Skip removing nodes in <code> (mozilla#647)

* skip removing nodes in <code> (especially common for comments in codeblocks) * don't limit maxDepth
royalgarter · Nov 23, 2020 · 290724c · 290724c
1 parent 3c83389
commit 290724c
Show file tree

Hide file tree

Showing 4 changed files with 373 additions and 0 deletions.
diff --git a/Readability.js b/Readability.js
@@ -912,6 +912,7 @@ Readability.prototype = {
           if (this.REGEXPS.unlikelyCandidates.test(matchString) &&
               !this.REGEXPS.okMaybeItsACandidate.test(matchString) &&
               !this._hasAncestorTag(node, "table") &&
+              !this._hasAncestorTag(node, "code") &&
               node.tagName !== "BODY" &&
               node.tagName !== "A") {
             this.log("Removing unlikely candidate - " + matchString);
@@ -2038,6 +2039,10 @@ Readability.prototype = {
         return false;
       }
 
+      if (this._hasAncestorTag(node, "code")) {
+        return false;
+      }
+
       var weight = this._getClassWeight(node);
 
       this.log("Cleaning Conditionally", node);

diff --git a/test/test-pages/v8-blog/expected-metadata.json b/test/test-pages/v8-blog/expected-metadata.json
@@ -0,0 +1,8 @@
+{
+  "title": "standalone WebAssembly binaries using Emscripten · V8",
+  "byline": null,
+  "dir": null,
+  "excerpt": "Emscripten now supports standalone Wasm files, which do not need JavaScript.",
+  "siteName": null,
+  "readerable": true
+}