DEV: Add support for aliases in HighlightJS languages (#20380)

We were only supporting the main name of each HighlightJS language. So, by default, you could not use `js` or `jsx` to highlight Javascript, given they are aliases for `javascript`.

This PR adds a list of aliases as a constant to core (built via a rake task), and then checks against the `highlighted_languages` site settings plus the list of aliases when processing a code block.

app/assets/javascripts/discourse/tests/unit/lib/pretty-text-test.js

@@ -18,7 +18,7 @@ const rawOpts = {
     enable_mentions: true,
     emoji_set: "twitter",
     external_emoji_url: "",
-    highlighted_languages: "json|ruby|javascript",
+    highlighted_languages: "json|ruby|javascript|xml",
     default_code_lang: "auto",
     enable_markdown_linkify: true,
     markdown_linkify_tlds: "com",
@@ -1776,4 +1776,23 @@ var bar = 'bar';
       "does not loop infinitely"
     );
   });
+
+  test("highlighted aliased languages", function (assert) {
+    // "js" is an alias of "javascript"
+    assert.cooked(
+      "```js\nvar foo ='foo';\nvar bar = 'bar';\n```",
+      `<pre><code class=\"lang-js\">var foo ='foo';
+var bar = 'bar';
+</code></pre>`,
+      "code block with js alias works"
+    );
+
+    // "html" is an alias of "xml"
+    assert.cooked(
+      "```html\n<strong>fun</strong> times\n```",
+      `<pre><code class=\"lang-html\">&lt;strong&gt;fun&lt;/strong&gt; times
+</code></pre>`,
+      "code block with html alias work"
+    );
+  });
 });

app/assets/javascripts/pretty-text/addon/highlightjs-aliases.js

@@ -0,0 +1,123 @@
+// DO NOT EDIT THIS FILE!!!
+// Update it by running `rake javascript:update_constants`
+
+export const HLJS_ALIASES = {
+  bash: ["sh"],
+  c: ["h"],
+  cpp: ["cc", "c++", "h++", "hpp", "hh", "hxx", "cxx"],
+  csharp: ["cs", "c#"],
+  diff: ["patch"],
+  go: ["golang"],
+  graphql: ["gql"],
+  ini: ["toml"],
+  java: ["jsp"],
+  javascript: ["js", "jsx", "mjs", "cjs"],
+  kotlin: ["kt", "kts"],
+  makefile: ["mk", "mak", "make"],
+  xml: [
+    "html",
+    "xhtml",
+    "rss",
+    "atom",
+    "xjb",
+    "xsd",
+    "xsl",
+    "plist",
+    "wsf",
+    "svg",
+  ],
+  markdown: ["md", "mkdown", "mkd"],
+  objectivec: ["mm", "objc", "obj-c", "obj-c++", "objective-c++"],
+  perl: ["pl", "pm"],
+  plaintext: ["text", "txt"],
+  python: ["py", "gyp", "ipython"],
+  "python-repl": ["pycon"],
+  ruby: ["rb", "gemspec", "podspec", "thor", "irb"],
+  rust: ["rs"],
+  shell: ["console", "shellsession"],
+  typescript: ["ts", "tsx"],
+  vbnet: ["vb"],
+  yaml: ["yml"],
+  actionscript: ["as"],
+  angelscript: ["asc"],
+  apache: ["apacheconf"],
+  applescript: ["osascript"],
+  arduino: ["ino"],
+  armasm: ["arm"],
+  asciidoc: ["adoc"],
+  autohotkey: ["ahk"],
+  axapta: ["x++"],
+  brainfuck: ["bf"],
+  "c-like": [],
+  capnproto: ["capnp"],
+  clean: ["icl", "dcl"],
+  clojure: ["clj", "edn"],
+  cmake: ["cmake.in"],
+  coffeescript: ["coffee", "cson", "iced"],
+  cos: ["cls"],
+  crmsh: ["crm", "pcmk"],
+  crystal: ["cr"],
+  delphi: ["dpr", "dfm", "pas", "pascal"],
+  django: ["jinja"],
+  dns: ["bind", "zone"],
+  dockerfile: ["docker"],
+  dos: ["bat", "cmd"],
+  dust: ["dst"],
+  elixir: ["ex", "exs"],
+  erlang: ["erl"],
+  excel: ["xlsx", "xls"],
+  fortran: ["f90", "f95"],
+  fsharp: ["fs", "f#"],
+  gams: ["gms"],
+  gauss: ["gss"],
+  gcode: ["nc"],
+  gherkin: ["feature"],
+  handlebars: ["hbs", "html.hbs", "html.handlebars", "htmlbars"],
+  haskell: ["hs"],
+  haxe: ["hx"],
+  htmlbars: ["hbs", "html.hbs", "html.handlebars", "htmlbars"],
+  http: ["https"],
+  hy: ["hylang"],
+  inform7: ["i7"],
+  "jboss-cli": ["wildfly-cli"],
+  "julia-repl": ["jldoctest"],
+  lasso: ["ls", "lassoscript"],
+  latex: ["tex"],
+  livescript: ["ls"],
+  mathematica: ["mma", "wl"],
+  mercury: ["m", "moo"],
+  mipsasm: ["mips"],
+  moonscript: ["moon"],
+  nestedtext: ["nt"],
+  nginx: ["nginxconf"],
+  nimrod: ["nim"],
+  nix: ["nixos"],
+  ocaml: ["ml"],
+  openscad: ["scad"],
+  pf: ["pf.conf"],
+  pgsql: ["postgres", "postgresql"],
+  powershell: ["pwsh", "ps", "ps1"],
+  processing: ["pde"],
+  puppet: ["pp"],
+  purebasic: ["pb", "pbi"],
+  q: ["k", "kdb"],
+  qml: ["qt"],
+  reasonml: ["re"],
+  roboconf: ["graph", "instances"],
+  routeros: ["mikrotik"],
+  scilab: ["sci"],
+  smalltalk: ["st"],
+  sml: ["ml"],
+  sql_more: ["mysql", "oracle"],
+  stan: ["stanfuncs"],
+  stata: ["do", "ado"],
+  step21: ["p21", "step", "stp"],
+  stylus: ["styl"],
+  tcl: ["tk"],
+  twig: ["craftcms"],
+  vbscript: ["vbs"],
+  verilog: ["v", "sv", "svh"],
+  xl: ["tao"],
+  xquery: ["xpath", "xq"],
+  zephir: ["zep"],
+};

