enh(cpp) simpler struct matching with beforeMatch

Resolves #2736.

Author: joshgoebel

src/languages/cpp.js

@@ -21,7 +21,7 @@ export default function(hljs) {
   const DECLTYPE_AUTO_RE = 'decltype\\(auto\\)';
   const NAMESPACE_RE = '[a-zA-Z_]\\w*::';
   const TEMPLATE_ARGUMENT_RE = '<[^<>]+>';
-  const FUNCTION_TYPE_RE = '(' +
+  const FUNCTION_TYPE_RE = '(?!struct)(' +
     DECLTYPE_AUTO_RE + '|' +
     regex.optional(NAMESPACE_RE) +
     '[a-zA-Z_]\\w*' + regex.optional(TEMPLATE_ARGUMENT_RE) +
@@ -233,7 +233,7 @@ export default function(hljs) {
       'atomic_bool atomic_char atomic_schar ' +
       'atomic_uchar atomic_short atomic_ushort atomic_int atomic_uint atomic_long atomic_ulong atomic_llong ' +
       'atomic_ullong new throw return ' +
-      'and and_eq bitand bitor compl not not_eq or or_eq xor xor_eq',
+      'and and_eq bitand bitor compl not not_eq or or_eq xor xor_eq struct',
     built_in: '_Bool _Complex _Imaginary',
     _relevance_hints: COMMON_CPP_HINTS,
     literal: 'true false nullptr NULL'
@@ -403,15 +403,10 @@ export default function(hljs) {
           keywords: CPP_KEYWORDS
         },
         {
-          className: 'class',
-          beginKeywords: 'enum class struct union',
-          end: /[{;:<>=]/,
-          contains: [
-            {
-              beginKeywords: "final class struct"
-            },
-            hljs.TITLE_MODE
-          ]
+          beforeMatch: /\b(enum|class|struct|union)\s+/,
+          keywords: "enum class struct union",
+          match: /\w+/,
+          className: "title"
         }
       ])
   };

src/lib/exts/before_match.js

@@ -0,0 +1,25 @@
+import * as regex from "../regex.js";
+
+// allow beforeMatch to act as a "qualifier" for the match
+// the full match begin must be [beforeMatch][begin]
+export const beforeMatchExt = (mode, parent) => {
+  if (!mode.beforeMatch) return;
+  // starts conflicts with endsParent which we need to make sure the child
+  // rule is not matched multiple times
+  if (mode.starts) throw new Error("beforeMatch cannot be used with starts");
+
+  const originalMode = Object.assign({}, mode);
+  Object.keys(mode).forEach((key) => { delete mode[key]; });
+
+  mode.keywords = originalMode.keywords;
+  mode.begin = regex.concat(originalMode.beforeMatch, regex.lookahead(originalMode.begin));
+  mode.starts = {
+    relevance: 0,
+    contains: [
+      Object.assign(originalMode, { endsParent: true })
+    ]
+  };
+  mode.relevance = 0;
+
+  delete originalMode.beforeMatch;
+};

src/lib/mode_compiler.js

@@ -1,6 +1,7 @@
 import * as regex from './regex.js';
 import { inherit } from './utils.js';
 import * as EXT from "./compiler_extensions.js";
+import { beforeMatchExt } from "./exts/before_match.js";
 import { compileKeywords } from "./compile_keywords.js";
 import { MultiClass } from "./ext/multi_class.js";
 
@@ -286,7 +287,8 @@ export function compileLanguage(language, { plugins }) {
       // do this early so compiler extensions generally don't have to worry about
       // the distinction between match/begin
       EXT.compileMatch,
-      MultiClass
+      MultiClass,
+      beforeMatchExt
     ].forEach(ext => ext(mode, parent));
 
     language.compilerExtensions.forEach(ext => ext(mode, parent));

test/markup/cpp/primitive-types.expect.txt

@@ -1,3 +1,3 @@
 <span class="hljs-keyword">const</span> <span class="hljs-keyword">uint64_t</span> MAX_INT_64;
 
