Browse Source

Update twitter-text from 1.14 to 3.1.0 and fix toot character counting (#15382)

* Update twitter-text from 1.14 to 3.1.0

* Disable emoji parsing

* Properly depend on twitter-text for url detection

* Fix some URLs being wrongly detected client-side

* Add test for server-side validation of non-autolinkable URLs

* Fix server-side status length counting
Claire 3 years ago
parent
commit
65db262550

+ 1 - 1
Gemfile

@@ -93,7 +93,7 @@ gem 'sprockets-rails', '~> 3.2', require: 'sprockets/railtie'
 gem 'stoplight', '~> 2.2.1'
 gem 'strong_migrations', '~> 0.7'
 gem 'tty-prompt', '~> 0.23', require: false
-gem 'twitter-text', '~> 1.14'
+gem 'twitter-text', '~> 3.1.0'
 gem 'tzinfo-data', '~> 1.2021'
 gem 'webpacker', '~> 5.2'
 gem 'webpush'

+ 3 - 2
Gemfile.lock

@@ -641,7 +641,8 @@ GEM
       tty-screen (~> 0.8)
       wisper (~> 2.0)
     tty-screen (0.8.1)
-    twitter-text (1.14.7)
+    twitter-text (3.1.0)
+      idn-ruby
       unf (~> 0.1.0)
     tzinfo (1.2.9)
       thread_safe (~> 0.1)
@@ -810,7 +811,7 @@ DEPENDENCIES
   strong_migrations (~> 0.7)
   thor (~> 1.1)
   tty-prompt (~> 0.23)
-  twitter-text (~> 1.14)
+  twitter-text (~> 3.1.0)
   tzinfo-data (~> 1.2021)
   webauthn (~> 3.0.0.alpha1)
   webmock (~> 3.12)

+ 1 - 1
app/javascript/mastodon/features/compose/util/counter.js

@@ -1,6 +1,6 @@
 import { urlRegex } from './url_regex';
 
-const urlPlaceholder = 'xxxxxxxxxxxxxxxxxxxxxxx';
+const urlPlaceholder = '$2xxxxxxxxxxxxxxxxxxxxxxx';
 
 export function countableText(inputText) {
   return inputText

+ 28 - 194
app/javascript/mastodon/features/compose/util/url_regex.js

@@ -1,196 +1,30 @@
-const regexen = {};
+import regexSupplant from 'twitter-text/dist/lib/regexSupplant';
+import validUrlPrecedingChars from 'twitter-text/dist/regexp/validUrlPrecedingChars';
+import validDomain from 'twitter-text/dist/regexp/validDomain';
+import validPortNumber from 'twitter-text/dist/regexp/validPortNumber';
+import validUrlPath from 'twitter-text/dist/regexp/validUrlPath';
+import validUrlQueryChars from 'twitter-text/dist/regexp/validUrlQueryChars';
+import validUrlQueryEndingChars from 'twitter-text/dist/regexp/validUrlQueryEndingChars';
 
-const regexSupplant = function(regex, flags) {
-  flags = flags || '';
-  if (typeof regex !== 'string') {
-    if (regex.global && flags.indexOf('g') < 0) {
-      flags += 'g';
-    }
-    if (regex.ignoreCase && flags.indexOf('i') < 0) {
-      flags += 'i';
-    }
-    if (regex.multiline && flags.indexOf('m') < 0) {
-      flags += 'm';
-    }
+// The difference with twitter-text's extractURL is that the protocol isn't
+// optional.
 
