Hello, I downloaded with wget 1.7 some pages like this:
wget -N -nH -w 5 -k -K -p \
http://www.example.com/page.htm \
http://www.example.com/anotherpage.htm \
http://www.example.com/thirdpage.htm \
...
http://www.example.com/evenmorepages.htm \
http://www.example.com/andmore.htm
The -k option was supposed to fixup the links so that links from one page to
another page that I also downloaded would work while links that linked to
another page (say, 'differentpage.htm' with an URL of
http://www.example.com/differentpage.htm ) would be converted from relative
(<a href="differentpage.htm">) to absolute (<a
href="http://www.example.com/differentpage.htm";>), so that offline viewing
works correctly: those links that lead to cached pages work fine, and links
that lead to other pages also work.
However, apparently the code handles HTML #anchors incorrectly. Some of
those pages had links that look like <a href="#internalanchor">, which
reference not another page but a place marked with <a
name="internalanchor">...</a>. However, those places were changed to read <a
href="http://www.example.com/#internalanchor";>, which is something quite
different! Instead of linking to an anchor in the current page, the link now
points to an anchor in the http://www.example.com/ (the root page, as it
were).
Please fix the algorithm so that internal anchors (that is, links starting
with #) do not get changed. They should always work since they are always
relative. URLs such as otherpage.htm#anchor appeared to work correctly (they
didn't get a http://www.example.com/ prepended), and I suppose
http://www.example.com/example.htm#anchor would also have worked. It's just
a bare #anchor which seems confused.
Cheers,
Philip
--
Philip Newton <[EMAIL PROTECTED]>
All opinions are my own, not my employer's.
If you're not part of the solution, you're part of the precipitate.
