On 19 Dec 2001 at 13:19, Hrvoje Niksic wrote:

> "Ian Abbott" <[EMAIL PROTECTED]> writes:
> 
> >> Does it?  For me this command retrieves only `index.html' and
> >> `a.html', and that's a bug.  `-i list' makes no different.
> > 
> > Well that's how it behaved for me
> 
> That's really strange, because the code that implements `-i' uses the
> same logic as the code that downloads command-line URLs in main().
> When I tried to repeat the behavior and failed, I wrongly assumed it
> was an oversight on your part.

> So the question is: what are we doing different?

After a bit more investigation, I now need to disclose more about
my test, as what I thought was an irrelevant detail turns out not
to be!

Although I said I used the following command line:

   wget -r -l 1 http://somesite/index.html http://somesite/a.html

what I actually used was more like the following:

   wget -r -l 1 http://somesite/~user/index.html \
   http://somesite/~user/a.html

which resulted in a.html being downloaded twice.

If I replace the ~'s on the command-line with %7E's then it
correctly downloads a.html only once.

For the command-line that used -i file, file contained the ~user
URLs, not %7Euser URLs, and it correctly downloads a.html only
once. If I replace the ~'s in file with %7E's then it still works
correctly.

So this particular bug seems to be a difference in the way URLs
specified on the command-line are transformed, as compared to URLs
specified in a -i file.

Reply via email to