-    regex = regex.source;
-  }
-  return new RegExp(regex.replace(/#\{(\w+)\}/g, function(match, name) {
-    var newRegex = regexen[name] || '';
-    if (typeof newRegex !== 'string') {
-      newRegex = newRegex.source;
-    }
-    return newRegex;
-  }), flags);
-};
-
-const stringSupplant = function(str, values) {
-  return str.replace(/#\{(\w+)\}/g, function(match, name) {
-    return values[name] || '';
-  });
-};
-
-export const urlRegex = (function() {
-  regexen.spaces_group = /\x09-\x0D\x20\x85\xA0\u1680\u180E\u2000-\u200A\u2028\u2029\u202F\u205F\u3000/;
-  regexen.invalid_chars_group = /\uFFFE\uFEFF\uFFFF\u202A-\u202E/;
-  regexen.punct = /\!'#%&'\(\)*\+,\\\-\.\/:;<=>\?@\[\]\^_{|}~\$/;
-  regexen.validUrlPrecedingChars = regexSupplant(/(?:[^A-Za-z0-9@@$###{invalid_chars_group}]|^)/);
-  regexen.invalidDomainChars = stringSupplant('#{punct}#{spaces_group}#{invalid_chars_group}', regexen);
-  regexen.validDomainChars = regexSupplant(/[^#{invalidDomainChars}]/);
-  regexen.validSubdomain = regexSupplant(/(?:(?:#{validDomainChars}(?:[_-]|#{validDomainChars})*)?#{validDomainChars}\.)/);
-  regexen.validDomainName = regexSupplant(/(?:(?:#{validDomainChars}(?:-|#{validDomainChars})*)?#{validDomainChars}\.)/);
-  regexen.validGTLD = regexSupplant(RegExp(
-    '(?:(?:' +
-      '삼성|닷컴|닷넷|香格里拉|餐厅|食品|飞利浦|電訊盈科|集团|通販|购物|谷歌|诺基亚|联通|网络|网站|网店|网址|组织机构|移动|珠宝|点看|游戏|淡马锡|机构|書籍|时尚|新闻|政府|' +
-      '政务|手表|手机|我爱你|慈善|微博|广东|工行|家電|娱乐|天主教|大拿|大众汽车|在线|嘉里大酒店|嘉里|商标|商店|商城|公益|公司|八卦|健康|信息|佛山|企业|中文网|中信|世界|' +
-      'ポイント|ファッション|セール|ストア|コム|グーグル|クラウド|みんな|คอม|संगठन|नेट|कॉम|همراه|موقع|موبايلي|كوم|كاثوليك|عرب|شبكة|' +
-      'بيتك|بازار|العليان|ارامكو|اتصالات|ابوظبي|קום|сайт|рус|орг|онлайн|москва|ком|католик|дети|' +
-      'zuerich|zone|zippo|zip|zero|zara|zappos|yun|youtube|you|yokohama|yoga|yodobashi|yandex|yamaxun|' +
-      'yahoo|yachts|xyz|xxx|xperia|xin|xihuan|xfinity|xerox|xbox|wtf|wtc|wow|world|works|work|woodside|' +
-      'wolterskluwer|wme|winners|wine|windows|win|williamhill|wiki|wien|whoswho|weir|weibo|wedding|wed|' +
-      'website|weber|webcam|weatherchannel|weather|watches|watch|warman|wanggou|wang|walter|walmart|' +
-      'wales|vuelos|voyage|voto|voting|vote|volvo|volkswagen|vodka|vlaanderen|vivo|viva|vistaprint|' +
-      'vista|vision|visa|virgin|vip|vin|villas|viking|vig|video|viajes|vet|versicherung|' +
-      'vermögensberatung|vermögensberater|verisign|ventures|vegas|vanguard|vana|vacations|ups|uol|uno|' +
-      'university|unicom|uconnect|ubs|ubank|tvs|tushu|tunes|tui|tube|trv|trust|travelersinsurance|' +
-      'travelers|travelchannel|travel|training|trading|trade|toys|toyota|town|tours|total|toshiba|' +
-      'toray|top|tools|tokyo|today|tmall|tkmaxx|tjx|tjmaxx|tirol|tires|tips|tiffany|tienda|tickets|' +
-      'tiaa|theatre|theater|thd|teva|tennis|temasek|telefonica|telecity|tel|technology|tech|team|tdk|' +
-      'tci|taxi|tax|tattoo|tatar|tatamotors|target|taobao|talk|taipei|tab|systems|symantec|sydney|' +
-      'swiss|swiftcover|swatch|suzuki|surgery|surf|support|supply|supplies|sucks|style|study|studio|' +
-      'stream|store|storage|stockholm|stcgroup|stc|statoil|statefarm|statebank|starhub|star|staples|' +
-      'stada|srt|srl|spreadbetting|spot|spiegel|space|soy|sony|song|solutions|solar|sohu|software|' +
-      'softbank|social|soccer|sncf|smile|smart|sling|skype|sky|skin|ski|site|singles|sina|silk|shriram|' +
-      'showtime|show|shouji|shopping|shop|shoes|shiksha|shia|shell|shaw|sharp|shangrila|sfr|sexy|sex|' +
-      'sew|seven|ses|services|sener|select|seek|security|secure|seat|search|scot|scor|scjohnson|' +
-      'science|schwarz|schule|school|scholarships|schmidt|schaeffler|scb|sca|sbs|sbi|saxo|save|sas|' +
-      'sarl|sapo|sap|sanofi|sandvikcoromant|sandvik|samsung|samsclub|salon|sale|sakura|safety|safe|' +
-      'saarland|ryukyu|rwe|run|ruhr|rugby|rsvp|room|rogers|rodeo|rocks|rocher|rmit|rip|rio|ril|' +
