These are chat archives for coala/coala-bears

31st
Jan 2019
Amol Grover
@frextrite
Jan 31 11:25

Currently we are using xenial repos. Read .travis.yml and .ci/deps.* to check this

We're still on trusty right? No mention of xenial images anywhere.

Amol Grover
@frextrite
Jan 31 11:58
Plus, how can we trigger a travis build with updated .travis.yml?
Any method that doesn't include creating a PR?
Abhinav Kaushlya
@abhishalya
Jan 31 12:21
@frextrite You can change the build config and run for your branch
You need to enable your fork on Travis
Amol Grover
@frextrite
Jan 31 12:35

Did a quick google search for it and guess what I found: http://api.coala.io/en/latest/Developers/Adding_CI.html

I must say coala is well documented.

Abhinav Kaushlya
@abhishalya
Jan 31 12:36
Yep ;)
Amol Grover
@frextrite
Jan 31 15:14
So I tried setting up Travis CI on my fork. Setup was done but sentinel is failing when I try to run a build.
John Vandenberg
@jayvdb
Jan 31 15:50

So I tried setting up Travis CI on my fork. Setup was done but sentinel is failing when I try to run a build.

we have another error happening now -- pycodestyle is failing ... looks like a one liner to prevent that

if someone creates issue and PR, ill merge it tmr, so that one is out of the way
Amol Grover
@frextrite
Jan 31 16:33
I dont think its pycodestyle
I looked into build history and many of the builds have failed with exit code 1 with no apparent reason
Abhinav Kaushlya
@abhishalya
Jan 31 16:52
@frextrite Its due to pycodestylebear as you can see here
Amol Grover
@frextrite
Jan 31 17:01
Yep, just saw it.
Abhinav Kaushlya
@abhishalya
Jan 31 17:04
@jayvdb This is my initial approach asdf-vm/asdf#451
Unfortunately, there is some issue with the windows system, as mentioned in sstephenson/bats#234
Sladyn
@sladyn98
Jan 31 17:06
The pycodestyle bear seems to be complaining for a lot of files.
Shashank Batra
@shashank-b
Jan 31 17:07

The pycodestyle bear seems to be complaining for a lot of files.

+1

Sladyn
@sladyn98
Jan 31 17:15
On further inspection of the travis logs
! W605 invalid escape sequence '\d' in GitCommitBear.py
I think this is why the bear is complaining
Amol Grover
@frextrite
Jan 31 17:18
There's 2 of them
nevermind
just searched and found 4 W605
Abhinav Kaushlya
@abhishalya
Jan 31 17:21
Create an issue with the files mentioned
Sladyn
@sladyn98
Jan 31 17:39
Yeah
I will create an issue
Newcomers could pick this issue and submit a PR,they are moslty one line fixes
Amol Grover
@frextrite
Jan 31 18:17
btw what's the intended solution? Disable W605 warning?
Naveen Naidu
@Naveenaidu
Jan 31 18:33
Yup! I think that should be the answer. Ignore that warning in the PycodestyleBear.py
But I'm just curious, why is the bear behaving in the manner? I mean what is actually causing the error
Sladyn
@sladyn98
Jan 31 18:38
maybe because '\' is the default escape string literal and if we wanted to use a '\d' we might have to do something like '\\d'
Naveen Naidu
@Naveenaidu
Jan 31 18:39
Shouldn't we be fixing the file which has a bad escape sequence then?
Instead of ignoring the error.
I take back my statement. I don't think we need to ignore the W605 warning. Rather it would be better to fix the error that we get due the Line 38 of GitCommitBear.
Sladyn
@sladyn98
Jan 31 18:43
Exactly
I think we should add '\' in front of the escape literals
Naveen Naidu
@Naveenaidu
Jan 31 18:47
Yup the issues PyCQA/pycodestyle#755 and PyCQA/pycodestyle#766 tells the same thing that you suggested.
I'll try those changes(change '\d+' to '\d+') on my fork and check if it passes the travis
Amol Grover
@frextrite
Jan 31 18:51
pycodestyle would still emit an error if you have an invalid escape sequence which is not in a raw string
Naveen Naidu
@Naveenaidu
Jan 31 18:53
But IMHO we still should not ignore the warning, as this would become an error in the Python 3.8, It's always better to fix the error rather than just ignore it
Amol Grover
@frextrite
Jan 31 18:53
in the above mentioned files the invalid escape sequences are in multi line comments, so it will continue to emit errors. the only solution imo is wrapping them up in r"" which would not make sense.
see this PyCQA/pycodestyle#822 for reference
the only thing that was fixed was the wrong line numbers that were emitted
Naveen Naidu
@Naveenaidu
Jan 31 18:57
Gotcha!! Thanks:) Then the only solution would be for now is to ignore the warning
Naveen Naidu
@Naveenaidu
Jan 31 19:04
Well! ignoring makes the Travis green :)
@jayvdb should I send a PR or should the issue be open for newcomers?
Ignorance is Bliss :P
Sladyn
@sladyn98
Jan 31 19:05
yeah @Naveenaidu send the PR @jayvdb will merge it tommorrow
Naveen Naidu
@Naveenaidu
Jan 31 19:06
:+1:
Sladyn
@sladyn98
Jan 31 19:14
Hopefully that will fix some of the errors in the corobo_build as well
Naveen Naidu
@Naveenaidu
Jan 31 19:14
Amol Grover
@frextrite
Jan 31 19:15
@Naveenaidu Either we could ignore it as you said, or maybe wrap the invalid sequence with r" ". This should work aswell imo.
Naveen Naidu
@Naveenaidu
Jan 31 19:16
I tried wrapping it around r" " but found out that there are more than 3 files that have this error
Or we could just add \ to the sequence as @sladyn98 suggested.
Sladyn
@sladyn98
Jan 31 19:17
Yeah I think ignoring it for now works fine as well, we might have a problem with regex expressions later though but we will wrap them in r" " :)
Naveen Naidu
@Naveenaidu
Jan 31 19:18
We gotta wait until John decides upon that :P