app/assets/javascripts/pretty-text/engines/discourse-markdown/code.js

@@ -1,3 +1,5 @@
+import { HLJS_ALIASES } from "pretty-text/highlightjs-aliases";
+
 // we need a custom renderer for code blocks cause we have a slightly non compliant
 // format with special handling for text and so on
 const TEXT_CODE_CLASSES = ["text", "pre", "plain"];
@@ -77,10 +79,20 @@ function render(tokens, idx, options, env, slf, md) {
 
 export function setup(helper) {
   helper.registerOptions((opts, siteSettings) => {
-    opts.defaultCodeLang = siteSettings.default_code_lang;
+    let languageAliases = [];
-    opts.acceptableCodeClasses = (siteSettings.highlighted_languages || "")
+    const languages = (siteSettings.highlighted_languages || "")
       .split("|")
-      .filter(Boolean)
+      .filter(Boolean);
+
+    languages.forEach((lang) => {
+      if (HLJS_ALIASES[lang]) {
+        languageAliases = languageAliases.concat(HLJS_ALIASES[lang]);
+      }
+    });
+
+    opts.defaultCodeLang = siteSettings.default_code_lang;
+    opts.acceptableCodeClasses = languages
+      .concat(languageAliases)
       .concat(["auto", "plaintext"]);
   });
 

lib/tasks/javascript.rake

@@ -182,6 +182,32 @@ task "javascript:update_constants" => :environment do
     export const replacements = #{Emoji.unicode_replacements_json};
   JS
 
+  langs = []
+  Dir
+    .glob("vendor/assets/javascripts/highlightjs/languages/*.min.js")
+    .each { |f| langs << File.basename(f, ".min.js") }
+  bundle = HighlightJs.bundle(langs)
+
+  ctx = MiniRacer::Context.new
+  hljs_aliases = ctx.eval(<<~JS)
+    #{bundle}
+
+    let aliases = {};
+    hljs.listLanguages().forEach((lang) => {
+      if (hljs.getLanguage(lang).aliases) {
+        aliases[lang] = hljs.getLanguage(lang).aliases;
+      }
+    });
+
+    aliases;
+  JS
+
+  write_template("pretty-text/addon/highlightjs-aliases.js", task_name, <<~JS)
+    export const HLJS_ALIASES = #{hljs_aliases.to_json};
+  JS
+
+  ctx.dispose
+
   write_template("pretty-text/addon/emoji/version.js", task_name, <<~JS)
     export const IMAGE_VERSION = "#{Emoji::EMOJI_VERSION}";
   JS