[vf] Multiple parser bug fixes #340
Conversation
jsotuyod
changed the title
| @@ -70,7 +70,7 @@ PARSER_END(VfParser) | |||
| | <#XMLNAME: (<ALPHA_CHAR> | "_" | ":") (<IDENTIFIER_CHAR>)* > | |||
| | <#QUOTED_STRING_NO_BREAKS: ( "'" ( ~["'", "\r", "\n"] )* "'" ) | |||
| | ( "\"" ( ~["\"", "\r", "\n"] )* "\"" ) > | |||
| | <#QUOTED_STRING: ( "'" ( ~["'"] )* "'" ) | ( "\"" ( ~["\""] )* "\"" ) > | |||
| | <#QUOTED_STRING: ( "'" ( ~["'"] )* "'" ) | ( "\"" ( ~["\""] | "\\\"" )* "\"" ) > | |||
There was a problem hiding this comment.
shouldn't you change ( "'" ( ~["'"] )* "'" ) to ( "'" ( ~["'"] | "\\'" )* "'" ) to allow strings with single quotes including escaped quotes?
There was a problem hiding this comment.
@jsotuyod This would be superfluous, VisualForce rewrites all double quotes in the source code into single quotes for the resulting page. Putting something like <abc ab=' \' '>abc</abc> causes a compilation error and <abc ab=' \' >abc</abc> does not.
There was a problem hiding this comment.
Ok, not what I expected, but if vf works that way...
|
All comments (both XML / HTML and EL) are currently part of the generated AST... this is usually not the case on grammars, moving them to special tokens which are ignored. Completely outside of this PR's scope, but is there any reason to not want it that way? |
|
You are absolutely right, I'm going to pull the new AST node and the other AST node. |
|
@sgorbaty awesome, make it on a different PR, I'll be merging this soon enough |
|
Oops. @jsotuyod, I just pushed it in. |
|
@sgorbaty lol! Don't worry, that was way faster than expected. |
|
Merged! Will be included in PMD 5.6.0 |
Fixing several VF parser bugs:
Added additional parser unit tests.