Sérgio Carvalho [2004-10-03 16:11 UTC] It'd be nice to have a link to the source code (.phps) instead of the whole .tgz package. It'd mean an easier code review for everyone (and hence more reviews).
Christian Stocker [2004-10-04 06:36 UTC] I don't know, what I should think of this package. Nice idea, missleading title, wrong implementation of XPath
1) You're not implementing "Not all of the XPath language, but an appropriate subset for what a big XML files is expected to contain.", but a very limited one, which is even wrong.
the get the title of the 2nd entry of my test file, I have to write
/planet/entry/title[2]
while it should be
/planet/entry[2]/title
2) More than one attribute matching doesn't work
3) namespaces are not supported (AFAICS)
4) It exhausted my memory (a 5 MB File hit the 16 MB memlimit, which XPath/PHP5 didn't)
The idea is nice and I certainly see a use for this package, but I think the name is missleading, since it promises too much
Olivier Guilyardi [2004-10-04 11:05 UTC] This is alpha software, please don't expect it to be bugfree or feature-rich.
1) For now, XPath support is limited. The current implementation tries to address the common situations where one want to extract a few nodes from a big XML file.
you wrote :
> the get the title of the 2nd entry of my test file, I have to write
> /planet/entry/title[2]
> while it should be
> /planet/entry[2]/title
This may be a bug. Please mail me some reproduce code, along with an XML data example.
2) About multiple attributes support : this will come. You're asking for a
Olivier Guilyardi [2004-10-04 11:11 UTC] 2) About multiple attributes support : this will come. You're asking for a new feature, and I'll take that in account.
3) Namespacing support is a tricky question I'm currently trying to address.
4) The Expat Parser on PHP5 is buggy (see http://bugs.php.net/bug.php?id=30257). My code implements a workaround, but this is quite heavy.
On PHP4, you should be able to parse 1Gb files with a 16mb limit.
|
