Comment parsing: parse "<blah" as an HTML tag only if "blah" is a known tag

name.  This should reduce the amount of warning false positives about bad HTML
in comments when the comment author intended to put a reference to a template.
This change will also enable us parse the comment as intended in these cases.

Fixes part 1 of PR13374.


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@162407 91177308-0d34-0410-b5e6-96231b3b80d8

1 files changed, 38 insertions, 2 deletions

diff --git a/lib/AST/CommentLexer.cpp b/lib/AST/CommentLexer.cpp
index b6516ec126..870db2be5f 100644
--- a/lib/AST/CommentLexer.cpp
+++ b/lib/AST/CommentLexer.cpp
@@ -28,6 +28,33 @@ bool isHTMLHexCharacterReferenceCharacter(char C) {
          (C >= 'a' && C <= 'f') ||
          (C >= 'A' && C <= 'F');
 }
+
+bool isHTMLTagName(StringRef Name) {
+  return llvm::StringSwitch<bool>(Name)
+      .Cases("em", "strong", true)
+      .Cases("tt", "i", "b", "big", "small", true)
+      .Cases("strike", "s", "u", "font", true)
+      .Case("a", true)
+      .Case("hr", true)
+      .Cases("div", "span", true)
+      .Cases("h1", "h2", "h3", true)
+      .Cases("h4", "h5", "h6", true)
+      .Case("code", true)
+      .Case("blockquote", true)
+      .Cases("sub", "sup", true)
+      .Case("img", true)
+      .Case("p", true)
+      .Case("br", true)
+      .Case("pre", true)
+      .Cases("ins", "del", true)
+      .Cases("ul", "ol", "li", true)
+      .Cases("dl", "dt", "dd", true)
+      .Cases("table", "caption", true)
+      .Cases("thead", "tfoot", "tbody", true)
+      .Cases("colgroup", "col", true)
+      .Cases("tr", "th", "td", true)
+      .Default(false);
+}
 } // unnamed namespace
 
 StringRef Lexer::resolveHTMLNamedCharacterReference(StringRef Name) const {
@@ -585,8 +612,12 @@ void Lexer::setupAndLexHTMLStartTag(Token &T) {
   assert(BufferPtr[0] == '<' &&
          isHTMLIdentifierStartingCharacter(BufferPtr[1]));
   const char *TagNameEnd = skipHTMLIdentifier(BufferPtr + 2, CommentEnd);
-
   StringRef Name(BufferPtr + 1, TagNameEnd - (BufferPtr + 1));
+  if (!isHTMLTagName(Name)) {
+    formTextToken(T, TagNameEnd);
+    return;
+  }
+
   formTokenWithChars(T, TagNameEnd, tok::html_start_tag);
   T.setHTMLTagStartName(Name);
 
@@ -665,11 +696,16 @@ void Lexer::setupAndLexHTMLEndTag(Token &T) {
 
   const char *TagNameBegin = skipWhitespace(BufferPtr + 2, CommentEnd);
   const char *TagNameEnd = skipHTMLIdentifier(TagNameBegin, CommentEnd);
+  StringRef Name(TagNameBegin, TagNameEnd - TagNameBegin);
+  if (!isHTMLTagName(Name)) {
+    formTextToken(T, TagNameEnd);
+    return;
+  }
 
   const char *End = skipWhitespace(TagNameEnd, CommentEnd);
 
   formTokenWithChars(T, End, tok::html_end_tag);
-  T.setHTMLTagEndName(StringRef(TagNameBegin, TagNameEnd - TagNameBegin));
+  T.setHTMLTagEndName(Name);
 
   if (BufferPtr != CommentEnd && *BufferPtr == '>')
     State = LS_HTMLEndTag;