Hello,
after many weeks of non-developing, I updated again my wikipedia->static
html code. The code basically works, maybe it needs a few tunings but it's
accetable as a starting point.
The main issue, as I see it, is that it's written as a completely
separated Perl program, that is, it uses none of the PHP code used on the
web site. This means that each change on the PHP side must be replicated
in the Perl program. For example, the new TOC feature is not present now
and should be coded anew.
I don't know if this will be a major problem or not. On one hand, the
Wikipedia structure seems mature enough not to expect big changes. On the
other hand, who knows what the "phase 4" software will bring along.. :-))
I didn't read the mailing list, but I saw some meta page about automatic
map generation from Blue Marble data. Veeeery interesting.
Other issues:
- TeX rendering is absent
- Size: the English wikipedia is now too big to fit into a CD-ROM, without
using compression techniques that would render it much less portable than
simple HTML files. The problem is not file size, but the huge number of
small files. A simple zip or gzip of the whole archive brings it down to
about 150MB, but now you need some sort of installation or browser
program. If images and other media are to be included, a CD-ROM is surely
not enough. A DVD would be OK. Size issues should not hamper the script as
a mirror generation tool.
- Javascript search (for single words) works quite well. It seems that I'm
not able to get string.replace() to work, so multi-word results are wrong.
Any expert on the matter?
- Time: complete running time is about 3 hours on a 1.3 Ghz Athlon, and
will increase as the main database grows. Rewriting it in C or C++ should
help, but I don't feel like it :-)
- Non-English wikipedias should be included too. Modifying the script for
this purpose should be easy.
The code is licensed under the GPL and is available to anyone who
requests it. I can put it somewhere on the meta wikipedia, or in the CVS,
if you think it's better. My time to further work on it in the following
months will be quite limited, so if anyone is willing contribute just say
so.
Cheers,
Alfio