Update of bug #66287 (group groff):
Status: None => Need Info
Assigned to: None => barx
_______________________________________________________
Follow-up Comment #1:
[comment #0 original submission:]
> The files
[http://git.savannah.gnu.org/cgit/groff.git/tree/contrib/mm/groff_mmse.7.man
contrib/mm/groff_mmse.7.man] and
[http://git.savannah.gnu.org/cgit/groff.git/tree/doc/meintro_fr.me.in
doc/meintro_fr.me.in] each have an encoding tag near the end of the file in a
Local Variables block. Unfortunately, preconv looks only at the first two
lines of a file for encoding information.
Only if the file isn't seekable...
> So if the user doesn't explicitly tell preconv what the encoding is for
these file, and none of preconv's other heuristics works, it will guess the
wrong encoding.
I recall attempting this before. The problem is that `preconv` is a
preprocessor and supposed to operate in a pipeline. For it to behave as you
desire, it would need to "sponge" up all of its input before writing it back
out.
I think the status quo is the best we can do for shipped files without heavily
refactoring preconv and potentially doing violence to the pipeline/filter
concept.
> Give preconv a better chance to succeed by putting the encoding tag where
preconv looks for it.
>
> This does split up file information between the top and bottom of the file.
But plenty of .mom files do this;
[http://git.savannah.gnu.org/cgit/groff.git/tree/contrib/mom/examples/sample_docs.mom
contrib/mom/examples/sample_docs.mom] and
[http://git.savannah.gnu.org/cgit/groff.git/tree/doc/automake.mom
doc/automake.mom] are two examples.
>
> To see preconv go awry with the current versions of the files, run this
command in a UTF-8 terminal:
> groff -k -Tutf8 -man contrib/mm/groff_mmse.7.man | fgrep Hagg
Hmm, can't reproduce a problem here with _groff_ 1.23.0 or Git HEAD.
$ ~/groff-stable/bin/groff -k -Tutf8 -man contrib/mm/groff_mmse.7.man | grep
-F Hagg
Jörgen Hägg ⟨jorgen.h...@axis.se⟩, Lund, Sweden
$ ~/groff-HEAD/bin/groff -k -Tutf8 -man contrib/mm/groff_mmse.7.man | grep -F
Hagg
Jörgen Hägg, Lund, Sweden
> and this command in a latin-1 terminal:
> groff -k -Tlatin1 -me -mfr -Ww doc/meintro_fr.me.in | fgrep Babey
...nor this one (in a Latin-1 terminal). Is something breaking preconv on
your system? Can you do some experiments with `preconv -d` and see what it
says?
$ ~/groff-stable/bin/groff -k -Tlatin1 -me -mfr -Ww doc/meintro_fr.me.in |
fgrep Babey
Adapté au français par Grégoire Babey
$ ~/groff-HEAD/bin/groff -k -Tlatin1 -me -mfr -Ww doc/meintro_fr.me.in | fgrep
Babey
Adapté au français par Grégoire Babey
_______________________________________________________
Reply to this item at:
<https://savannah.gnu.org/bugs/?66287>
_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/
signature.asc
Description: PGP signature
