[PATCH 8 of 8] hgext: make fetch use absolute_import

Pulkit Goyal 7895pulkit at gmail.com
Sun Apr 10 11:03:16 EDT 2016


Okay, well that's a cool way.

On Sun, Apr 10, 2016 at 8:31 PM, timeless <timeless at gmail.com> wrote:
> yes, I use that script to run the tests, and thus at the end of each
> cycle I have globs which are a lot easier to manage.
>
> I didn't integrate it into the test itself because I want to be able
> to review the output and jump to lines.
>
> On Sun, Apr 10, 2016 at 10:59 AM, Pulkit Goyal <7895pulkit at gmail.com> wrote:
>> You mean manually because running the test yield this.
>>
>> On Sun, Apr 10, 2016 at 8:26 PM, timeless <timeless at gmail.com> wrote:
>>> (py)[timeless at gcc2-power8 tests]$ cat ~/bin/utpy3
>>> #!/bin/sh
>>> cd ~/hg/crew/tests/;
>>> perl -ni -e 'next if /^(?:<<<<<<|======|>>>>>>)/; print' test-check-py3-compat.t
>>> ./run-tests.py -l -j120 --with-python3=~/hg/py3/bin/python3.5
>>> test-check-py3-compat.t -i
>>> perl -pi -e 's/i\*n\.py/i18n.py/g;s/(named (.)urllib)\*./${1}2$2/;
>>> next unless /invalid syntax|error importing/; next unless s/\d+/*/g;
>>> s/i\*n\.py/i18n.py/g; s/(named (.)urllib)\*./${1}2$2/;s/ \(glob\)//g;
>>> s/$/ (glob)/' ~/hg/crew/tests/test-check-py3-compat.t
>>>
>>> obviously you would need to adjust the paths.
>>>
>>> On Sun, Apr 10, 2016 at 10:47 AM, Pulkit Goyal <7895pulkit at gmail.com> wrote:
>>>> Okay I will send an another series of same patches with these fixed
>>>> up. One more thing, when I am updating test-check-py3-compat.t, I am
>>>> getting line numbers in the second half in patch numbers 6,7 and 8
>>>> whereas it should be replaced with *.
>>>>
>>>> On Sun, Apr 10, 2016 at 7:43 AM, timeless <timeless at gmail.com> wrote:
>>>>> Pulkit Goyal wrote:
>>>>>>
>>>>>> +    lock,
>>>>>
>>>>> lock as lockmod
>>>>>
>>>>>> +    util,
>>>>>> +)
>>>>>>
>>>>>>  cmdtable = {}
>>>>>>  command = cmdutil.command(cmdtable)
>>>>>> @@ -151,4 +159,4 @@
>>>>>>          return err
>>>>>>
>>>>>>      finally:
>>>>>> -        release(lock, wlock)
>>>>> this almost certainly shouldn't work correctly:
>>>>>> +        lock.release(lock, wlock)


More information about the Mercurial-devel mailing list