAAvB,CAAb;MACIra,eAAJ;;MAEIzB,IAAJ,EAAU;aACCR,EAAE8c,gBAAF,EAAoBtc,IAApB,CAAyBA,KAAK,CAAL,CAAzB,CAAT;GADF,MAEO;;aAEIR,EAAE8c,gBAAF,EAAoB1U,IAApB,EAAT;;SAEK+L,SAAS9S,IAAT,EAAeY,MAAf,EAAuBsS,IAAvB,CAAP;;;AAGF,SAASyI,aAAT,CAAuBzI,IAAvB,EAA6B;MACnBlT,IADmB,GACCkT,IADD,CACnBlT,IADmB;MACb4b,SADa,GACC1I,IADD,CACb0I,SADa;;;;;SAKpBN,oBAAYpI,IAAZ,IAAkBqI,gBAAgBK,UAAU5b,IAAV,CAAlC,QACLsa,iBAAiBta,IAAjB,EAAuBkT,IAAvB,CADF;;;AAIF,IAAM2I,gBAAgB;SAAA,qBACwB;QAApCD,SAAoC,yDAAxBtB,gBAAwB;QAANpH,IAAM;gBACFA,IADE;QAClC4I,WADkC,SAClCA,WADkC;QACrBC,cADqB,SACrBA,cADqB;;;QAGtCH,UAAUd,MAAV,KAAqB,GAAzB,EAA8B,OAAOc,UAAUrB,OAAV,CAAkBrH,IAAlB,CAAP;;wBAGzBA,IADL;;;;QAKI4I,WAAJ,EAAiB;UACTjb,WAAU8a,2BACXzI,IADW,IACLlT,MAAM,SADD,EACYwb,aAAa,IADzB,EAC+BlO,OAAOyO;SADtD;aAGO;;OAAP;;QAIIzO,QAAQqO,2BAAmBzI,IAAnB,IAAyBlT,MAAM,OAA/B,IAAd;QACMya,iBAAiBkB,2BAAmBzI,IAAnB,IAAyBlT,MAAM,gBAA/B,IAAvB;QACMwQ,SAASmL,2BAAmBzI,IAAnB,IAAyBlT,MAAM,QAA/B,IAAf;QACM2a,gBAAgBgB,2BAAmBzI,IAAnB,IAAyBlT,MAAM,eAA/B,IAAtB;QACMa,UAAU8a,2BACXzI,IADW,IACLlT,MAAM,SADD,EACYwb,aAAa,IADzB,EAC+BlO;OAD/C;QAGMoN,iBAAiBiB,2BAAmBzI,IAAnB,IAAyBlT,MAAM,gBAA/B,EAAiDa,gBAAjD,IAAvB;QACMgQ,MAAM8K,2BAAmBzI,IAAnB,IAAyBlT,MAAM,KAA/B,EAAsCa,gBAAtC,IAAZ;QACMsZ,UAAUwB,2BAAmBzI,IAAnB,IAAyBlT,MAAM,SAA/B,EAA0Ca,gBAA1C,IAAhB;QACM+Z,aAAae,2BAAmBzI,IAAnB,IAAyBlT,MAAM,YAA/B,EAA6Ca,gBAA7C,IAAnB;;yBACwB8a,2BAAmBzI,IAAnB,IAAyBlT,MAAM,gBAA/B,IA7BkB;;QA6BlC3B,GA7BkC,kBA6BlCA,GA7BkC;QA6B7Byc,MA7B6B,kBA6B7BA,MA7B6B;;;WA+BnC;kBAAA;sBAAA;oBAAA;oCAAA;oCAAA;cAAA;kCAAA;cAAA;oBAAA;sBAAA;;KAAP;;CAhCJ,CAgDA;;ACjJA;sDAAe;QAEXH,aAFW,SAEXA,aAFW;QAGXvL,IAHW,SAGXA,IAHW;QAIXzQ,CAJW,SAIXA,CAJW;QAKXmV,SALW,SAKXA,SALW;QAMXlT,MANW,SAMXA,MANW;QAOXob,SAPW,SAOXA,SAPW;QAQX1O,KARW,SAQXA,KARW;QASXjP,GATW,SASXA,GATW;;;;;;;iBAAA,GAaD,CAbC;wBAAA,GAcQ,CAACgM,aAAahM,GAAb,CAAD,CAdR;;;;;;kBAkBNsc,iBAAiBsB,QAAQ,EAlBnB;;;;;qBAmBF,CAAT;;mBACUzb,SAAS0b,MAAT,CAAgBvB,aAAhB,CApBC;;;aAAA;;mBAqBJhc,EAAEyQ,IAAF,EAAP;;yBArBW,GAuBW;mBACfuL,aADe;wBAAA;kBAAA;kCAAA;2BAKP,IALO;8BAMJrN,KANI;;aAvBX;0BAAA,GAiCYuO,cAActB,OAAd,CAAsByB,SAAtB,EAAiCG,aAAjC,CAjCZ;;;yBAmCE7Q,IAAb,CAAkBqP,aAAlB;kCAEK/Z,MADL;sCAGMA,OAAOC,OADX,yCAGaob,KAHb,uBAIIG,eAAevb,OAJnB;;;4BAQcub,eAAezB,aAA/B;;;;;0DAIG/Z,MAlDQ;2BAmDEqb,KAnDF;8BAoDKA;;;;;;;;;GApDpB;;WAA8BI,eAA9B;;;;SAA8BA,eAA9B;;;ACAA,IAAMC,OAAO;OAAA,iBACCje,GADD,EACM+Q,IADN,EACuB;;;QAAX8D,IAAW,yDAAJ,EAAI;;;;;;;;qBACCA,QAAQ,IADT;wCACxBqJ,aADwB;2BAAA,sCACR,IADQ;uBAAA,GAEdxB,aAAa1c,GAAb,CAFc;;;;qBAKhBmC,SAAS0b,MAAT,CAAgB7d,GAAhB,EAAqB+Q,IAArB,CALgB;;;eAAA;;qBAMzBzQ,EAAEyQ,IAAF,EAAP;;;;uBANgC,GAUdzQ,EAAE,MAAF,EAAUqD,GAAV,CAAc,UAACjD,CAAD,EAAIC,IAAJ;uBAAaL,EAAEK,IAAF,EAAQG,IAAR,CAAa,MAAb,CAAb;eAAd,EAAiDwP,OAAjD,EAVc;oBAAA,GAYnBkN,cAActB,OAAd,CAAsByB,SAAtB,EAAiC,EAAE3d,QAAF,EAAO+Q,UAAP,EAAazQ,IAAb,EAAgBmV,oBAAhB,EAAjC,CAZmB;wBAaClT,MAbD;mBAAA,WAaxB0M,KAbwB;2BAAA,WAajBqN,aAbiB;;;;oBAgB5B4B,iBAAiB5B,aAhBW;;;;;;qBAiBf0B,gBACb;oCAAA;4CAAA;0BAAA;oBAAA;oCAAA;8BAAA;4BAAA;;eADa,CAjBe;;;oBAAA;;;;;oCA+BzBzb,MADL;6BAEe,CAFf;gCAGkB;;;;+CAIbA,MArCyB;;;;;;;;;;CADpC,CA2CA;;"}
\ No newline at end of file