On 11/11/2009 01:35 PM, Pavel Sanda wrote:
hi,
i'm afraid we have broken something wrt configure workings.
when i open the document which has uninstalled latex class,
export and view menu is errorneusly filled "inaccesible" items.
moreover when i do export->plain latex i see error messages:
/usr/bin/python2.6: can't open file 'scripts/elyxer.py': [Errno 2] No such file
or directory
Error: Cannot convert file
----------------------------------------
An error occurred whilst running python -tt scripts/elyxer.py --directory
'/home/pa
i guess there is something fishy in the way we detect elyxer.
I don't see the problems with the export or view menu in current trunk
or branch (just updated). It's still possible I broke something with the
Graph.* changes, which are only in trunk. I know I broke them
temporarily and then had to fix it. ;-) So try updating.
I do know why you're seeing the other error message, though. Since LyX
thinks it can't export LaTeX directly---as JMarc has said, it actually
could, just not DVI, so that needs fixing----LyX is finding the path LyX
--> HTML --> LaTeX, the first via elyxer and the second via html2latex.
The former is failing. I'll see if I can fix this and enable LaTeX
output directly. I don't think we should be exporting LaTeX that way!
The bigger problem is that I don't have elyxer on this machine. What's
happening is that configure.py has this:
path, elyxer = checkProg('a LyX -> HTML converter', ['elyxer.py', 'elyxer'],
rc_entry = [ r'\converter lyx html "python -tt elyxer.py --directory $$r
$$i $$o" ""' ])
and the last entry here---which is the only entry---is being used in the
"not found" case. So it gets written whether you have elyxer or not.
I think this will fix things:
path, elyxer = checkProg('a LyX -> HTML converter', ['elyxer.py',
'elyxer'],
rc_entry = [ r'\converter lyx html "python -tt elyxer.py
--directory $$r $$i $$o" ""' ],
r'\converter lyx html "python -tt elyxer.py --directory
$$r $$i $$o" ""' ],
r'\converter lyx html "" ""' ])
But it needs to be tested this time.
Also, why aren't the LaTeX --> HTML entries being checked if elyxer is
found? It's a different sort of converter, and there could conceivably
be conversion paths for which such a thing was wanted. In trunk,
certainly, the test should still be done, since we hope to allow a
choice between alternate conversion routes. So that "else" should be
removed.
Richard
