Ara, your latest modified classes use XJavaDoc.getInstance(), but the CVS
version of XJavaDoc doesn't have a getInstance() method. --> Tons of compile
errors.

I have committed an updated version of the grammar which opens up for a
serious optimisation/refactoring, and will let us remove the JavadocVisitor
alltogether. It will be the parser that populates the SourceClass instances
with information in stead of the JavadocVisitor (which we'll throw away).
After a source file is parsed, we'll have a SourceClass object with
everything initialised.

IMO, the SourceClass should not have any refs to other XClass objects. It
should only have the String names of the classes it points to. the getters
will return an XClass object of course, but the body of the getters will do
a lookup in the singleton, using the member class name as a key.

As you know, my laptop is fucked, and I'm travelling in the week, so I won't
be able to complete this before the end of the week. Feel free to continue
from where I left, and I'll join back in ASAP.

The only reason I can see to use JJTree now is for printing the whole source
back after it's modified, but I think we can get around that too. It would
be nice to throw away all the visitors. It's a nice pattern okay, but it
slows things down and adds complexity (in our case).

Have fun!


_______________________________________________
Xdoclet-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/xdoclet-devel

Reply via email to