One of the biggest problems with the current Kid implementation is that it compiles Kid templates into Python modules in an intermediate step. This makes tracking down errors in your Kid templates a very annoying thing since the errors reported are in the compiled Python module, and it is not obvious to which location in the original Kid template the erroneous line actually corresponds. Even worse, you usually don't have the Python source code of the compiled templates available, but only the compiled Python file (you can compile to Python source in Kid as well, but this is not the default for performance and other reasons).
Though it surely looked ingenious in the beginning, maybe this was a wrong design decision, since it turned out that it really complicates everything and does not even bring the desired performance advantage. Luckily, this whole problem has been tackled and solved by the Genshi project recently which is an alternative implementation that diverges in some aspects and will supersede Kid eventually. Therefore I don't think we should change the original implementation because then we could better put our efforts into the Genshi project. Instead, I think it is good to keep Kid the way that it is, as a reference implementation and proof of concept, making it more stable, fixing bugs, improving details, and hopefully releasing a stable version 1.0 some time soon. As a step into that direction (and a small Christmas present to all Kid users ;-) I have just implemented (in the SVN trunk) an improved error reporting mechanism that is able to track down errors to their corresponding location in the original Kid (XML) template. I have realized this by adding a debug mode to the codewriter in which it requests location information from the Expat parser and tries to match this with the location of an error that has occurred in the compiled Kid template. The debug mode and error tracking is only and automatically activated once an error really occurred, so it does not slow down the Kid templates in production when everything works fine, and it does not need to be explicitly activated. In Python <2.4 the new error reporting does not work, you'll get the old messages then. This is because Python <2.4 does not have the necessary Expat parser attributes. All unit tests are still working, and I have added some more tests for this new error tracking. Please let me know what you think of this new feature and whether it works for you, since I haven't tested it really heavily yet. -- Christoph ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ kid-template-discuss mailing list kid-template-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/kid-template-discuss
