Factor out common URL regular expression code

This also eliminates the differences between the regular expressions.

src/node/utils/ExportHtml.js

@@ -22,6 +22,7 @@ const hooks = require('ep_etherpad-lite/static/js/pluginfw/hooks');
 const eejs = require('ep_etherpad-lite/node/eejs');
 const _analyzeLine = require('./ExportHelper')._analyzeLine;
 const _encodeWhitespace = require('./ExportHelper')._encodeWhitespace;
+const padutils = require('../../static/js/pad_utils').padutils;
 
 async function getPadHTML(pad, revNum) {
   let atext = pad.atext;
@@ -191,7 +192,7 @@ async function getHTMLFromAtext(pad, atext, authorColors) {
       }
     }
 
-    const urls = _findURLs(text);
+    const urls = padutils.findURLs(text);
 
     let idx = 0;
 
@@ -459,30 +460,6 @@ exports.getPadHTMLDocument = async function (padId, revNum) {
   });
 };
 
-// copied from ACE
-const _REGEX_WORDCHAR = /[\u0030-\u0039\u0041-\u005A\u0061-\u007A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u1FFF\u3040-\u9FFF\uF900-\uFDFF\uFE70-\uFEFE\uFF10-\uFF19\uFF21-\uFF3A\uFF41-\uFF5A\uFF66-\uFFDC]/;
-const _REGEX_SPACE = /\s/;
-const _REGEX_URLCHAR = new RegExp(`(${/[-:@a-zA-Z0-9_.,~%+\/\\?=&#;()$]/.source}|${_REGEX_WORDCHAR.source})`);
-const _REGEX_URL = new RegExp(`${/(?:(?:https?|s?ftp|ftps|file|smb|afp|nfs|(x-)?man|gopher|txmt):\/\/|mailto:)/.source + _REGEX_URLCHAR.source}*(?![:.,;])${_REGEX_URLCHAR.source}`, 'g');
-
-// returns null if no URLs, or [[startIndex1, url1], [startIndex2, url2], ...]
-
-
-function _findURLs(text) {
-  _REGEX_URL.lastIndex = 0;
-  let urls = null;
-  let execResult;
-  while ((execResult = _REGEX_URL.exec(text))) {
-    urls = (urls || []);
-    const startIndex = execResult.index;
-    const url = execResult[0];
-    urls.push([startIndex, url]);
-  }
-
-  return urls;
-}
-
-
 // copied from ACE
 function _processSpaces(s) {
   const doesWrap = true;

src/static/js/ace2_inner.js

@@ -19,6 +19,9 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
+
+const padutils = require('./pad_utils').padutils;
+
 let _, $, jQuery, plugins, Ace2Common;
 const browser = require('./browser');
 if (browser.msie) {
@@ -2806,13 +2809,9 @@ function Ace2Inner() {
     }
   }
 
-  // set of "letter or digit" chars is based on section 20.5.16 of the original Java Language Spec
-  const REGEX_WORDCHAR = /[\u0030-\u0039\u0041-\u005A\u0061-\u007A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u1FFF\u3040-\u9FFF\uF900-\uFDFF\uFE70-\uFEFE\uFF10-\uFF19\uFF21-\uFF3A\uFF41-\uFF5A\uFF66-\uFFDC]/;
   const REGEX_SPACE = /\s/;
 
-  function isWordChar(c) {
+  const isWordChar = (c) => padutils.wordCharRegex.test(c);
-    return !!REGEX_WORDCHAR.exec(c);
-  }
   editorInfo.ace_isWordChar = isWordChar;
 
   function isSpaceChar(c) {

src/static/js/linestylefilter.js

@@ -33,6 +33,7 @@ const hooks = require('./pluginfw/hooks');
 const linestylefilter = {};
 const _ = require('./underscore');
 const AttributeManager = require('./AttributeManager');
+const padutils = require('./pad_utils').padutils;
 
 linestylefilter.ATTRIB_CLASSES = {
   bold: 'tag:b',
@@ -224,11 +225,7 @@ linestylefilter.getRegexpFilter = function (regExp, tag) {
 };
 
 
-linestylefilter.REGEX_WORDCHAR = /[\u0030-\u0039\u0041-\u005A\u0061-\u007A\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF\u0100-\u1FFF\u3040-\u9FFF\uF900-\uFDFF\uFE70-\uFEFE\uFF10-\uFF19\uFF21-\uFF3A\uFF41-\uFF5A\uFF66-\uFFDC]/;
+linestylefilter.getURLFilter = linestylefilter.getRegexpFilter(padutils.urlRegex, 'url');
-linestylefilter.REGEX_URLCHAR = new RegExp(`(${/[-:@a-zA-Z0-9_.,~%+\/\\?=&#!;()$]/.source}|${linestylefilter.REGEX_WORDCHAR.source})`);
-linestylefilter.REGEX_URL = new RegExp(`${/(?:(?:https?|s?ftp|ftps|file|nfs):\/\/|(about|geo|mailto|tel):|www\.)/.source + linestylefilter.REGEX_URLCHAR.source}*(?![:.,;])${linestylefilter.REGEX_URLCHAR.source}`, 'g');
-linestylefilter.getURLFilter = linestylefilter.getRegexpFilter(
-    linestylefilter.REGEX_URL, 'url');
 
 linestylefilter.textAndClassFuncSplitter = function (func, splitPointsOpt) {
   let nextPointIndex = 0;

src/static/js/pad_utils.js

@@ -39,6 +39,55 @@ const randomString = (len) => {
   return randomstring;
 };
 
