v1.1.6.4 still breaks pylint

[feenes]

For versions 1.1.6.3 and 1.1.6.4 I get following error during our CI runs.

We had no issue (except contradicting dependencies with flake8) before

The related error log:

$ prospector
Messages
========

.
  Line: None
    pylint: failure / Tool pylint failed to run (exception was raised)



Check Information
=================
         Started: 2019-06-13 08:48:04.082400
        Finished: 2019-06-13 08:48:19.840839
      Time Taken: 15.76 seconds
       Formatter: grouped
        Profiles: .prospector.yaml, no_doc_warnings, no_member_warnings, no_test_warnings, strictness_medium, strictness_high, strictness_veryhigh
      Strictness: from profile
  Libraries Used: 
       Tools Run: pep8, profile-validator, pyflakes, pylint
  Messages Found: 1

ERROR: Job failed: exit status 1

pip check output:

$ cat pipcheck.txt
No broken requirements found.

Have to change our CI pipeline to add output of pip freeze

[carlio]

Can you rerun with the -X flag to prospector? That'll show you the stacktrace caused by pylint, then add that here, which would be useful for debugging this.

[feenes]

changed our CI pipeline:

WIth -X I don't get the previous error message anymore. Probably this is what should happen when calling with ',-X' ?

Here the output of prospector -X:

$ prospector -X
Traceback (most recent call last):
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/bin/prospector", line 10, in <module>
    sys.exit(main())
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/prospector/run.py", line 173, in main
    prospector.execute()
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/prospector/run.py", line 66, in execute
    messages += tool.run(found_files)
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/prospector/tools/pylint/__init__.py", line 254, in run
    self._linter.check(self._args)
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/pylint/lint.py", line 943, in check
    self._do_check(files_or_modules)
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/pylint/lint.py", line 1081, in _do_check
    self.add_message(msgid, line, None, args)
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/pylint/utils.py", line 496, in add_message
    message_definition, line, node, args, confidence, col_offset
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/pylint/utils.py", line 571, in add_one_message
    confidence,
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/prospector/tools/pylint/collector.py", line 21, in handle_message
    self.add_message(msg.msg_id, location, msg.msg)
  File "/home/gitlab-runner/builds/30c73d6d/0/x/y/eenv/lib/python3.6/site-packages/prospector/tools/pylint/collector.py", line 32, in add_message
    msg_data = self._message_store.get_message_definition(msg_id)
AttributeError: 'MessagesStore' object has no attribute 'get_message_definition'
ERROR: Job failed: exit status 1

And here the output of pip freeze:

asn1crypto==0.24.0
astroid==2.2.5
atomicwrites==1.3.0
attrs==19.1.0
backcall==0.1.0
bcrypt==3.1.6
begins==0.9
boto3==1.9.167
botocore==1.12.167
cairocffi==1.0.2
CairoSVG==2.4.0
certifi==2019.3.9
cffi==1.12.3
chardet==3.0.4
Click==7.0
coverage==4.5.3
cryptography==2.7
cssselect==1.0.3
cssselect2==0.2.1
decorator==4.4.0
defusedxml==0.6.0
Django==1.11.21
django-cors-headers==3.0.2
django-debug-toolbar==1.4
django-extensions==2.1.9
django-filter==1.1.0
django-jsonfield==1.2.0
djangorestframework==3.6.4
docutils==0.14
dodgy==0.1.9
entrypoints==0.3
Fabric3==1.14.post1
factory-boy==2.12.0
Faker==1.0.7
filemagic==1.6
future==0.17.1
holidays==0.9.10
html5lib==1.0.1
idna==2.8
importlib-metadata==0.18
ipython==7.5.0
ipython-genutils==0.2.0
isort==4.3.20
jedi==0.13.3
jmespath==0.9.4
lazy-object-proxy==1.4.1
lxml==4.3.4
mccabe==0.6.1
minibelt==0.1.1
mock==3.0.5
more-itertools==7.0.0
packaging==19.0
paramiko==2.5.0
parso==0.4.0
pep8==1.7.1
pep8-naming==0.4.1
pexpect==4.7.0
pickleshare==0.7.5
Pillow==6.0.0
pluggy==0.12.0
prompt-toolkit==2.0.9
prospector==1.1.6.4
psycopg2-binary==2.8.2
ptyprocess==0.6.0
py==1.8.0
pycodestyle==2.4.0
pycparser==2.19
pydocstyle==3.0.0
pyflakes==1.6.0
pygal==2.4.0
Pygments==2.4.2
pylint==2.3.1
pylint-celery==0.3
pylint-common==0.2.5
pylint-django==2.0.9
pylint-flask==0.6
pylint-plugin-utils==0.5
PyNaCl==1.3.0
pyotp==2.2.7
pyparsing==2.4.0
PyPDF2==1.26.0
Pyphen==0.9.5
pytest==4.6.3
pytest-django==3.5.0
python-dateutil==2.8.0
pytz==2019.1
PyYAML==5.1.1
requests==2.22.0
requirements-detector==0.6
responses==0.10.6
s3transfer==0.2.1
setoptconf==0.2.0
six==1.12.0
snowballstemmer==1.2.1
sqlparse==0.3.0
text-unidecode==1.2
tinycss==0.4
tinycss2==1.0.2
traitlets==4.3.2
typed-ast==1.4.0
urllib3==1.25.3
uWSGI==2.0.18
wcwidth==0.1.7
WeasyPrint==0.36
webencodings==0.5.1
wrapt==1.11.1
yapf==0.27.0
zipp==0.5.1

