Re: [PATCH pkgs] Texinfo 5.x fixes, etc.

 Ar an dara lá is fiche de mí Bealtaine, scríobh Jerry James: 

 > On Mon, May 12, 2014 at 10:45 AM, Jerry James <[email protected]> wrote:
 > > $ iconv -f iso-2022-jp -t utf-8 skk.texi > skk.texi.utf8
 > > iconv: illegal input sequence at position 154746
 > >
 > > That is line 4779 of skk.texi, right after the "W:" if I am counting
 > > characters correctly.  Any advice on how to proceed?
 > 
 > I've got a framework in place that seems to generate HTML from the
 > info files correctly, except for this problem.  It's not just iconv
 > that is confused by skk.texi, either; XEmacs 21.5 also has problems
 > with it.
 > 
 > Would one of you who reads Japanese please take a look at lines 4771
 > and 4780 of skk.texi?  There is something wrong with those two lines.
 > If I use XEmacs itself to convert this file to UTF-8, then makeinfo
 > still chokes on the results, complaining:

The characters on line 4779 that iconv chokes on are all encoded in JISX
0208, details from #'split-char as so:

((japanese-jisx0208 34 49) (japanese-jisx0208 34 81) (japanese-jisx0208 34 113))

Converted to ku/ten (row/column) indexes, these are:

2 17
2 49
2 81

Looking at page 814 of Ken Lund’s CJKV Information Processing, second
edition (2002), these all have no mappings, they are blank spaces. They also
don’t have mappings in our japanese-jisx0208-1978 character set.

The relevant text in skk.texi is describing how to enter characters by means
of their ku/ten code (type \ within the japanese-skk input method), and
indeed, if I type C-u C-\ japanese-skk RET \ RET XEmacs prints that line
(4779) in the minibuffer, unencodable characters and all.

It’d be reasonable to use replacement characters, something like U+FFFD, for
them in skk.texi, and iconv and makeinfo shouldn’t choke on those. U+3013 is
in JISX0208 and is intended for this purpose, as I understand it. To be
clear, the characters to replace are those after W:, R: and Y:.

I’m not confident that there will be problems with lines 4771 and 4780 of
skk.texi once line 4779 is fixed. Give it a shot.

I’m also irritated the Unicode coding systems don’t error more cleanly on
this, I should have noticed and fixed this long ago. Oh well.

Thanks for the work, Jerry!

 > utf8 "\xE04576" does not map to Unicode at
 > /usr/share/texinfo/Texinfo/Parser.pm line 1909, <FH> line 4771.
 > utf8 "\xE04596" does not map to Unicode at
 > /usr/share/texinfo/Texinfo/Parser.pm line 1909, <FH> line 4771.
 > utf8 "\xE045B6" does not map to Unicode at
 > /usr/share/texinfo/Texinfo/Parser.pm line 1909, <FH> line 4771.
 > skk.texi:4780: misplaced {
 > skk.texi:4780: misplaced }
 > skk.texi:4780: misplaced {
 > skk.texi:4780: misplaced }
 > skk.texi:4780: misplaced {
 > skk.texi:4780: misplaced }
 > make[2]: *** [texi/skk_toc.html] Error 1
 > 
 > And reloading the UTF-8 version of the file in XEmacs results in a
 > bunch of weird characters showing up.  I have no idea what should be
 > on those lines, so I need some help to proceed.  Thanks,

-- 
‘Liston operated so fast that he once accidentally amputated an assistant’s 
fingers along with a patient’s leg, […] The patient and the assistant both 
died of sepsis, and a spectator reportedly died of shock, resulting in the 
only known procedure with a 300% mortality.’ (Atul Gawande, NEJM, 2012)

_______________________________________________
XEmacs-Patches mailing list
[email protected]
http://lists.xemacs.org/mailman/listinfo/xemacs-patches