bug-lilypond
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Issue 1522 in lilypond: Windows - problem with accented characters i


From: lilypond
Subject: Re: Issue 1522 in lilypond: Windows - problem with accented characters in file name
Date: Mon, 28 Feb 2011 03:01:16 +0000


Comment #4 on issue 1522 by address@hidden: Windows - problem with accented characters in file name
http://code.google.com/p/lilypond/issues/detail?id=1522

Confirming comment 3, we can avoid this issue by un-doing the part of commit 11d5b22 that partially fixed issue 1429. This change involves only .scm code, so if other windows users want to test, they can apply the attached diff by hand. Of course the flashing DOS box of issue 1429 comes back.

Should go back to the old way, live with flashing DOS boxes in 2.14, and make spawn() work on windows during 2.15 ?

If we want to use spawn(), the problem appears to be that MS windows gives us wchar filenames, but glib's spawn wants utf8 input, because: 1) The error message matches the one given by glib's gspawn-win32.c, when it tries to convert from UTF8 to wchar. 2) Output on a DOS box shows non-ANSI characters as if they were utf16 encoded. That is, u-umlaut U+00FC shows as a raised 'n', xFC in the default MS-DOS code-page. 3) Hard-coding filenames for the command-line directly into backend-library.scm, and saving the .scm file as utf-8, gives the desired .pdf output file.

This might be solved by inserting a utf-8 normalization function somewhere, but I do not know Scheme well enough to try that myself.


Attachments:
        1522.txt  794 bytes




reply via email to

[Prev in Thread] Current Thread [Next in Thread]