I'm seeing several problems. I'm working on Linux. Terminal programs are
xfce-terminal and xterm. I built the parse-info branch of Maxima with
SBCL, CMUCL, and Clisp. ./configure options included --enable-lang-pt
and --enable-lang-pt-utf8. Texinfo is Texinfo 5. Default LANG is
LANG=en_US.UTF-8; no LC_<FOO> variables are set.
Same results for xterm as for xfce-terminal. Same results for all three
Lisp implementations, I think (didn't notice any difference, maybe I
didn't look hard enough).
* English documentation has spurious characters before and after
variable names (where quote marks might occur).
* Documentation for some items is truncated. E.g. read_matrix in both
English and Portuguese (UTF-8, didn't try non-UTF-8).
* Off-by-one problem occurs for some items, not others. E.g. writefile
Portuguese (UTF-8) shows documentation for mod_threshold.
* @deffnx items (i.e. more than one function prototype) are displayed
in list shown by ??. Should display only first one (@deffn).
* At present, list shown by ?? shows name of section in parentheses
e.g. pfeformat (Functions and Variables for Display). I think that
context is helpful so I'd rather not give it up.
* With CMUCL + LANG=pt_PT.UTF-8, ? fort causes about 20 lines of
compiler warning message to be emitted (something about "signed word
to integer coercion").
Some problems might be caused by different output of Texinfo 5.
I don't have Texinfo 4 installed so I can't test that at the moment.
Hope this helps,
Robert Dodier