On Monday, 17. February 2003 14:37, you wrote: > > >> Look at LibXSLT.pm for example - the code used to parse the XML is > > >> used all over AxKit in several places. We should either let get dom do > > >> that work (parsing the FH or strref if neccessary), and let all others > > >> only call get dom instead of trying several methods, or provide other > > >> helper methods in case get dom is not able to exactly replace the > > >> current code. > > > > > > Brilliant idea! +1 from me. > > > > +1 here, too. A lot less messy that way. > > Aren't we implying that things always then have to be in memory? That was > the idea behind get fh() in the first place - some Language modules or > other parts of axkit might be happier with a filehandle than a DOM or > string.
We don't prevent them doing it their way. But fact is that most, if not all, current language modules just parse the DOM exactly like LibXSLT.pm does it. What those really need is just a simple "sub get_dom". By putting the code into Provider.pm, every provider gets a decent get_dom and we can remove the current mess from most lanuguage modules. -- CU Joerg PGP Public Key at http://ich.bin.kein.hoschi.de/~trouble/public_key.asc PGP Key fingerprint = D34F 57C4 99D8 8F16 E16E 7779 CDDC 41A4 4C48 6F94