-      'rightathome|ricoh|richardli|rich|rexroth|reviews|review|restaurant|rest|republican|report|' +
-      'repair|rentals|rent|ren|reliance|reit|reisen|reise|rehab|redumbrella|redstone|red|recipes|' +
-      'realty|realtor|realestate|read|raid|radio|racing|qvc|quest|quebec|qpon|pwc|pub|prudential|pru|' +
-      'protection|property|properties|promo|progressive|prof|productions|prod|pro|prime|press|praxi|' +
-      'pramerica|post|porn|politie|poker|pohl|pnc|plus|plumbing|playstation|play|place|pizza|pioneer|' +
-      'pink|ping|pin|pid|pictures|pictet|pics|piaget|physio|photos|photography|photo|phone|philips|phd|' +
-      'pharmacy|pfizer|pet|pccw|pay|passagens|party|parts|partners|pars|paris|panerai|panasonic|' +
-      'pamperedchef|page|ovh|ott|otsuka|osaka|origins|orientexpress|organic|org|orange|oracle|open|ooo|' +
-      'onyourside|online|onl|ong|one|omega|ollo|oldnavy|olayangroup|olayan|okinawa|office|off|observer|' +
-      'obi|nyc|ntt|nrw|nra|nowtv|nowruz|now|norton|northwesternmutual|nokia|nissay|nissan|ninja|nikon|' +
-      'nike|nico|nhk|ngo|nfl|nexus|nextdirect|next|news|newholland|new|neustar|network|netflix|netbank|' +
-      'net|nec|nba|navy|natura|nationwide|name|nagoya|nadex|nab|mutuelle|mutual|museum|mtr|mtpc|mtn|' +
-      'msd|movistar|movie|mov|motorcycles|moto|moscow|mortgage|mormon|mopar|montblanc|monster|money|' +
-      'monash|mom|moi|moe|moda|mobily|mobile|mobi|mma|mls|mlb|mitsubishi|mit|mint|mini|mil|microsoft|' +
-      'miami|metlife|merckmsd|meo|menu|men|memorial|meme|melbourne|meet|media|med|mckinsey|mcdonalds|' +
-      'mcd|mba|mattel|maserati|marshalls|marriott|markets|marketing|market|map|mango|management|man|' +
-      'makeup|maison|maif|madrid|macys|luxury|luxe|lupin|lundbeck|ltda|ltd|lplfinancial|lpl|love|lotto|' +
-      'lotte|london|lol|loft|locus|locker|loans|loan|lixil|living|live|lipsy|link|linde|lincoln|limo|' +
-      'limited|lilly|like|lighting|lifestyle|lifeinsurance|life|lidl|liaison|lgbt|lexus|lego|legal|' +
-      'lefrak|leclerc|lease|lds|lawyer|law|latrobe|latino|lat|lasalle|lanxess|landrover|land|lancome|' +
-      'lancia|lancaster|lamer|lamborghini|ladbrokes|lacaixa|kyoto|kuokgroup|kred|krd|kpn|kpmg|kosher|' +
-      'komatsu|koeln|kiwi|kitchen|kindle|kinder|kim|kia|kfh|kerryproperties|kerrylogistics|kerryhotels|' +
-      'kddi|kaufen|juniper|juegos|jprs|jpmorgan|joy|jot|joburg|jobs|jnj|jmp|jll|jlc|jio|jewelry|jetzt|' +
-      'jeep|jcp|jcb|java|jaguar|iwc|iveco|itv|itau|istanbul|ist|ismaili|iselect|irish|ipiranga|' +
-      'investments|intuit|international|intel|int|insure|insurance|institute|ink|ing|info|infiniti|' +
-      'industries|immobilien|immo|imdb|imamat|ikano|iinet|ifm|ieee|icu|ice|icbc|ibm|hyundai|hyatt|' +
-      'hughes|htc|hsbc|how|house|hotmail|hotels|hoteles|hot|hosting|host|hospital|horse|honeywell|' +
-      'honda|homesense|homes|homegoods|homedepot|holiday|holdings|hockey|hkt|hiv|hitachi|hisamitsu|' +
-      'hiphop|hgtv|hermes|here|helsinki|help|healthcare|health|hdfcbank|hdfc|hbo|haus|hangout|hamburg|' +
-      'hair|guru|guitars|guide|guge|gucci|guardian|group|grocery|gripe|green|gratis|graphics|grainger|' +
-      'gov|got|gop|google|goog|goodyear|goodhands|goo|golf|goldpoint|gold|godaddy|gmx|gmo|gmbh|gmail|' +
-      'globo|global|gle|glass|glade|giving|gives|gifts|gift|ggee|george|genting|gent|gea|gdn|gbiz|' +
-      'garden|gap|games|game|gallup|gallo|gallery|gal|fyi|futbol|furniture|fund|fun|fujixerox|fujitsu|' +
-      'ftr|frontier|frontdoor|frogans|frl|fresenius|free|fox|foundation|forum|forsale|forex|ford|' +
-      'football|foodnetwork|food|foo|fly|flsmidth|flowers|florist|flir|flights|flickr|fitness|fit|' +
-      'fishing|fish|firmdale|firestone|fire|financial|finance|final|film|fido|fidelity|fiat|ferrero|' +
