TortoiseMerge fails to restore Unicode characters

[GoogleCodeExporter]

As reported here:
http://tortoisesvn.tigris.org/ds/viewMessage.do?dsForumId=4061&dsMessageId=30718
44

Steps to reproduce:

1. Commit a file that contains Unicode characters. In my case there was this 
line:

counter.text = "∞"

2. Remove Unicode characters from the line, for example:

counter.text = "42"

3. Right-click the file and select "TortoiseSVN > Diff"
4. Right-click the changed line in the right pane and select "Use other text 
block"
5. Save the file in TortoiseMerge
6. Refresh the file in a text editor

Defect: the line is not unchanged. In my case the line became:

counter.text = "8"

File encoding is UTF-8 without BOM. I checked that the bug is still present in 
nightly build 25160.

Original issue reported on code.google.com by tortoisesvn on 21 Jan 2014 at 7:17

[GoogleCodeExporter]

Problem is that the edited file without the unicode chars is treated as ANSI, 
but the unedited file is in utf8. Since after editing the file in TMerge it is 
still saved as ANSI, the "∞" gets automatically replaced by "8" on saving.

Original comment by tortoisesvn on 21 Jan 2014 at 7:18

[GoogleCodeExporter]

This issue was closed by revision r25161.

Original comment by tortoisesvn on 21 Jan 2014 at 9:23

• Changed state: Fixed

[GoogleCodeExporter]

This issue was closed by revision r25162.

Original comment by tortoisesvn on 21 Jan 2014 at 9:24