Op Maandag 2008-01-28 skryf Charles Brossollet:
> Hi all,
> 
> This is not a question, nor a bug report, but a message to thank the
> lxml team.
> I had a script using ElementTree, that was taking 10s. I tried to
> switch to cElementTree, but this broke my code. So I gave a try to
> lxml, changed nothing to my code except the import statement, and
> zooom, it runs in 1 second.
> 
> Thank you very much for you work, and I hope you will get more
> visibility in the xml python libs, because you deserve it!
> 
> Charles Brossollet

My story is perhaps not that interesting as I switched from minidom (am
I allowed here?). I can't find my benchmarks now, but I guess they are
not really interesting. ElementTree wasn't really an option because of
roundtrip issues.

Speed improved substantially and memory consumption dropped enormously.
Some large XML files could actually be processed for the first time.

Thanks a lot for lxml - it is now used in the latest release of the
Translate Toolkit.

Regards
Friedel