-      'ferrari|feedback|fedex|fast|fashion|farmers|farm|fans|fan|family|faith|fairwinds|fail|fage|' +
-      'extraspace|express|exposed|expert|exchange|everbank|events|eus|eurovision|etisalat|esurance|' +
-      'estate|esq|erni|ericsson|equipment|epson|epost|enterprises|engineering|engineer|energy|emerck|' +
-      'email|education|edu|edeka|eco|eat|earth|dvr|dvag|durban|dupont|duns|dunlop|duck|dubai|dtv|drive|' +
-      'download|dot|doosan|domains|doha|dog|dodge|doctor|docs|dnp|diy|dish|discover|discount|directory|' +
-      'direct|digital|diet|diamonds|dhl|dev|design|desi|dentist|dental|democrat|delta|deloitte|dell|' +
-      'delivery|degree|deals|dealer|deal|dds|dclk|day|datsun|dating|date|data|dance|dad|dabur|cyou|' +
-      'cymru|cuisinella|csc|cruises|cruise|crs|crown|cricket|creditunion|creditcard|credit|courses|' +
-      'coupons|coupon|country|corsica|coop|cool|cookingchannel|cooking|contractors|contact|consulting|' +
-      'construction|condos|comsec|computer|compare|company|community|commbank|comcast|com|cologne|' +
-      'college|coffee|codes|coach|clubmed|club|cloud|clothing|clinique|clinic|click|cleaning|claims|' +
-      'cityeats|city|citic|citi|citadel|cisco|circle|cipriani|church|chrysler|chrome|christmas|chloe|' +
-      'chintai|cheap|chat|chase|channel|chanel|cfd|cfa|cern|ceo|center|ceb|cbs|cbre|cbn|cba|catholic|' +
-      'catering|cat|casino|cash|caseih|case|casa|cartier|cars|careers|career|care|cards|caravan|car|' +
-      'capitalone|capital|capetown|canon|cancerresearch|camp|camera|cam|calvinklein|call|cal|cafe|cab|' +
-      'bzh|buzz|buy|business|builders|build|bugatti|budapest|brussels|brother|broker|broadway|' +
-      'bridgestone|bradesco|box|boutique|bot|boston|bostik|bosch|boots|booking|book|boo|bond|bom|bofa|' +
-      'boehringer|boats|bnpparibas|bnl|bmw|bms|blue|bloomberg|blog|blockbuster|blanco|blackfriday|' +
-      'black|biz|bio|bingo|bing|bike|bid|bible|bharti|bet|bestbuy|best|berlin|bentley|beer|beauty|' +
-      'beats|bcn|bcg|bbva|bbt|bbc|bayern|bauhaus|basketball|baseball|bargains|barefoot|barclays|' +
-      'barclaycard|barcelona|bar|bank|band|bananarepublic|banamex|baidu|baby|azure|axa|aws|avianca|' +
-      'autos|auto|author|auspost|audio|audible|audi|auction|attorney|athleta|associates|asia|asda|arte|' +
-      'art|arpa|army|archi|aramco|arab|aquarelle|apple|app|apartments|aol|anz|anquan|android|analytics|' +
-      'amsterdam|amica|amfam|amex|americanfamily|americanexpress|alstom|alsace|ally|allstate|allfinanz|' +
-      'alipay|alibaba|alfaromeo|akdn|airtel|airforce|airbus|aigo|aig|agency|agakhan|africa|afl|' +
-      'afamilycompany|aetna|aero|aeg|adult|ads|adac|actor|active|aco|accountants|accountant|accenture|' +
-      'academy|abudhabi|abogado|able|abc|abbvie|abbott|abb|abarth|aarp|aaa|onion' +
-    ')(?=[^0-9a-zA-Z@]|$))'));
-  regexen.validCCTLD = regexSupplant(RegExp(
-    '(?:(?:' +
-      '한국|香港|澳門|新加坡|台灣|台湾|中國|中国|გე|ไทย|ලංකා|ഭാരതം|ಭಾರತ|భారత్|சிங்கப்பூர்|இலங்கை|இந்தியா|ଭାରତ|ભારત|ਭਾਰਤ|' +
-      'ভাৰত|ভারত|বাংলা|भारोत|भारतम्|भारत|ڀارت|پاکستان|مليسيا|مصر|قطر|فلسطين|عمان|عراق|سورية|سودان|تونس|' +
-      'بھارت|بارت|ایران|امارات|المغرب|السعودية|الجزائر|الاردن|հայ|қаз|укр|срб|рф|мон|мкд|ею|бел|бг|ελ|' +
-      'zw|zm|za|yt|ye|ws|wf|vu|vn|vi|vg|ve|vc|va|uz|uy|us|um|uk|ug|ua|tz|tw|tv|tt|tr|tp|to|tn|tm|tl|tk|' +
-      'tj|th|tg|tf|td|tc|sz|sy|sx|sv|su|st|ss|sr|so|sn|sm|sl|sk|sj|si|sh|sg|se|sd|sc|sb|sa|rw|ru|rs|ro|' +
-      're|qa|py|pw|pt|ps|pr|pn|pm|pl|pk|ph|pg|pf|pe|pa|om|nz|nu|nr|np|no|nl|ni|ng|nf|ne|nc|na|mz|my|mx|' +
-      'mw|mv|mu|mt|ms|mr|mq|mp|mo|mn|mm|ml|mk|mh|mg|mf|me|md|mc|ma|ly|lv|lu|lt|ls|lr|lk|li|lc|lb|la|kz|' +
-      'ky|kw|kr|kp|kn|km|ki|kh|kg|ke|jp|jo|jm|je|it|is|ir|iq|io|in|im|il|ie|id|hu|ht|hr|hn|hm|hk|gy|gw|' +
-      'gu|gt|gs|gr|gq|gp|gn|gm|gl|gi|gh|gg|gf|ge|gd|gb|ga|fr|fo|fm|fk|fj|fi|eu|et|es|er|eh|eg|ee|ec|dz|' +
