enh(cpp) Highlight all function dispatches (#3005)

* convert cpp built-ins to relevance hints (and match ALL dispatches) * fix arduino to use function dispatch also * document the new keywords `_ident` behavior
highlightjs · Mar 7, 2021 · d2f3ab5 · d2f3ab5
1 parent 9d17540
commit d2f3ab5
Show file tree

Hide file tree

Showing 9 changed files with 170 additions and 29 deletions.
diff --git a/CHANGES.md b/CHANGES.md
@@ -12,6 +12,7 @@ New Languages:
 
 Language grammar improvements:
 
+- enh(cpp) Highlight all function dispatches (#3005) [Josh Goebel][]
 - enh(python) support type hints and better type support (#2972) [Josh Goebel][]
 - enh(gml) Add additional GML 2.3 keywords (#2984) [xDGameStudios][]
 - fix(cpp) constructor support for initializers (#3001) [Josh Goebel][]

diff --git a/docs/language-guide.rst b/docs/language-guide.rst
@@ -87,12 +87,14 @@ object, each property of which defines its own group of keywords:
   {
     keywords: {
       keyword: 'else for if while',
-      literal: ['false','true','null']
+      literal: ['false','true','null'],
+      _relevance_only: 'one two three four'
     }
   }
 
 The group name becomes the class name in the generated markup, enabling different
-theming for different kinds of keywords.
+theming for different kinds of keywords.  Any property starting with a ``_`` will
+only use those keywords to increase relevance, they will not be highlighted.
 
 To detect keywords, highlight.js breaks the processed chunk of code into separate
 words — a process called lexing. By default, "words" are matched with the regexp

diff --git a/src/highlight.js b/src/highlight.js
@@ -172,8 +172,14 @@ const HLJS = function(hljs) {
           buf = "";
 
           relevance += keywordRelevance;
-          const cssClass = language.classNameAliases[kind] || kind;
-          emitter.addKeyword(match[0], cssClass);
+          if (kind.startsWith("_")) {
+            // _ implied for relevance only, do not highlight
+            // by applying a class name
+            buf += match[0];
+          } else {
+            const cssClass = language.classNameAliases[kind] || kind;
+            emitter.addKeyword(match[0], cssClass);
+          }
         } else {
           buf += match[0];
         }

diff --git a/src/languages/arduino.js b/src/languages/arduino.js
@@ -13,7 +13,6 @@ export default function(hljs) {
     keyword:
       'boolean byte word String',
     built_in:
-      'setup loop ' +
       'KeyboardController MouseController SoftwareSerial ' +
       'EthernetServer EthernetClient LiquidCrystal ' +
       'RobotControl GSMVoiceCall EthernetUDP EsploraTFT ' +
@@ -24,7 +23,9 @@ export default function(hljs) {
       'WiFiUDP GSM_SMS Mailbox USBHost Firmata PImage ' +
       'Client Server GSMPIN FileIO Bridge Serial ' +
       'EEPROM Stream Mouse Audio Servo File Task ' +
-      'GPRS WiFi Wire TFT GSM SPI SD ' +
+      'GPRS WiFi Wire TFT GSM SPI SD ',
+    _:
+      'setup loop ' +
       'runShellCommandAsynchronously analogWriteResolution ' +
       'retrieveCallingNumber printFirmwareVersion ' +
       'analogReadResolution sendDigitalPortPair ' +
@@ -102,6 +103,7 @@ export default function(hljs) {
   kws.keyword += ' ' + ARDUINO_KW.keyword;
   kws.literal += ' ' + ARDUINO_KW.literal;
   kws.built_in += ' ' + ARDUINO_KW.built_in;
+  kws._ += ' ' + ARDUINO_KW._;
 
   ARDUINO.name = 'Arduino';
   ARDUINO.aliases = ['ino'];

diff --git a/src/languages/cpp.js b/src/languages/cpp.js
@@ -107,6 +107,122 @@ export default function(hljs) {
 
   const FUNCTION_TITLE = regex.optional(NAMESPACE_RE) + hljs.IDENT_RE + '\\s*\\(';
 
