hiya, sorry to bother... I've been consistenly recompiling the newest
wget with slight change in the file "url.c", mainly because of the
default "non-clobber" behaviour. I'm sure there's been lots of
controversial discussion already on this, so I'll be quick.

the site I am trying to mirror is a shopping cart "template" based
system, which has re-occuring filenames for different pages. I use the
following command:

./wget -r -p -k -nH -nd -w0.8 -P/home/www/static/aug-05/mirror/ -linf -K
-U"Mozilla/4.0+(compatible;+MSIE+6.0;+Windows+NT+5.0)"
-I/images2,/cgi-bin/lf.storefront -X/cgi-bin/lf.storefront/*/Offer/*
-oaug-spider.log -b http://dev.lookfantastic.com/cgi-bin/lf.storefront/

because of the re-occuring endings in the URI's, this command will fail
to save some pages, due to the way non-clobber works and that all files
are saved in the same folder.

My recompiled wget ignores the --recursive non-clobber suppression, and
saves files appended with a number, though it would be nice to have some
extra combination of commands to re-anable it.

Is there a useful way around this 'default' behaviour, without
replicating the directory tree locally?

regards

- Niel

Reply via email to