-      'do|dm|dk|dj|de|cz|cy|cx|cw|cv|cu|cr|co|cn|cm|cl|ck|ci|ch|cg|cf|cd|cc|ca|bz|by|bw|bv|bt|bs|br|bq|' +
-      'bo|bn|bm|bl|bj|bi|bh|bg|bf|be|bd|bb|ba|az|ax|aw|au|at|as|ar|aq|ao|an|am|al|ai|ag|af|ae|ad|ac' +
-    ')(?=[^0-9a-zA-Z@]|$))'));
-  regexen.validPunycode = /(?:xn--[0-9a-z]+)/;
-  regexen.validSpecialCCTLD = /(?:(?:co|tv)(?=[^0-9a-zA-Z@]|$))/;
-  regexen.validDomain = regexSupplant(/(?:#{validSubdomain}*#{validDomainName}(?:#{validGTLD}|#{validCCTLD}|#{validPunycode}))/);
-  regexen.validPortNumber = /[0-9]+/;
-  regexen.pd = /\u002d\u058a\u05be\u1400\u1806\u2010-\u2015\u2e17\u2e1a\u2e3a\u2e40\u301c\u3030\u30a0\ufe31\ufe58\ufe63\uff0d/;
-  regexen.validGeneralUrlPathChars = regexSupplant(/[^#{spaces_group}\(\)\?]/i);
-  // Allow URL paths to contain up to two nested levels of balanced parens
-  //  1. Used in Wikipedia URLs like /Primer_(film)
-  //  2. Used in IIS sessions like /S(dfd346)/
-  //  3. Used in Rdio URLs like /track/We_Up_(Album_Version_(Edited))/
-  regexen.validUrlBalancedParens = regexSupplant(
-    '\\('                                   +
-      '(?:'                                 +
-        '#{validGeneralUrlPathChars}+'      +
-        '|'                                 +
-        // allow one nested level of balanced parentheses
-        '(?:'                               +
-          '#{validGeneralUrlPathChars}*'    +
-          '\\('                             +
-            '#{validGeneralUrlPathChars}+'  +
-          '\\)'                             +
-          '#{validGeneralUrlPathChars}*'    +
-        ')'                                 +
-      ')'                                   +
-    '\\)',
-    'i');
-  // Valid end-of-path characters (so /foo. does not gobble the period).
-  // 1. Allow =&# for empty URL parameters and other URL-join artifacts
-  regexen.validUrlPathEndingChars = regexSupplant(/[^#{spaces_group}\(\)\?!\*';:=\,\.\$%\[\]#{pd}~&\|@]|(?:#{validUrlBalancedParens})/i);
-  // Allow @ in a url, but only in the middle. Catch things like http://example.com/@user/
-  regexen.validUrlPath = regexSupplant('(?:' +
-    '(?:' +
-      '#{validGeneralUrlPathChars}*' +
-        '(?:#{validUrlBalancedParens}#{validGeneralUrlPathChars}*)*' +
-        '#{validUrlPathEndingChars}'+
-      ')|(?:@#{validGeneralUrlPathChars}+\/)'+
-    ')', 'i');
-  regexen.validUrlQueryChars = /[a-z0-9!?\*'@\(\);:&=\+\$\/%#\[\]\-_\.,~|]/i;
-  regexen.validUrlQueryEndingChars = /[a-z0-9_&=#\/]/i;
-  regexen.validUrl = regexSupplant(
-    '('                                                          + // $1 URL
-      '(https?:\\/\\/)'                                          + // $2 Protocol
-      '(#{validDomain})'                                         + // $3 Domain(s)
-      '(?::(#{validPortNumber}))?'                               + // $4 Port number (optional)
-      '(\\/#{validUrlPath}*)?'                                   + // $5 URL Path
-      '(\\?#{validUrlQueryChars}*#{validUrlQueryEndingChars})?'  + // $6 Query String
-    ')',
-    'gi');
-  return regexen.validUrl;
-}());
+export const urlRegex = regexSupplant(
+  '('                                                          + // $1 URL
+    '(#{validUrlPrecedingChars})'                              + // $2
+    '(https?:\\/\\/)'                                          + // $3 Protocol
+    '(#{validDomain})'                                         + // $4 Domain(s)
+    '(?::(#{validPortNumber}))?'                               + // $5 Port number (optional)
+    '(\\/#{validUrlPath}*)?'                                   + // $6 URL Path
+    '(\\?#{validUrlQueryChars}*#{validUrlQueryEndingChars})?'  + // $7 Query String
+  ')',
+  {
+    validUrlPrecedingChars,
+    validDomain,
+    validPortNumber,
+    validUrlPath,
+    validUrlQueryChars,
+    validUrlQueryEndingChars,
+  },
+  'gi',
+);