+  const COMMON_CPP_HINTS = [
+    'asin',
+    'atan2',
+    'atan',
+    'calloc',
+    'ceil',
+    'cosh',
+    'cos',
+    'exit',
+    'exp',
+    'fabs',
+    'floor',
+    'fmod',
+    'fprintf',
+    'fputs',
+    'free',
+    'frexp',
+    'auto_ptr',
+    'deque',
+    'list',
+    'queue',
+    'stack',
+    'vector',
+    'map',
+    'set',
+    'pair',
+    'bitset',
+    'multiset',
+    'multimap',
+    'unordered_set',
+    'fscanf',
+    'future',
+    'isalnum',
+    'isalpha',
+    'iscntrl',
+    'isdigit',
+    'isgraph',
+    'islower',
+    'isprint',
+    'ispunct',
+    'isspace',
+    'isupper',
+    'isxdigit',
+    'tolower',
+    'toupper',
+    'labs',
+    'ldexp',
+    'log10',
+    'log',
+    'malloc',
+    'realloc',
+    'memchr',
+    'memcmp',
+    'memcpy',
+    'memset',
+    'modf',
+    'pow',
+    'printf',
+    'putchar',
+    'puts',
+    'scanf',
+    'sinh',
+    'sin',
+    'snprintf',
+    'sprintf',
+    'sqrt',
+    'sscanf',
+    'strcat',
+    'strchr',
+    'strcmp',
+    'strcpy',
+    'strcspn',
+    'strlen',
+    'strncat',
+    'strncmp',
+    'strncpy',
+    'strpbrk',
+    'strrchr',
+    'strspn',
+    'strstr',
+    'tanh',
+    'tan',
+    'unordered_map',
+    'unordered_multiset',
+    'unordered_multimap',
+    'priority_queue',
+    'make_pair',
+    'array',
+    'shared_ptr',
+    'abort',
+    'terminate',
+    'abs',
+    'acos',
+    'vfprintf',
+    'vprintf',
+    'vsprintf',
+    'endl',
+    'initializer_list',
+    'unique_ptr',
+    'complex',
+    'imaginary',
+    'std',
+    'string',
+    'wstring',
+    'cin',
+    'cout',
+    'cerr',
+    'clog',
+    'stdin',
+    'stdout',
+    'stderr',
+    'stringstream',
+    'istringstream',
+    'ostringstream'
+  ];
+
   const CPP_KEYWORDS = {
     keyword: 'int float while private char char8_t char16_t char32_t catch import module export virtual operator sizeof ' +
       'dynamic_cast|10 typedef const_cast|10 const for static_cast|10 union namespace ' +
@@ -120,19 +236,27 @@ export default function(hljs) {
       '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',
-    built_in: 'std string wstring cin cout cerr clog stdin stdout stderr stringstream istringstream ostringstream ' +
-      'auto_ptr deque list queue stack vector map set pair bitset multiset multimap unordered_set ' +
-      'unordered_map unordered_multiset unordered_multimap priority_queue make_pair array shared_ptr abort terminate abs acos ' +
-      'asin atan2 atan calloc ceil cosh cos exit exp fabs floor fmod fprintf fputs free frexp ' +
-      'fscanf future isalnum isalpha iscntrl isdigit isgraph islower isprint ispunct isspace isupper ' +
-      'isxdigit tolower toupper labs ldexp log10 log malloc realloc memchr memcmp memcpy memset modf pow ' +
-      'printf putchar puts scanf sinh sin snprintf sprintf sqrt sscanf strcat strchr strcmp ' +
-      'strcpy strcspn strlen strncat strncmp strncpy strpbrk strrchr strspn strstr tanh tan ' +
-      'vfprintf vprintf vsprintf endl initializer_list unique_ptr _Bool complex _Complex imaginary _Imaginary',
+    built_in: '_Bool _Complex _Imaginary',
+    _relevance_hints: COMMON_CPP_HINTS,
     literal: 'true false nullptr NULL'
   };
 
