Re: --convert-links should not convert # hrefs

[Christopher Gait-Smith]

The site I originally found this bug on was https://extravm.com, using 
the command:

`wget --page-requisites --convert-links https://extravm.com/`

The issue appeared on the first carousel, between lines 120 - 121 in 
index.html.
When the href was changed to "index.html#carouselHero", the slider did 
not work. Changing it back
to just "#carouselHero" made the slider work.

I also noticed that wGET2 does not convert an href="#example" to 
href="site#example"

Besides, there would be no point to explicitly make the href point to 
the page it is on, as it does that by default. The only reason why an 
href with a # should be converted is if its href="newurl#something"


TL;DR href="#example" should not be converted to 
href="file.extension#example" and should remain in the original form. 
href="file.html#example" would need to be converted, though.


On 2022/03/26 20:50, Tim Rühsen wrote:
> On 26.03.22 11:20, Christopher Gait-Smith wrote:
> > When using the --convert-links in wGET, sites will often become 
> > broken if they contain:
> >
> > <a href="#example"></a>
> >
> > Will get converted to:
> >
> > <a href="index.html#example"></a>
> >
> >
> > I believe that hrefs that just have a #whatever should not get 
> > converted to file.html#whatever
>
> I can see how this breaks. Without a file name, wget assumes 
> `index.html`.
> But this file a) doesn't need to exist on the server or b) the default 
> file is something else (e.g. index.php) but index.html also exists on 
> the server.
>
> Do you have one or two examples of pages or domains where this happens ?
>
> Regards, Tim