historic grail python browser "semi-recovered"

[lkcl]

On Jun 10, 6:17 pm, MRAB <pyt... at mrabarnett.plus.com> wrote:
> lkcl wrote:
> > On Jun 9, 11:03 pm, rantingrick <rantingr... at gmail.com> wrote:
> >> On Jun 9, 4:29 pm, lkcl <luke.leigh... at gmail.com> wrote:
>
> >>> um, please don't ask me why but i foundgrail, the python-based web
> >>>browser, and have managed to hack it into submission sufficiently to
> >>> view e.g.http://www.google.co.uk.  out of sheer apathy i happened to
> >>> have python2.4 still installed which was the only way i could get it
> >>> to run without having to rewrite regex expressions (which i don't
> >>> understand).
> >>> if anyone else would be interested in resurrecting this historic web
> >>>browser, just for fits and giggles, please let me know.
> >> Hi lkcl,
>
> >> My current conquest to bring a new (or fix the current GUI) in
> >> Python's stdlib is receiving much resistance. I many need a project to
> >> convince my opponents of my worth. Tell you what i do, send me a text
> >> file with a pathname and all the line numbers that have broken regexs
> >> using a common sep --space is fine for me-- and i'll fix them for you.
> >> Here is a sample...
>
> >  ok i've committed a file REGEX.CONVERSIONS.REQUIRED into the git
> > repository,
> >http://github.com/lkcl/grailbrowser
> > git://github.com/lkcl/grailbrowser.git
>
> >  i used "grep -n" so it's filename:lineno:  {ignore the actual stuff}
>
> >  unfortunately, SGMLLexer.py contains some _vast_ regexs spanning 5-6
> > lines, which means that a simple grep ain't gonna cut it.  there's a
> > batch of regex's spanning from line 650 to line 699 and a few more
> > besides.
>
> >  of course, it has to be borne in mind that this code was written for
> > python 1.5 initially, at a time when python xml/sax/dom/sgml code
> > probably didn't exist.
>
> >  but leaving aside the fact that it all needs to be ripped up and
> > modernised i'm more concerned about getting these 35,000 lines of code
> > operational, doing as small transitions as possible.
>
> The regex module was called 'regex'.

 yes.  there's a python module in 2.4 called reconvert.py which can
"understand" and convert _most_ regex expressions to re.

> I see that the name 're' is used as
> a name in the code.

 bizarre, isn't it?