Re: revert-buffer-with-coding-system fails to work for CP1250

[Kenichi Handa]

In article <address@hidden>, Peter Dyballa <address@hidden> writes:

>>> The problems with CP1250 and CP1251 are solved with this patch. No
>>> other failure could be found, except that there is just one
>>> difference, with and without the patch: a thai-tis620 encoded buffer
>>> (without NO BREAK SPACE) shows Thai glyphs, but when the same
>>> contents is reverted to iso-8859-11 (which is thai-tis620 with NO
>>> BREAK SPACE) only empty boxes are shown.
[...]
> I launch both Emacsen with -Q. Then: C-x d <my tests dir> RET. Some  
> cursor movements to position the cursor on the file. Then: C-x RET c  
> tis620 RET v, or C-x RET c iso-8859-11 RET v.

Ok, as in your operation above, revert-buffer is not used.
So the reason you can't see Thai characters by visiting with
coding system iso-8859-11 is not the problem of
revert-bufffer.  Perhaps, your iso10646-1 font doesn't
contain Thai glyphs.

Coding systems defined in code-pages decode non-ASCII
characters into charsets mule-unicode-xxxx-yyyy (including
Thai characters).  But, tis620 decodes Thai characters into
the charset thai-tis620.

> These are my proper encoding files. The test files were generated by  
> removing the first lines in vi(m) and saving under the same name in a  
> "puristic" directory inside my test cases directory:

"Removing the first lines" of whic file?

> [2 ISO 8859-10.txt <text/plain (quoted-printable)>]

??? This is ISO 8859-10 file, not 8859-11 file.  But, if the
problem is because of font, you don't have to resend
8859-11.

---
Kenichi Handa
address@hidden