+[8.50b2] Google search fails with Codepage=0

[MVV]

I have text document in ANSI encoding that contains national characters (Cyrillic in my case).
Google search from Lister works fine with encoding 'As configured for current font'.
However with 'ANSI (local code page)' it passes same text in wrong encoding while text looks 100% identical in both modes.

Required INI values:

Code: Select all

[Lister]
Font1=Courier New,-13,204
Codepage=0

[ghisler(Author)]

I don't have this problem with my PC (using Western encoding), so I will have to check it with Cyrillic.

[ghisler(Author)]

Unfortunately I failed to reproduce it. Here is what I tried:
1. Set the locale for non-unicode programs to Russian
2. Rebooted Windows as requested
3. Opened Lister with some Russian ANSI text
4. Changed encoding to "local code page"
5. Selected text and sent it to Google via right click
-> Firefox was looking for the correct Russian text

So what is wrong? What do I need to choose for the font to reproduce it?

[MVV]

Test file:

Code: Select all

MIME-Version: 1.0 
Content-Type: application/octet-stream; name="1.txt" 
Content-Transfer-Encoding: base64 
Content-Disposition: attachment; filename="1.txt" 

7ODs4CDs++vgIPDg7PM=

Any word may be used for test. It was reproduced in both russian/english Win7 with Russian codepage for non-Unicode programs.

Ah, tested it with clean INI and failed to reproduce. You need to set font in Lister:

Code: Select all

[Lister]
Font1=Courier New,-13,204
Codepage=0

[ghisler(Author)]

Yes, these settings seem to trigger the error, thanks!

Btw, the same problem seems to occur with "Copy to clipboard" in Lister. I will fix it.

[MVV]

Bug still exists in 8.50b3...

BTW it is interesting that it copies correct text if I switch keyboard layout to RU...

[ghisler(Author)]

That's strange, I have just re-tested it with 32-bit and 64-bit version, and it work with English, Russian, German and even Korean keyboard layout.

[MVV]

I tried it with INI that contains only three lines (mentioned in first post).

Sorry, I've tested only text copying.
Google search works fine in 8.50b3.
However text is still copied in wrong encoding (Ctrl+C or context menu) when keyboard layout is EN. Instead of "подразделение" it copies "ïîäðàçäåëåíèå".

[ghisler(Author)]

OK, I will check that - but I'm now using CP_ACP when encoding is set to ANSI. Is this wrong?

[MVV]

It seems that something is wrong. I think you should use same piece of code for both google and text copy (code for google works fine).

[ghisler(Author)]

I thought that I am using the same code, but maybe I made a mistake (I developped this on a virtual computer so I don't need to reboot my main PC).

[ghisler(Author)]

OK, I could reproduce it now in Lister!

[MVV]

Thanks, works in 8.50b4!