From 11fcaea98b16679912d9d8328c56696467a0cc18 Mon Sep 17 00:00:00 2001
From: Josh Goebel <me@joshgoebel.com>
Date: Thu, 6 Feb 2020 17:20:25 -0500
Subject: [PATCH] enh(parser) better regular expression detection (#2380)

* enh(parser) better regular expression detection

- fewer false positives
- Uses a look-ahead to make sure we have a WHOLE regex and not
  just a partial (which is more likely a math expression)
---
 CHANGES.md                              |  1 +
 src/highlight.js                        | 31 ++++++++++++++++---------
 test/markup/javascript/regex.expect.txt |  5 ++++
 test/markup/javascript/regex.txt        |  5 ++++
 4 files changed, 31 insertions(+), 11 deletions(-)
 create mode 100644 test/markup/javascript/regex.expect.txt
 create mode 100644 test/markup/javascript/regex.txt

diff --git a/CHANGES.md b/CHANGES.md
index c12ea5f3ed..e64932fdc4 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -10,6 +10,7 @@ New themes:
 
 Core Changes:
 
+- improve regular expression detect (less false-positives) (#2380) [Josh Goebel][]
 - make `noHighlightRe` and `languagePrefixRe` configurable (#2374) [Josh Goebel][]
 
 Language Improvements:
diff --git a/src/highlight.js b/src/highlight.js
index 201a8287f9..bd15c8393b 100644
--- a/src/highlight.js
+++ b/src/highlight.js
@@ -1102,17 +1102,26 @@ https://highlightjs.org/
     relevance: 0
   };
   hljs.REGEXP_MODE = {
-    className: 'regexp',
-    begin: /\//, end: /\/[gimuy]*/,
-    illegal: /\n/,
-    contains: [
-      hljs.BACKSLASH_ESCAPE,
-      {
-        begin: /\[/, end: /\]/,
-        relevance: 0,
-        contains: [hljs.BACKSLASH_ESCAPE]
-      }
-    ]
+    // this outer rule makes sure we actually have a WHOLE regex and not simply
+    // an expression such as:
+    //
+    //     3 / something
+    //
+    // (which will then blow up when regex's `illegal` sees the newline)
+    begin: /(?=\/[^\/\n]*\/)/,
+    contains: [{
+      className: 'regexp',
+      begin: /\//, end: /\/[gimuy]*/,
+      illegal: /\n/,
+      contains: [
+        hljs.BACKSLASH_ESCAPE,
+        {
+          begin: /\[/, end: /\]/,
+          relevance: 0,
+          contains: [hljs.BACKSLASH_ESCAPE]
+        }
+      ]
+    }]
   };
   hljs.TITLE_MODE = {
     className: 'title',
diff --git a/test/markup/javascript/regex.expect.txt b/test/markup/javascript/regex.expect.txt
new file mode 100644
index 0000000000..21342a7889
--- /dev/null
+++ b/test/markup/javascript/regex.expect.txt
@@ -0,0 +1,5 @@
+<span class="hljs-comment">// examples that have proven problematic in the past</span>
+<span class="hljs-string">`Bad <span class="hljs-subst">${foo / <span class="hljs-number">1000</span>}</span>`</span>
+foo = <span class="hljs-number">2</span> / <span class="hljs-number">2</span> + <span class="hljs-number">2</span> / <span class="hljs-number">2</span>;
+foo = <span class="hljs-regexp">/ 2 + 2 /</span>;
+b = a++ / <span class="hljs-number">2</span>;
diff --git a/test/markup/javascript/regex.txt b/test/markup/javascript/regex.txt
new file mode 100644
index 0000000000..da0ee68a12
--- /dev/null
+++ b/test/markup/javascript/regex.txt
@@ -0,0 +1,5 @@
+// examples that have proven problematic in the past
+`Bad ${foo / 1000}`
+foo = 2 / 2 + 2 / 2;
+foo = / 2 + 2 /;
+b = a++ / 2;