+ 4 - 4
app/lib/extractor.rb

@@ -1,20 +1,20 @@
 # frozen_string_literal: true
 
 module Extractor
-  extend Twitter::Extractor
+  extend Twitter::TwitterText::Extractor
 
   module_function
 
   # :yields: username, list_slug, start, end
   def extract_mentions_or_lists_with_indices(text)
-    return [] unless Twitter::Regex[:at_signs].match?(text)
+    return [] unless Twitter::TwitterText::Regex[:at_signs].match?(text)
 
     possible_entries = []
 
     text.to_s.scan(Account::MENTION_RE) do |screen_name, _|
       match_data = $LAST_MATCH_INFO
       after = $'
-      unless Twitter::Regex[:end_mention_match].match?(after)
+      unless Twitter::TwitterText::Regex[:end_mention_match].match?(after)
         start_position = match_data.char_begin(1) - 1
         end_position = match_data.char_end(1)
         possible_entries << {
@@ -44,7 +44,7 @@ module Extractor
       if %r{\A://}.match?(after)
         hash_text.match(/(.+)(https?\Z)/) do |matched|
           hash_text = matched[1]
-          end_position -= matched[2].char_length
+          end_position -= matched[2].codepoint_length
         end
       end
 

+ 1 - 1
app/lib/formatter.rb

@@ -260,7 +260,7 @@ class Formatter
 
     html_attrs[:rel] = "me #{html_attrs[:rel]}" if options[:me]
 
-    Twitter::Autolink.send(:link_to_text, entity, link_html(entity[:url]), url, html_attrs)
+    Twitter::TwitterText::Autolink.send(:link_to_text, entity, link_html(entity[:url]), url, html_attrs)
   rescue Addressable::URI::InvalidURIError, IDN::Idna::IdnaError
     encode(entity[:url])
   end

+ 6 - 6
app/services/fetch_link_card_service.rb

@@ -2,12 +2,12 @@
 
 class FetchLinkCardService < BaseService
   URL_PATTERN = %r{
-    (                                                                                                 #   $1 URL
-      (https?:\/\/)                                                                                   #   $2 Protocol (required)
-      (#{Twitter::Regex[:valid_domain]})                                                              #   $3 Domain(s)
-      (?::(#{Twitter::Regex[:valid_port_number]}))?                                                   #   $4 Port number (optional)
-      (/#{Twitter::Regex[:valid_url_path]}*)?                                                         #   $5 URL Path and anchor
-      (\?#{Twitter::Regex[:valid_url_query_chars]}*#{Twitter::Regex[:valid_url_query_ending_chars]})? #   $6 Query String
+    (                                                                                                                           #   $1 URL
+      (https?:\/\/)                                                                                                             #   $2 Protocol (required)
+      (#{Twitter::TwitterText::Regex[:valid_domain]})                                                                           #   $3 Domain(s)
+      (?::(#{Twitter::TwitterText::Regex[:valid_port_number]}))?                                                                #   $4 Port number (optional)
+      (/#{Twitter::TwitterText::Regex[:valid_url_path]}*)?                                                                      #   $5 URL Path and anchor
+      (\?#{Twitter::TwitterText::Regex[:valid_url_query_chars]}*#{Twitter::TwitterText::Regex[:valid_url_query_ending_chars]})? #   $6 Query String
     )
   }iox
 

