File Encodings for result file when using TortoiseMerge

Hi all,

i have some issues with the TortoiseMerge picking surprising encodings
for the results of a merge.

We are trying to move some source files from ISO-8859-1 to UTF-8 encoding.

e.g.
/svn/branches/v1/test.py is ISO-8859-1 encoded containing some umlauts
/svn/trunk/test.py is UTF-8 encoded

Now a change (unrelated to the umlauts) happens to test.py in the
ISO-8859-1 branch, and i need to merge.

e.g.: a conflict happens because both sides have modified 'some change',
but the diff is larger due to UTF-8/ISO-8859-1 changes.

<<<<<<< .working
y = "some change 2"
y = "abcchange"
x = "some evil üäößÖÄÜ Umlauts"
=======
y = "some change 1"
x = "some evil üäößÖÄÜ Umlauts"
>>>>>>> .merge-right.r101860

Working Copy is UTF-8, merge-right ISO-8859-1 (or maybe CP1252, doesn't
matter really). This is just fine.

Now running TortoiseMerge from TortoiseSVN to resolve the conflict, the
working copy is on the right, the merge file on the left, encoding is
correctly detected (as seen in the lower bar of the window) and
displayed (e.g. no broken UTF-8 stuff, both umlaut lines look identical
in the left, right and bottom Window).

But when i cherry pick some lines:
y = "some change 1" --> ISO
y = "abcchange" --> UTF-8
x = "some evil üäößÖÄÜ Umlauts" --> UTF-8

and save the file, the merged file contains ISO-8859-1 umlauts, even
though i explictly selected some lines from the UTF-8 (right) side for
merge.

So, merging from Branch -> TRUNK breaks my TRUNK files encoding.

If i do the merge down from TRUNK -> Branch, and pick the UTF-8 side, i
get UTF-8 encoded umlauts as a result. If i pick the ISO-8859-1 side, i
also get UTF-8 encoded umlauts.

So merging from TRUNK -> Branch also breaks my file encodings.

It seems like a deliberate 'feature' to always pick the left sides
encoding, but it is very inconvenient for merging up and down between
branches with differing file encodings.

So, my questions:

1. Is there a way to configure TortoiseSVN/TortoiseMerge to pick the
encoding of the working copy file for the result automatically, or force
an output encoding to use?
2. If not, how can i configure TortoiseSVN/TortoiseMerge to reverse the
order of files easily so it does the right thing (only confusing the
users a little bit)?
3. Bonus Question: Can i force an encoding to use for all diff windows,
so encoding differences become visible?

If it matters:
TortoiseSVN 1.7.12, Build 24070 - 64 Bit , 2013/03/29 08:00:43
Subversion 1.7.9,
apr 1.4.6
apr-utils 1.3.12
neon 0.29.6
OpenSSL 1.0.1e 11 Feb 2013
zlib 1.2.7

against an SVN 1.8 server.

Michael

-- 
Michael Schlenker
Software Architect
CONTACT Software GmbH           Tel.:   +49 (421) 20153-80
Wiener Straße 1-3               Fax:    +49 (421) 20153-41
28359 Bremen
http://www.contact.de/          E-Mail: msc@contact.de
Sitz der Gesellschaft: Bremen
Geschäftsführer: Karl Heinz Zachries, Ralf Holtgrefe
Eingetragen im Handelsregister des Amtsgerichts Bremen unter HRB 13215
------------------------------------------------------
http://tortoisesvn.tigris.org/ds/viewMessage.do?dsForumId=4061&dsMessageId=3064361
To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_tortoisesvn.tigris.org].