Re: Malformed UTF-8 character: \xe8\x6e\x69

On Fri, Mar 05, 2021 at 11:21:30PM -0500, Paul Dufresne wrote:
> Been building https://codeload.github.com/hebisch/gpc/zip/master with: 
> 
> ftp://ftp.gnu.org/gnu/gcc/gcc-3.4.6/gcc-core-3.4.6.tar.bz2
>
<snip>
> LANG=hr_HR LC_ALL=hr_HR makeinfo --split-size=5000000?? -I ../../gpc1/gcc-3.4.6/gcc/p/doc -I ../../gpc1/gcc-3.4.6/gcc/p/doc/hr -I ../../gpc1/gcc-3.4.6/gcc/p/doc/generated -I ../../gpc1/gcc-3.4.6/gcc/p/doc/en -o ../../gpc1/gcc-3.4.6/gcc/p/doc/info/gpc-hr.info \
> 
> ?????????????????? ../../gpc1/gcc-3.4.6/gcc/p/doc/hr/gpc.texi
> 
> utf8 "\xE8" does not map to Unicode at /usr/share/texinfo/Texinfo/ParserNonXS.pm line 1796, <FH> line 3.
> 
> Malformed UTF-8 character: \xe8\x6e\x69 (unexpected non-continuation byte 0x6e, immediately after start byte 0xe8; need 3 bytes, got 1) in pattern match (m//) at /usr/share/texinfo/Texinfo/ParserNonXS.pm line 3364.
> 
> Malformed UTF-8 character (fatal) at /usr/share/texinfo/Texinfo/ParserNonXS.pm line 3364.
<snip>
> At first I had tought it was because I did not have hr locale, but I did "sudo dpkg-reconfigure locales" where I included hr locales.

GPC documentation was written using 8-bit codepages.  At time were the
files we written (and even now) on some (many???) systems documentation
utilities worked only in 8-bit mode.  Apparenly your system defaults
everything to UTF-8.  One workaround could be to try to modify

LANG=hr_HR LC_ALL=hr_HR

into something like:

LANG=hr_HR.ISO-8859-2 LC_ALL=hr_HR.ISO-8859-2

I do not know which ISO codepage is correct for hr.

Or just disable documentation build (configure tests for version
of makeinfo, changing test one can disable documentation build).

Checking if documentation build works can be done by testing
corresponding commands by hand (that can save a lot of time
compared to repeated compilation of everthing).

-- 
                              Waldek Hebisch