Re: AW: svnlook proplist & unicode characters

On 17.12.2014 11:27, Philip Martin wrote:
> Your mail included svnlook: E160013: Pfad »/a/o¨« existiert nicht and
> what I see is 'o' '0xC2' '0xA8' so the decomposed character U+0308 has
> been converted to U+00A8.

U+0308 is not a decomposed character; it is the combining diaeresis
mark. U+00A8 is the non-combining diaeresis (from the Latin 1 subset),
so something along the line must have performed an incorrect conversion
from the combining to the non-combining character.

The correct decomposed representation of ö is U+006F U+0308, but the
output of svnlook has U+006F U+00A8 ... which means that some process
converted the Unicode string without any awareness of combining characters.

I couldn't guess which process that would be, but on Windows, Subversion
uses the Windows API for character set conversion. On Windows,
characters entered by any standard input method are precomposed. The
most common way for decomposed characters to find their way into the
repository is by using an OSX-based Subversion client to create the paths.

-- Brane

-- Brane
Received on 2014-12-17 14:07:15 CET