-<span class="hljs-class"><span class="hljs-keyword">struct</span> <span class="hljs-title">position_tag</span>;</span>
+<span class="hljs-keyword">struct</span> <span class="hljs-title hljs-title-class">position_tag</span>;

test/markup/cpp/template_complexity.expect.txt

@@ -1,24 +1,24 @@
-<span class="hljs-keyword">template</span> &lt;<span class="hljs-class"><span class="hljs-keyword">class</span> <span class="hljs-title">T</span>&gt;</span> <span class="hljs-comment">// comment</span>
+<span class="hljs-keyword">template</span> &lt;<span class="hljs-keyword">class</span> <span class="hljs-title hljs-title-class">T</span>&gt; <span class="hljs-comment">// comment</span>
 <span class="hljs-function"><span class="hljs-keyword">auto</span> <span class="hljs-title">foo</span><span class="hljs-params">(T x)</span> </span>{ ... };
 
 <span class="hljs-keyword">namespace</span> impl {
     <span class="hljs-keyword">template</span>&lt;<span class="hljs-keyword">typename</span> T&gt;
-    <span class="hljs-class"><span class="hljs-keyword">struct</span> <span class="hljs-title">is_streamable</span>&lt;</span>T, std::<span class="hljs-keyword">void_t</span>&lt;<span class="hljs-keyword">decltype</span>(std::declval&lt;std::wostream &amp;&gt;() &lt;&lt; std::declval&lt;T&gt;())&gt;&gt; : std::true_type { };
+    <span class="hljs-keyword">struct</span> <span class="hljs-title hljs-title-class">is_streamable</span>&lt;T, std::<span class="hljs-keyword">void_t</span>&lt;<span class="hljs-keyword">decltype</span>(std::declval&lt;std::wostream &amp;&gt;() &lt;&lt; std::declval&lt;T&gt;())&gt;&gt; : std::true_type { };
 }
 
 <span class="hljs-comment">// Disable overload for already valid operands.</span>
-<span class="hljs-keyword">template</span>&lt;<span class="hljs-class"><span class="hljs-keyword">class</span> <span class="hljs-title">T</span>, <span class="hljs-keyword">class</span> =</span> std::<span class="hljs-keyword">enable_if_t</span>&lt;!impl::is_streamable_v&lt;<span class="hljs-keyword">const</span> T &amp;&gt; &amp;&amp; std::is_convertible_v&lt;<span class="hljs-keyword">const</span> T &amp;, std::wstring_view&gt;&gt;&gt;
+<span class="hljs-keyword">template</span>&lt;<span class="hljs-keyword">class</span> <span class="hljs-title hljs-title-class">T</span>, class = std::<span class="hljs-keyword">enable_if_t</span>&lt;!impl::is_streamable_v&lt;<span class="hljs-keyword">const</span> T &amp;&gt; &amp;&amp; std::is_convertible_v&lt;<span class="hljs-keyword">const</span> T &amp;, std::wstring_view&gt;&gt;&gt;
 std::wostream &amp;<span class="hljs-keyword">operator</span> &lt;&lt;(std::wostream &amp;stream, <span class="hljs-keyword">const</span> T &amp;thing)
 {
     <span class="hljs-keyword">return</span> stream &lt;&lt; <span class="hljs-keyword">static_cast</span>&lt;std::wstring_view&gt;(thing);
 }
 
-<span class="hljs-class"><span class="hljs-keyword">enum</span> <span class="hljs-keyword">struct</span> <span class="hljs-title">DataHolder</span> {</span> };
-<span class="hljs-class"><span class="hljs-keyword">enum</span> <span class="hljs-keyword">class</span> <span class="hljs-title">DataThingy</span> {</span> };
-<span class="hljs-class"><span class="hljs-keyword">enum</span> <span class="hljs-keyword">class</span> <span class="hljs-title">Boolean</span> :</span> <span class="hljs-keyword">char</span> {
+<span class="hljs-keyword">enum</span> <span class="hljs-title hljs-title-class"><span class="hljs-keyword">struct</span></span> DataHolder { };
+<span class="hljs-keyword">enum</span> <span class="hljs-title hljs-title-class"><span class="hljs-keyword">class</span></span> DataThingy { };
+<span class="hljs-keyword">enum</span> <span class="hljs-title hljs-title-class"><span class="hljs-keyword">class</span></span> Boolean : <span class="hljs-keyword">char</span> {
     True, False, FileNotFound
 };
 
-<span class="hljs-class"><span class="hljs-keyword">union</span> <span class="hljs-title">Soy</span>
-{</span>
+<span class="hljs-keyword">union</span> <span class="hljs-title hljs-title-class">Soy</span>
+{
 };