Merge pull request #695 from skyweb331/main

Improve markdown italicizing when following tags
Expensify · May 29, 2024 · 315f810 · 315f810
2 parents 05c03d4 + 4ef55d7
commit 315f810
Show file tree

Hide file tree

Showing 2 changed files with 23 additions and 12 deletions.
diff --git a/__tests__/ExpensiMark-HTML-test.js b/__tests__/ExpensiMark-HTML-test.js
@@ -57,6 +57,16 @@ test('Test italic markdown replacement', () => {
     expect(parser.replace(italicTestStartString)).toBe(italicTestReplacedString);
 });
 
+test('Test italic markdown replacement', () => {
+    const italicTestStartString = 'Note that _this is correctly italicized_\n'
+        + 'Note that `_this is correctly not italicized_` and _following inline contents are italicized_'
+
+    const italicTestReplacedString = 'Note that <em>this is correctly italicized</em><br />'
+        + 'Note that <code>_this is correctly not italicized_</code> and <em>following inline contents are italicized</em>'
+
+    expect(parser.replace(italicTestStartString)).toBe(italicTestReplacedString);
+});
+
 // Multi-line text wrapped in _ is successfully replaced with <em></em>
 test('Test multi-line italic markdown replacement', () => {
     const testString = '_Here is a multi-line\ncomment that should\nbe italic_ \n_\n_test\n_';

diff --git a/lib/ExpensiMark.js b/lib/ExpensiMark.js
@@ -278,24 +278,25 @@ export default class ExpensiMark {
                     return `<blockquote>${isStartingWithSpace ? ' ' : ''}${replacedText}</blockquote>`;
                 },
             },
+            /**
+             * Use \b in this case because it will match on words, letters,
+             * and _: https://www.rexegg.com/regex-boundaries.html#wordboundary
+             * Use [\s\S]* instead of .* to match newline
+             */
             {
-                /**
-                 * Use \b in this case because it will match on words, letters,
-                 * and _: https://www.rexegg.com/regex-boundaries.html#wordboundary
-                 * The !_blank is to prevent the `target="_blank">` section of the
-                 * link replacement from being captured Additionally, something like
-                 * `\b\_([^<>]*?)\_\b` doesn't work because it won't replace
-                 * `_https://www.test.com_`
-                 * Use [\s\S]* instead of .* to match newline
-                 */
                 name: 'italic',
-                regex: /(?<!<[^>]*)(\b_+|\b)(?!_blank")_((?![\s_])[\s\S]*?[^\s_](?<!\s))_(?![^\W_])(?![^<]*>)(?![^<]*(<\/pre>|<\/code>|<\/a>|<\/mention-user>|_blank))/g,
+                regex: /(<(pre|code|a|mention-user)[^>]*>(.*?)<\/\2>)|((\b_+|\b)_((?![\s_])[\s\S]*?[^\s_](?<!\s))_(?![^\W_])(?![^<]*>)(?![^<]*(<\/pre>|<\/code>|<\/a>|<\/mention-user>)))/g,
+                replacement: (match, html, tag, content, text, extraLeadingUnderscores, textWithinUnderscores) => {
+                    // Skip any <pre>, <code>, <a>, <mention-user> tag contents
+                    if (html) {
+                        return html;
+                    }
 
-                // We want to add extraLeadingUnderscores back before the <em> tag unless textWithinUnderscores starts with valid email
-                replacement: (match, extraLeadingUnderscores, textWithinUnderscores) => {
+                    // If any tags are included inside underscores, ignore it. ie. _abc <pre>pre tag</pre> abc_
                     if (textWithinUnderscores.includes('</pre>') || this.containsNonPairTag(textWithinUnderscores)) {
                         return match;
                     }
+
                     if (String(textWithinUnderscores).match(`^${Constants.CONST.REG_EXP.MARKDOWN_EMAIL}`)) {
                         return `<em>${extraLeadingUnderscores}${textWithinUnderscores}</em>`;
                     }