+  const FUNCTION_DISPATCH = {
+    className: "function.dispatch",
+    relevance: 0,
+    keywords: CPP_KEYWORDS,
+    begin: regex.concat(
+      /\b/,
+      /(?!decltype)/,
+      /(?!if)/,
+      /(?!for)/,
+      /(?!while)/,
+      hljs.IDENT_RE,
+      regex.lookahead(/\s*\(/))
+  };
+
   const EXPRESSION_CONTAINS = [
+    FUNCTION_DISPATCH,
     PREPROCESSOR,
     CPP_PRIMITIVE_TYPES,
     C_LINE_COMMENT_MODE,
@@ -141,6 +265,7 @@ export default function(hljs) {
     STRINGS
   ];
 
+
   const EXPRESSION_CONTEXT = {
     // This mode covers expression context where we can't expect a function
     // definition and shouldn't highlight anything that looks like one:
@@ -256,9 +381,13 @@ export default function(hljs) {
     ],
     keywords: CPP_KEYWORDS,
     illegal: '</',
+    classNameAliases: {
+      "function.dispatch": "built_in"
+    },
     contains: [].concat(
       EXPRESSION_CONTEXT,
       FUNCTION_DECLARATION,
+      FUNCTION_DISPATCH,
       EXPRESSION_CONTAINS,
       [
         PREPROCESSOR,

diff --git a/test/markup/cpp/function-declarations.expect.txt b/test/markup/cpp/function-declarations.expect.txt
@@ -1,5 +1,5 @@
-<span class="hljs-function"><span class="hljs-keyword">decltype</span>(<span class="hljs-keyword">auto</span>) <span class="hljs-title">look_up_a_string_1</span><span class="hljs-params">()</span> </span>{ <span class="hljs-keyword">return</span> lookup1(); }
-<span class="hljs-function"><span class="hljs-keyword">void</span> <span class="hljs-title">look_up_a_string_2</span><span class="hljs-params">()</span> </span>{ <span class="hljs-keyword">return</span> lookup2(); }
+<span class="hljs-function"><span class="hljs-keyword">decltype</span>(<span class="hljs-keyword">auto</span>) <span class="hljs-title">look_up_a_string_1</span><span class="hljs-params">()</span> </span>{ <span class="hljs-keyword">return</span> <span class="hljs-built_in">lookup1</span>(); }
+<span class="hljs-function"><span class="hljs-keyword">void</span> <span class="hljs-title">look_up_a_string_2</span><span class="hljs-params">()</span> </span>{ <span class="hljs-keyword">return</span> <span class="hljs-built_in">lookup2</span>(); }
 <span class="hljs-function"><span class="hljs-keyword">friend</span> <span class="hljs-keyword">void</span> <span class="hljs-title">A::showB</span><span class="hljs-params">(B x)</span> </span>{}
 <span class="hljs-function"><span class="hljs-keyword">friend</span> <span class="hljs-keyword">void</span> <span class="hljs-title">showB</span><span class="hljs-params">(B x)</span> </span>{}
 <span class="hljs-function"><span class="hljs-keyword">friend</span> <span class="hljs-keyword">void</span> <span class="hljs-title">showB</span><span class="hljs-params">(B::SomeType x)</span> </span>{}
@@ -17,7 +17,8 @@
 <span class="hljs-function"><span class="hljs-keyword">template</span>&lt;<span class="hljs-keyword">typename</span> T...&gt;
 <span class="hljs-keyword">void</span> <span class="hljs-title">foo</span><span class="hljs-params">(T... args)</span> </span>{}
 
-test();
+<span class="hljs-built_in">test</span>();
 
