simon-git: putty (main): Simon Tatham

[Commits to Tartarus hosted VCS]

TL;DR:
  a216d861 Windows mb_to_wc: support internal SBCSes.

Repository:     https://git.tartarus.org/simon/putty.git
On the web:     https://git.tartarus.org/?p=simon/putty.git
Branch updated: main
Committer:      Simon Tatham <anakin at pobox.com>
Date:           2022-05-31 13:23:49

commit a216d86106d40c38f05f1ffc03996be54d590aa6
web diff https://git.tartarus.org/?p=simon/putty.git;a=commitdiff;h=a216d86106d40c38f05f1ffc03996be54d590aa6;hp=eb7f5aff5c2234296b9ec69675e02aefa6bb773d
Author: Simon Tatham <anakin at pobox.com>
Date:   Tue May 31 13:13:57 2022 +0100

    Windows mb_to_wc: support internal SBCSes.
    
    A user points out that the new charset-aware window title setting
    doesn't work if the configured character set is one of the entries in
    cp_list[] based on a hard-coded Unicode translation table, such as the
    ISO 8859 family.
    
    That's because the Windows mb_to_wc() function assumes that the code
    page it's given will always be OK to pass to the Windows API function
    MultiByteToWideChar, forgetting that for those internally implemented
    single-byte character sets are not.
    
    This commit adds a manual implementation of SBCS -> Unicode based on
    those tables, which restores the ability to set a window title
    specified in ISO 8859.
    
    However, it's not a full fix to windows/unicode.c in general, because
    wc_to_mb has a similar blind spot: it's only prepared to convert
    Unicode to an internally implemented SBCS if that SBCS happens to be
    the one currently set in ucsdata->line_codepage, because that's when
    we've already prepared the reverse lookup table. Probably we ought to
    sort that out, and arrange that it can make the reverse lookup table
    if suddenly called on to do a different conversion. But that needs
    more refactoring, so I haven't done it in this commit.

 windows/unicode.c | 29 +++++++++++++++++++++++++++++
 1 file changed, 29 insertions(+)