+// Set of "letter or digit" chars is based on section 20.5.16 of the original Java Language Spec.
+const wordCharRegex = new RegExp(`[${[
+  '\u0030-\u0039',
+  '\u0041-\u005A',
+  '\u0061-\u007A',
+  '\u00C0-\u00D6',
+  '\u00D8-\u00F6',
+  '\u00F8-\u00FF',
+  '\u0100-\u1FFF',
+  '\u3040-\u9FFF',
+  '\uF900-\uFDFF',
+  '\uFE70-\uFEFE',
+  '\uFF10-\uFF19',
+  '\uFF21-\uFF3A',
+  '\uFF41-\uFF5A',
+  '\uFF66-\uFFDC',
+].join('')}]`);
+
+const urlRegex = (() => {
+  // TODO: wordCharRegex matches many characters that are not permitted in URIs. Are they included
+  // here as an attempt to support IRIs? (See https://tools.ietf.org/html/rfc3987.)
+  const urlChar = `[-:@_.,~%+/?=&#!;()$${wordCharRegex.source.slice(1, -1)}]`;
+  // Matches a single character that should not be considered part of the URL if it is the last
+  // character that matches urlChar.
+  const postUrlPunct = '[:.,;]';
+  // Schemes that must be followed by ://
+  const withAuth = `(?:${[
+    '(?:x-)?man',
+    'afp',
+    'file',
+    'ftps?',
+    'gopher',
+    'https?',
+    'nfs',
+    'sftp',
+    'smb',
+    'txmt',
+  ].join('|')})://`;
+  // Schemes that do not need to be followed by ://
+  const withoutAuth = `(?:${[
+    'about',
+    'geo',
+    'mailto',
+    'tel',
+  ].join('|')}):`;
+  return new RegExp(
+      `(?:${withAuth}|${withoutAuth}|www\\.)${urlChar}*(?!${postUrlPunct})${urlChar}`, 'g');
+})();
+
 const padutils = {
   escapeHtml: (x) => Security.escapeHTML(String(x)),
   uniqueId: () => {
@@ -75,45 +124,24 @@ const padutils = {
     const hourmin = `${d.getHours()}:${(`0${d.getMinutes()}`).slice(-2)}`;
     return `${dayOfWeek} ${month} ${dayOfMonth} ${year} ${hourmin}`;
   },
+  wordCharRegex,
+  urlRegex,
+  // returns null if no URLs, or [[startIndex1, url1], [startIndex2, url2], ...]
   findURLs: (text) => {
-    // copied from ACE
+    // Copy padutils.urlRegex so that the use of .exec() below (which mutates the RegExp object)
-    const _REGEX_WORDCHAR = new RegExp(`[${[
+    // does not break other concurrent uses of padutils.urlRegex.
-      '\u0030-\u0039',
+    const urlRegex = new RegExp(padutils.urlRegex, 'g');
-      '\u0041-\u005A',
+    urlRegex.lastIndex = 0;
-      '\u0061-\u007A',
+    let urls = null;
-      '\u00C0-\u00D6',
+    let execResult;
-      '\u00D8-\u00F6',
+    // TODO: Switch to String.prototype.matchAll() after support for Node.js < 12.0.0 is dropped.
-      '\u00F8-\u00FF',
+    while ((execResult = urlRegex.exec(text))) {
-      '\u0100-\u1FFF',
+      urls = (urls || []);
-      '\u3040-\u9FFF',
+      const startIndex = execResult.index;
-      '\uF900-\uFDFF',
+      const url = execResult[0];
-      '\uFE70-\uFEFE',
+      urls.push([startIndex, url]);
-      '\uFF10-\uFF19',
+    }
-      '\uFF21-\uFF3A',
+    return urls;
-      '\uFF41-\uFF5A',
-      '\uFF66-\uFFDC',
-    ].join('')}]`);
-    const _REGEX_URLCHAR = new RegExp(`([-:@a-zA-Z0-9_.,~%+/?=&#;()$]|${_REGEX_WORDCHAR.source})`);
-    const _REGEX_URL = new RegExp(
-        '(?:(?:https?|s?ftp|ftps|file|nfs)://|(about|geo|mailto|tel):)' +
-        `${_REGEX_URLCHAR.source}*(?![:.,;])${_REGEX_URLCHAR.source}`, 'g');
-
-    // returns null if no URLs, or [[startIndex1, url1], [startIndex2, url2], ...]
-    const _findURLs = (text) => {
-      _REGEX_URL.lastIndex = 0;
-      let urls = null;
-      let execResult;
-      while ((execResult = _REGEX_URL.exec(text))) {
-        urls = (urls || []);
-        const startIndex = execResult.index;
-        const url = execResult[0];
-        urls.push([startIndex, url]);
-      }
-
-      return urls;
-    };
-
-    return _findURLs(text);
   },
   escapeHtmlWithClickableLinks: (text, target) => {
     let idx = 0;