 <span class="hljs-function"><span class="hljs-keyword">void</span> <span class="hljs-title">A</span><span class="hljs-params">()</span>: a(<span class="hljs-number">10</span>) {</span>}
 <span class="hljs-function"><span class="hljs-keyword">explicit</span> <span class="hljs-title">A</span><span class="hljs-params">()</span>: a(<span class="hljs-number">10</span>) {</span>}
+
diff --git a/test/markup/cpp/function-title.expect.txt b/test/markup/cpp/function-title.expect.txt
@@ -1,9 +1,9 @@
 <span class="hljs-function"><span class="hljs-keyword">int</span> <span class="hljs-title">main</span><span class="hljs-params">()</span> </span>{
-  A a = <span class="hljs-keyword">new</span> A();
-  <span class="hljs-keyword">int</span> b = b * sum(<span class="hljs-number">1</span>, <span class="hljs-number">2</span>);
-  <span class="hljs-keyword">if</span> (a-&gt;check1())
+  A a = <span class="hljs-keyword">new</span> <span class="hljs-built_in">A</span>();
+  <span class="hljs-keyword">int</span> b = b * <span class="hljs-built_in">sum</span>(<span class="hljs-number">1</span>, <span class="hljs-number">2</span>);
+  <span class="hljs-keyword">if</span> (a-&gt;<span class="hljs-built_in">check1</span>())
     <span class="hljs-keyword">return</span> <span class="hljs-number">3</span>;
-  <span class="hljs-keyword">else</span> <span class="hljs-keyword">if</span> (a-&gt;check2())
+  <span class="hljs-keyword">else</span> <span class="hljs-keyword">if</span> (a-&gt;<span class="hljs-built_in">check2</span>())
     <span class="hljs-keyword">return</span> <span class="hljs-number">4</span>;
-  <span class="hljs-keyword">return</span> a-&gt;result();
+  <span class="hljs-keyword">return</span> a-&gt;<span class="hljs-built_in">result</span>();
 }
diff --git a/test/markup/cpp/preprocessor.expect.txt b/test/markup/cpp/preprocessor.expect.txt
@@ -20,11 +20,11 @@
 
 <span class="hljs-keyword">if</span> (p) {
 <span class="hljs-meta">#<span class="hljs-meta-keyword">ifdef</span> DEBUG</span>
-    onething();
+    <span class="hljs-built_in">onething</span>();
 <span class="hljs-meta">#<span class="hljs-meta-keyword">endif</span></span>
 } <span class="hljs-keyword">else</span> <span class="hljs-keyword">if</span> (errno) {
 <span class="hljs-meta">#<span class="hljs-meta-keyword">ifdef</span> DEBUG</span>
-    anotherthing();
+    <span class="hljs-built_in">anotherthing</span>();
 <span class="hljs-meta">#<span class="hljs-meta-keyword">endif</span></span>
 }
 

diff --git a/test/markup/cpp/template_complexity.expect.txt b/test/markup/cpp/template_complexity.expect.txt
@@ -3,14 +3,14 @@
 
 <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, <span class="hljs-built_in">std</span>::<span class="hljs-keyword">void_t</span>&lt;<span class="hljs-keyword">decltype</span>(<span class="hljs-built_in">std</span>::declval&lt;<span class="hljs-built_in">std</span>::wostream &amp;&gt;() &lt;&lt; <span class="hljs-built_in">std</span>::declval&lt;T&gt;())&gt;&gt; : <span class="hljs-built_in">std</span>::true_type { };
+    <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-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> <span class="hljs-built_in">std</span>::<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; <span class="hljs-built_in">std</span>::is_convertible_v&lt;<span class="hljs-keyword">const</span> T &amp;, <span class="hljs-built_in">std</span>::wstring_view&gt;&gt;&gt;
-<span class="hljs-built_in">std</span>::wostream &amp;<span class="hljs-keyword">operator</span> &lt;&lt;(<span class="hljs-built_in">std</span>::wostream &amp;stream, <span class="hljs-keyword">const</span> T &amp;thing)
+<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;
+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;<span class="hljs-built_in">std</span>::wstring_view&gt;(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> };