Boolean attributes in XPath are stripped of their is, has or uses prefix, although it greatly contributes to readability:
http://github.com/pmd/pmd/blob/master/pmd-core/src/main/java/net/sourceforge/pmd/lang/ast/xpath/AttributeAxisIterator.java#L36-L44
SwitchStatement[@IsDefault = true()] would be much more readable than just @Default = true(), and same for eg @IsArray for dimensionable nodes. The problem is even greater with has, since it's common practice to eg have a hasXX and getXX on the same class, which would be truncated to the same attribute name. If we're going to flesh our the AST APIs, we need to avoid that.
It would be great to preserve the prefix on boolean attributes. The problem here is compatibility, so we would have to keep the two forms along for some time, one being deprecated.
The new behaviour needs to be described in the language specific doc (it's part of the AST API).
⚠️ For ease of migration, we should probably do this on master and expose the attributes in both ways and deprecate the old way.
Boolean attributes in XPath are stripped of their
is,hasorusesprefix, although it greatly contributes to readability:http://github.com/pmd/pmd/blob/master/pmd-core/src/main/java/net/sourceforge/pmd/lang/ast/xpath/AttributeAxisIterator.java#L36-L44
SwitchStatement[@IsDefault = true()]would be much more readable than just@Default = true(), and same for eg@IsArrayfor dimensionable nodes. The problem is even greater withhas, since it's common practice to eg have ahasXXandgetXXon the same class, which would be truncated to the same attribute name. If we're going to flesh our the AST APIs, we need to avoid that.It would be great to preserve the prefix on boolean attributes. The problem here is compatibility, so we would have to keep the two forms along for some time, one being deprecated.
The new behaviour needs to be described in the language specific doc (it's part of the AST API).