From 5c125b978d128e697f7a9f78d6045087fc697f53 Mon Sep 17 00:00:00 2001
From: Josh Goebel <me@joshgoebel.com>
Date: Thu, 7 May 2020 16:41:03 -0400
Subject: [PATCH] fix(javascript) fix regex inside parens after a non-regex
 (#2531)

* make the object attr container smarter
* deal with multi-line comments also
* comments in any order, spanning multiple lines

Essentially makes the object attr container much more sensitive by allowing it to look-ahead thru comments to find object keys - and therefore prevent them from being incorrectly matched by the "value container" rule.
---
 CHANGES.md                                    |  2 ++
 src/languages/javascript.js                   | 25 ++++++++++++++++---
 src/lib/regex.js                              |  4 +++
 test/markup/javascript/object-attr.expect.txt | 10 +++++++-
 test/markup/javascript/object-attr.txt        | 11 +++++++-
 5 files changed, 46 insertions(+), 6 deletions(-)

diff --git a/CHANGES.md b/CHANGES.md
index 2cf4ad9ace..6c3f7225ce 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -15,6 +15,8 @@ Deprecations:
 - `endSameAsBegin` is now deprecated. (#2261) [Josh Goebel][]
 
 Language Improvements:
+
+- fix(javascript) fix regex inside parens after a non-regex (#2530) [Josh Goebel][]
 - enh(typescript) use identifier to match potential keywords, preventing false positivites (#2519) [Josh Goebel][]
 - enh(javascript) use identifier to match potential keywords, preventing false positivites (#2519) [Josh Goebel][]
 - [enh] Add `OPTIMIZE:` and `HACK:` to the labels highlighted inside comments [Josh Goebel][]
diff --git a/src/languages/javascript.js b/src/languages/javascript.js
index 35b12dba02..c8334357bc 100644
--- a/src/languages/javascript.js
+++ b/src/languages/javascript.js
@@ -6,6 +6,7 @@ Website: https://developer.mozilla.org/en-US/docs/Web/JavaScript
 */
 
 import * as ECMAScript from "./lib/ecmascript";
+import * as regex from "../lib/regex";
 
 export default function(hljs) {
   var IDENT_RE = ECMAScript.IDENT_RE;
@@ -149,13 +150,29 @@ export default function(hljs) {
       hljs.C_BLOCK_COMMENT_MODE,
       NUMBER,
       { // object attr container
-        begin: /[{,\n]\s*/, relevance: 0,
+        begin: regex.concat(/[{,\n]\s*/,
+          // we need to look ahead to make sure that we actually have an
+          // attribute coming up so we don't steal a comma from a potential
+          // "value" container
+          //
+          // NOTE: this might not work how you think.  We don't actually always
+          // enter this mode and stay.  Instead it might merely match `,
+          // <comments up next>` and then immediately end after the , because it
+          // fails to find any actual attrs. But this still does the job because
+          // it prevents the value contain rule from grabbing this instead and
+          // prevening this rule from firing when we actually DO have keys.
+          regex.lookahead(regex.concat(
+            // we also need to allow for multiple possible comments inbetween
+            // the first key:value pairing
+            /(((\/\/.*)|(\/\*(.|\n)*\*\/))\s*)*/,
+            IDENT_RE + '\\s*:'))),
+        relevance: 0,
         contains: [
           {
-            begin: IDENT_RE + '\\s*:', returnBegin: true,
+            className: 'attr',
+            begin: IDENT_RE + regex.lookahead('\\s*:'),
             relevance: 0,
-            contains: [{className: 'attr', begin: IDENT_RE, relevance: 0}]
-          }
+          },
         ]
       },
       { // "value" container
diff --git a/src/lib/regex.js b/src/lib/regex.js
index 003dc9cbca..80a9ce1b70 100644
--- a/src/lib/regex.js
+++ b/src/lib/regex.js
@@ -8,6 +8,10 @@ export function source(re) {
   return (re && re.source) || re;
 }
 
+export function lookahead(regex) {
+  return concat('(?=', regex, ')');
+}
+
 export function concat(...args) {
   const joined = args.map((x) => source(x)).join("");
   return joined;
diff --git a/test/markup/javascript/object-attr.expect.txt b/test/markup/javascript/object-attr.expect.txt
index e92b578cea..54e9d7301f 100644
--- a/test/markup/javascript/object-attr.expect.txt
+++ b/test/markup/javascript/object-attr.expect.txt
@@ -1,6 +1,14 @@
 {
   <span class="hljs-attr">key</span>: value, <span class="hljs-comment">// with comment</span>
   <span class="hljs-attr">key2</span>: value,
+  <span class="hljs-attr">key2clone</span>: value,
   <span class="hljs-string">'key-3'</span>: value,
-  <span class="hljs-attr">key4</span>: <span class="hljs-literal">false</span> ? <span class="hljs-literal">undefined</span> : <span class="hljs-literal">true</span>
+  <span class="hljs-attr">key4</span>: <span class="hljs-literal">false</span> ? <span class="hljs-literal">undefined</span> : <span class="hljs-literal">true</span>,
+  <span class="hljs-attr">key5</span>: value, <span class="hljs-comment">/* with a multiline comment */</span>
+  <span class="hljs-attr">key6</span>: value,
+  <span class="hljs-attr">key7</span>: value, <span class="hljs-comment">/* with a multiline comment */</span> <span class="hljs-comment">// another comment</span>
+  <span class="hljs-attr">key8</span>: value,
+  <span class="hljs-attr">key9</span>: value, <span class="hljs-comment">/* with a REAL multiline
+comment */</span>
+  <span class="hljs-attr">key10</span>: value,
 }
diff --git a/test/markup/javascript/object-attr.txt b/test/markup/javascript/object-attr.txt
index 442c0a17ae..b024927090 100644
--- a/test/markup/javascript/object-attr.txt
+++ b/test/markup/javascript/object-attr.txt
@@ -1,6 +1,15 @@
 {
   key: value, // with comment
   key2: value,
+  key2clone: value,
   'key-3': value,
-  key4: false ? undefined : true
+  key4: false ? undefined : true,
+  key5: value, /* with a multiline comment */
+  key6: value,
+  key7: value, /* with a multiline comment */ // another comment
+  key8: value,
+  key9: value, /* with a REAL multiline
+comment */
+  key10: value,
 }
+