Debian LMDE + Conky + gcalcli = issue

[ghost]

Hello

After switching from LinuxMint to LinuxMint Debian Edition I've got a problem with gcalcli.
After little investigation I've founded cause (copied below).

I'll be much appreciated for any help
Best regards to all.

My error:
Conky: desktop window (10000a5) is subwindow of root window (be)
Conky: window type - override
Conky: drawing to created window (0x2400001)
Conky: drawing to double buffer
Traceback (most recent call last):
File "/usr/bin/gcalcli", line 1529, in
BowChickaWowWow()
File "/usr/bin/gcalcli", line 1474, in BowChickaWowWow
gcal.CalQuery(args[0], count=int(args[1]))
File "/usr/bin/gcalcli", line 1007, in CalQuery
self._GraphEvents(cmd, start, count, eventList)
File "/usr/bin/gcalcli", line 692, in _GraphEvents
PrintMsg(CLR_NRM(), line + "\n")
File "/usr/bin/gcalcli", line 232, in PrintMsg
sys.stdout.write(msg)
UnicodeEncodeError: 'ascii' codec can't encode character u'\u0144' in position 95: ordinal not in range(128)

[insanum]

Did you try the latest code in the git master branch?

[ghost]

Yes.
I've copied content of https://raw.github.com/insanum/gcalcli/master/gcalcli, but still have the same error.

Oh, just checked right now and from command line i've good response (good calendar) - only when i run gcalcli from conky script i have that issue.

[ghost]

OK, so developer wasn't very helpful and I had to find workaround:

1. git clone http://www.tablix.org/~avian/git/unidecode.git
2. cd unidecode
3. sudo python setup.py install
4. sudo vi /usr/bin/gcalcli
5. above “from Queue import Queue” add “from unidecode import unidecode”
6. find “def PrintMsg(color, msg)”, and else statement should look like this "sys.stdout.write(unidecode(msg))"

All thanks goes to "Gilles Lenfant" (http://www.peterbe.com/plog/unicode-to-ascii#c358f0d)

[BeniaminK]

Or in PrintMsg function, after every msg one should encode to utf-8, so msg.encode("utf-8") should work properly.

[Unkuiri]

Following what is said here: http://stackoverflow.com/questions/8016236/python-unicode-handling-differences-between-print-and-sys-stdout-write
"A better solution might be to replace sys.stdout with a wrapper". This is accomplished by adding this:

class StdOut(object):
    def write(self, string):
        if isinstance(string, unicode):
            string = string.encode(sys.__stdout__.encoding)
        sys.__stdout__.write(string)

sys.stdout = StdOut()

to the part just next to the import part, for example on line 62 or 63 and on the def PrintMsg(color, msg) part the else statement should look like sys.stdout.write(msg)

Hope this helps

[insanum]

@ihatelogin don't you just hate it when the developer isn't helpful? Sucks I know, but sometimes those developers are just way to busy and simple python scripts are never a priority.

There was a new "--locale" option added to gcalcli. That might help address this issue. If not it is encouraged you take some time to create a github pull request that contains a proper fix.

[tresni]

This is probably the same issue I talked about over here in #8 and is likely fixed with b713a8c . Right now I've tagged that as 2.5 milestone as I really need more people then just me to test it. I'm actually inclined to mark this as a duplicate.

[insanum]

Got an email that unicode in the month names break the calw and calm output. Don't know if @tresni fix fixes this problem.

[tresni]

Probably not, though I think I know where the issue is. I believe the east_asian_width calls only apply to descriptions, not month/day names so that could definitely be an issue.

[insanum]

We should probably filter all terminal output through a single function that does the proper conversions. This unicode stuff has plagued gcalcli for quite some time now. Maybe unifying the output path will help. Thoughts?

[tresni]

Agree. I've found a couple other odd output errors. If I set my output width too big (variable depending on terminal window size) and have more one meeting in the day, I get newlines in the output where it shouldn't be...