Re: [Tracker] html_filter generate a extra file

[Laurent Aguerreche <laurent aguerreche free fr>]

Le dimanche 12 novembre 2006 Ã 21:42 +0000, Jamie McCracken a Ãcrit :
> Laurent Aguerreche wrote:
> > Le dimanche 12 novembre 2006 Ã 18:33 +0000, Jamie McCracken a Ãcrit :
> > > Carles Briansà wrote:
> > > > The html_filter generate a extra file that contains a tags of the
> > > > original file.
> > > >
> > > > Example:
> > > >
> > > > I save a file named debian.html in home directory
> > > > Then the trackerd detect and added correctly
> > > > Later in the directory of debian.html, have a debian_t.txt that contain
> > > > a tags of debian.html file.
> > > >
> > > > I see that the file is generated by htmless program.
> > > okay I will investigate (htmless is not my code)
> >
> > I wonder if we should replace htmless by an internet browser in text
> > mode like links or w3m. Htmless does not seem to know é (which
> > should be replaced by 'Ã') for instance and seems to not treat UTF-8.
> > With links, accentuated characters like 'Ã' become ''e' but I saw it to
> > work correctly with UTF-8 files and not with ISO-8859-1 files.
> > With w3m the extracted text is in UTF8 and there are not things like
> > é, à, etc.
>
> feel free to submit a patch - we should support utf8 so w3m looks best

I sent a new version of the html_filter executable in my previous
mail.  ;-)
But if you meant that I should commit it in CVS, I just cannot do it
currently since I did not receive any answer for my CVS account request
for GNOME.


Laurent.