Re: SVN support under Windows.



On 21 December 2015 at 23:50, Gilboa Davara <gilboad gmail com> wrote:
On Fri, Dec 11, 2015 at 10:31 PM, Kai Willadsen <kai willadsen gmail com> wrote:
On 11 December 2015 at 00:48, Gilboa Davara <gilboad gmail com> wrote:
The issue is indeed MSys2 related.
For some odd reason, Meld *always* gets itself (...) as the first
command line parameter (argv[1]).

What's argv[0] in this case? python?

I can't find any indication that this is how msys is supposed to work.
It would be good to know whether we're doing something to mess it up,
or whether it's doing odd things. If this is reliable and detectable
behaviour then I'm not above adding a hack for it to the launcher
script, but obviously it would be better not to.

cheers,
Kai

Sorry for the late reply.
I somehow missed your message.

Something is seriously weird here.
I added prints to main, and the sys.argv seems valid.

Logging code:

+print("ArgV size: " + str(len(sys.argv)))
+print("ArgV0: \"" + sys.argv[0] + "\"")
+print("ArgV1: \"" + sys.argv[1] + "\"")

Output:

trunk $ /mingw64/bin/meld .
Couldn't bind the translation domain. Some translations won't work.
[Error 126] The specified module could not be found
ArgV size: 2
ArgV0: "C:/Applications/Development/MSys64/mingw64/bin/meld"
ArgV1: "."

Alright this seems pretty odd.

It's possible that our optparse.OptionParser subclass is causing
something weird to happen, or conceivably GTK+ isn't passing along the
command line as we'd expect. If you're still looking at this, I'd
suggest adding some debugging to MeldApp.parse_args() to see whether
command_line is sane, and whether something odd is happening when we
actually parse.

Specifically, we do:
    rawargs = command_line.get_arguments()[1:]
and then:
    options, args = parser.parse_args(rawargs)

both of which *could* be doing unexpected things.

cheers,
Kai


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]