[sphuber]

This is the same problem that I reported in #321

[feenes]

Yes this looks identical.

What is weird however is, that I have the problem only with 1.1.6.3 and 1.1.6.4 which were release only a few days ago, but you encountered the problem already in March.

With 1.1.6.2 I do not have this issue

[sphuber]

Could it be that when you ran with 1.1.6.2 you happened to run with another version of pylint that was compatible. But now in the new runs, prospector will also install pylint==2.3.* which causes the failure

[feenes]

I will force 1.1.6.2 and dump the output of pip freeze

[feenes]

@sphuber You're right
My yesterday's fix
pip install prospector -U 'prospector!=1.1.6.3'

will install prospector==1.1.6.4 and pylint==2.3.1

whereas
pip install -U 'prospector!=1.1.6.3,!=1.1.6.4'
will install prospector==1.1.6.2 and pylint==2.1.1

[feenes]

Please advice:

Normally I do not like too much to limit the max versions of packages in my pip installs, but if it is probable, that the next release will still have this issue I'll probably go for the time being for something like
pip install -U 'prospector==1.1.6.2 in order to not block the developers waiting for our CI runs to pass.

[chocoelho]

@carlio the issue is within get_message_definition that got renamed and returns now a list, by changing the call to how Pylint uses it internally, it works again https://github.com/PyCQA/pylint/blob/master/pylint/message/message_store.py#L42

[chocoelho]

Though, there's an issue I noticed yesterday EOD an issue with pylint-django, described in here: pylint-dev/pylint-django#232

For now, it's better to use 1.1.6.2 while this is under fixing.

[chocoelho]

I'm already discussing inside pylint-dev/pylint-django#232 to have support for the latest pylint version with a new release of pylint-django ASAP.

[jabdoa2]

We got the same issue but we are not using django.

[volans-]

Any news on this?

@jabdoa2: that's independent of your project, prospector has pylint-django as dependency, see https://github.com/PyCQA/prospector/blob/master/setup.py#L35

[chocoelho]

Hey there folks, sorry about the delay. As we're still reviewing #342, we published a pre-release version 1.1.7.dev0. It'd be nice if you could give it a try and help us out testing the next version candidate.

[sphuber]

Thanks a lot for the new release. I tested it for our deployment on py2 and py3 and both now work with prospector==1.1.7dev0

[volans-]

@chocoelho thanks for the test release. I can confirm both 1.1.6.2 and 1.1.7.dev0 works fine while 1.1.6.4 fails for the repository where I had the issue. I tested it on Python 3.5, 3.6, 3.7.

[chocoelho]

That's great! Will start using pre-releases before making sudden changes as it was intended with 1.1.6.3 and 1.1.6.4. Sorry about that y'all, and we expect to release 1.1.7 soon.

[ggravlingen]

I came here after also experiencing some problems with pylint and prospector. Tried the pre-release of 1.1.7 and still get an error. Had another error before upgrading though. Perhaps I should move this comment to an issue instead to keep this PR clean?

Traceback (most recent call last):
  File "/home/vsts/work/1/s/venv/lib/python3.5/site-packages/astroid/bases.py", line 183, in getattr
    values = self._proxied.instance_attr(name, context)
  File "/home/vsts/work/1/s/venv/lib/python3.5/site-packages/astroid/scoped_nodes.py", line 2311, in instance_attr
    target=self, attribute=name, context=context
astroid.exceptions.AttributeInferenceError: 'index' not found on <ClassDef.SingleBlockManager l.1435 at 0x7f1a68f67908>.

Full stack trace here:
https://hastebin.com/uhokilefup.sql

What I'm trying to wrap my head around at the moment is if there's anything in my code or if there is really something going on with Astroid.

Edit: since it's a Pandas-file that causes the crash I believe it's linked to these issues: pylint-dev/pylint#2843 and pylint-dev/pylint#2811 rather than Prospector so please disregard.

[chocoelho]

Closing this as prospector 1.1.7 should fix this issue. Feel free to reopen it if this is erroring for you.