+ 8 - 1
app/validators/status_length_validator.rb

@@ -2,6 +2,13 @@
 
 class StatusLengthValidator < ActiveModel::Validator
   MAX_CHARS = 500
+  URL_PATTERN = %r{
+    (?:
+      (#{Twitter::TwitterText::Regex[:valid_url_preceding_chars]})
+      (#{FetchLinkCardService::URL_PATTERN})
+    )
+  }iox
+  URL_PLACEHOLDER = "\1#{'x' * 23}"
 
   def validate(status)
     return unless status.local? && !status.reblog?
@@ -28,7 +35,7 @@ class StatusLengthValidator < ActiveModel::Validator
     return '' if @status.text.nil?
 
     @status.text.dup.tap do |new_text|
-      new_text.gsub!(FetchLinkCardService::URL_PATTERN, 'x' * 23)
+      new_text.gsub!(URL_PATTERN, URL_PLACEHOLDER)
       new_text.gsub!(Account::MENTION_RE, '@\2')
     end
   end

+ 8 - 2
config/initializers/twitter_regex.rb

@@ -1,4 +1,10 @@
-module Twitter
+module Twitter::TwitterText
+  class Configuration
+    def emoji_parsing_enabled
+      false
+    end
+  end
+
   class Regex
     REGEXEN[:valid_general_url_path_chars] = /[^\p{White_Space}<>\(\)\?]/iou
     REGEXEN[:valid_url_path_ending_chars] = /[^\p{White_Space}\(\)\?!\*"'「」<>;:=\,\.\$%\[\]~&\|@]|(?:#{REGEXEN[:valid_url_balanced_parens]})/iou
@@ -79,7 +85,7 @@ module Twitter
       return [] unless text && text.index(":")
       urls = []
 
-      text.to_s.scan(Twitter::Regex[:valid_extended_uri]) do
+      text.to_s.scan(Twitter::TwitterText::Regex[:valid_extended_uri]) do
         valid_uri_match_data = $~
 
         start_position = valid_uri_match_data.char_begin(3)

+ 1 - 0
package.json

@@ -162,6 +162,7 @@
     "tesseract.js": "^2.1.1",
     "throng": "^4.0.0",
     "tiny-queue": "^0.2.1",
+    "twitter-text": "3.1.0",
     "uuid": "^8.3.1",
     "webpack": "^4.46.0",
     "webpack-assets-manifest": "^4.0.1",

+ 8 - 0
spec/lib/formatter_spec.rb

@@ -21,6 +21,14 @@ RSpec.describe Formatter do
       end
     end
 
+    context 'given a stand-alone URL with a newer TLD' do
+      let(:text) { 'http://example.gay' }
+
+      it 'matches the full URL' do
+        is_expected.to include 'href="http://example.gay"'
+      end
+    end
+
     context 'given a stand-alone IDN URL' do
       let(:text) { 'https://nic.みんな/' }
 

+ 8 - 0
spec/validators/status_length_validator_spec.rb

@@ -42,6 +42,14 @@ describe StatusLengthValidator do
       expect(status.errors).to_not have_received(:add)
     end
 
+    it 'does not count non-autolinkable URLs as 23 characters flat' do
+      text   = ('a' * 476) + "http://#{'b' * 30}.com/example"
+      status = double(spoiler_text: '', text: text, errors: double(add: nil), local?: true, reblog?: false)
+
+      subject.validate(status)
+      expect(status.errors).to have_received(:add)
+    end
+
     it 'counts only the front part of remote usernames' do
       text   = ('a' * 475) + " @alice@#{'b' * 30}.com"
       status = double(spoiler_text: '', text: text, errors: double(add: nil), local?: true, reblog?: false)

+ 22 - 2
yarn.lock

@@ -939,7 +939,7 @@
   dependencies:
     regenerator-runtime "^0.12.0"
 
