[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
update on XPath parser work
I am now working on the XPath parser.
I have abandoned SableCC. It has an ideal feature
set for XML work, but fails to complete its
code generation on even a partial grammar.
Really too bad---I hope they will continue to
work on it.
I have (once again) looked over using Metamata
tools: for now, I am going to pursue other alternatives
because the old JavaCC aparrently has problems with
Unicode and the new mparse can't product AST, yet.
Also, mparse doesn't generate a complete set of .java files:
alot of the work is actually done via binary class files
Yes they are free, but with no source, debugging
or maintenance is all but impossible. "Been there,
done that": open source isn't just about free, its about
sustainability and maintainability.
So, now its down to a hand-written parser. I have
just downloaded the gnu.regexp to see what it might
provide. I also have a bunch of parser and scanner
code that is XML specific that I can use.
So far, most of the XML-related projects I have seen
have had a hand-written parser that unfortunately was
fairly-well "spaghetti'd into the overall code--but I
am still looking at some more implementations.
Since this XPath may have a number of different uses in Ozone
I want it to be have a clean, reuseable interface.
The hard work is going to be in actually selecting
XML tree nodes based on the parsed Xpath .
Comments greatly appreciated.
Best Wishes,
Ann Tecklenburg