-"@babel/runtime@^7.1.2", "@babel/runtime@^7.10.2", "@babel/runtime@^7.11.2", "@babel/runtime@^7.12.1", "@babel/runtime@^7.12.5", "@babel/runtime@^7.13.8", "@babel/runtime@^7.2.0", "@babel/runtime@^7.4.4", "@babel/runtime@^7.5.5", "@babel/runtime@^7.6.3", "@babel/runtime@^7.7.2", "@babel/runtime@^7.8.4", "@babel/runtime@^7.9.2":
+"@babel/runtime@^7.1.2", "@babel/runtime@^7.10.2", "@babel/runtime@^7.11.2", "@babel/runtime@^7.12.1", "@babel/runtime@^7.12.5", "@babel/runtime@^7.13.8", "@babel/runtime@^7.2.0", "@babel/runtime@^7.3.1", "@babel/runtime@^7.4.4", "@babel/runtime@^7.5.5", "@babel/runtime@^7.6.3", "@babel/runtime@^7.7.2", "@babel/runtime@^7.8.4", "@babel/runtime@^7.9.2":
   version "7.13.8"
   resolved "https://registry.yarnpkg.com/@babel/runtime/-/runtime-7.13.8.tgz#cc886a85c072df1de23670dc1aa59fc116c4017c"
   integrity sha512-CwQljpw6qSayc0fRG1soxHAKs1CnQMOChm4mlQP6My0kf9upVGizj/KhlTTgyUnETmHpcUXjaluNAkteRFuafg==
@@ -3217,6 +3217,11 @@ core-js@^2.4.0:
   resolved "https://registry.yarnpkg.com/core-js/-/core-js-2.6.11.tgz#38831469f9922bded8ee21c9dc46985e0399308c"
   integrity sha512-5wjnpaT/3dV+XB4borEsnAYQchn00XSgTAWKDkEqv+K8KevjbzmofK6hfJ9TZIlpj2N0xQpazy7PiRQiWHqzWg==
 
+core-js@^2.5.0:
+  version "2.6.12"
+  resolved "https://registry.yarnpkg.com/core-js/-/core-js-2.6.12.tgz#d9333dfa7b065e347cc5682219d6f690859cc2ec"
+  integrity sha512-Kb2wC0fvsWfQrgk8HU5lW6U/Lcs8+9aaYcy4ZFc6DDlo4nZ7n70dEgE5rtR0oG6ufKDUnrwfWL1mXR5ljDatrQ==
+
 core-util-is@1.0.2, core-util-is@~1.0.0:
   version "1.0.2"
   resolved "https://registry.yarnpkg.com/core-util-is/-/core-util-is-1.0.2.tgz#b5fd54220aa2bc5ab57aab7140c940754503c1a7"
@@ -8649,7 +8654,7 @@ punycode@1.3.2:
   resolved "https://registry.yarnpkg.com/punycode/-/punycode-1.3.2.tgz#9653a036fb7c1ee42342f2325cceefea3926c48d"
   integrity sha1-llOgNvt8HuQjQvIyXM7v6jkmxI0=
 
-punycode@^1.2.4:
+punycode@1.4.1, punycode@^1.2.4:
   version "1.4.1"
   resolved "https://registry.yarnpkg.com/punycode/-/punycode-1.4.1.tgz#c0d5a63b2718800ad8e1eb0fa5269c84dd41845e"
   integrity sha1-wNWmOycYgArY4esPpSachN1BhF4=
@@ -10722,6 +10727,21 @@ tweetnacl@^0.14.3, tweetnacl@~0.14.0:
   resolved "https://registry.yarnpkg.com/tweetnacl/-/tweetnacl-0.14.5.tgz#5ae68177f192d4456269d108afa93ff8743f4f64"
   integrity sha1-WuaBd/GS1EViadEIr6k/+HQ/T2Q=
 
+twemoji-parser@^11.0.2:
+  version "11.0.2"
+  resolved "https://registry.yarnpkg.com/twemoji-parser/-/twemoji-parser-11.0.2.tgz#24e87c2008abe8544c962f193b88b331de32b446"
+  integrity sha512-5kO2XCcpAql6zjdLwRwJjYvAZyDy3+Uj7v1ipBzLthQmDL7Ce19bEqHr3ImSNeoSW2OA8u02XmARbXHaNO8GhA==
+
+twitter-text@3.1.0:
+  version "3.1.0"
+  resolved "https://registry.yarnpkg.com/twitter-text/-/twitter-text-3.1.0.tgz#798e932b289f506efe2a1f03fe917ba30627f125"
+  integrity sha512-nulfUi3FN6z0LUjYipJid+eiwXvOLb8Ass7Jy/6zsXmZK3URte043m8fL3FyDzrK+WLpyqhHuR/TcARTN/iuGQ==
+  dependencies:
+    "@babel/runtime" "^7.3.1"
+    core-js "^2.5.0"
+    punycode "1.4.1"
+    twemoji-parser "^11.0.2"
+
 type-check@^0.4.0, type-check@~0.4.0:
   version "0.4.0"
   resolved "https://registry.yarnpkg.com/type-check/-/type-check-0.4.0.tgz#07b8203bfa7056c0657050e3ccd2c37730bab8f1"