fail to run initial behave and test error

[igncampa]

Description

I pulled the repo to my machine and followed the contributing guide. Running behave and py.test both gave me an error:

behave command gives the following error:

package root: /home/igncampa/Repos/pgcli
fixture dir: /home/igncampa/Repos/pgcli/tests/features/fixture_data
HOOK-ERROR in before_all: OperationalError: fe_sendauth: no password supplied
HOOK-ERROR in after_all: AttributeError: 'Context' object has no attribute 'cn'

ABORTED: By user.
0 features passed, 0 failed, 0 skipped, 8 untested
0 scenarios passed, 0 failed, 0 skipped, 16 untested
0 steps passed, 0 failed, 0 skipped, 0 undefined, 85 untested
Took 0m0.000s

also running py.test one of the tests fails:
test_naive_completion.py:65: AssertionError

in case it matters, on my pg_hba.conf I have:
local all postgres trust

Your environment

• Please provide your OS and version information.
  Description: Ubuntu 18.04.1 LTS

• Please provide your CLI version.
  psql (PostgreSQL) 10.5 (Ubuntu 10.5-1.pgdg18.04+1)

• What is the output of pip freeze command.
  autopep8==1.3.3
  behave==1.2.6
  certifi==2018.8.24
  chardet==3.0.4
  cli-helpers==1.0.2
  click==6.7
  codecov==2.0.15
  colorama==0.3.9
  configobj==5.0.6
  coverage==4.3.4
  docformatter==1.0
  docutils==0.14
  humanize==0.5.1
  idna==2.7
  mock==2.0.0
  parse==1.8.4
  parse-type==0.4.2
  pbr==4.2.0
  pep8radius==0.9.2
  pexpect==3.3
  -e git+git@github.com:igncampa/pgcli.git@d3bdb89#egg=pgcli
  pgspecial==1.11.2
  pkg-resources==0.0.0
  pkginfo==1.4.2
  pluggy==0.7.1
  prompt-toolkit==1.0.15
  psycopg2==2.7.5
  psycopg2-binary==2.7.5
  py==1.6.0
  pycodestyle==2.4.0
  Pygments==2.2.0
  pytest==3.0.7
  requests==2.19.1
  requests-toolbelt==0.8.0
  setproctitle==1.1.10
  six==1.11.0
  sqlparse==0.2.4
  tabulate==0.8.2
  terminaltables==3.1.0
  toml==0.9.6
  tox==3.4.0
  tqdm==4.26.0
  twine==1.11.0
  untokenize==0.1.1
  urllib3==1.23
  virtualenv==16.0.0
  wcwidth==0.1.7
  yapf==0.24.0

[j-bennet]

I recently encountered this error when local connections had trust auth, but host didn't. Changing host to trust works:

# IPv4 local connections:
host    all             all             127.0.0.1/32            trust
# IPv6 local connections:
host    all             all             ::1/128                 trust

What kind of error do you get in test_naive_completion.py? Try py.test with verbose flag:

py.test -svv tests/test_naive_completion.py

[igncampa]

I started a fresh terminal after setting the lines to trust on my pg_hba.conf but the problem persists. No change in the output.

The output of running test_naive_completion.py is the following:

============================= test session starts ==============================
platform linux -- Python 3.6.5, pytest-3.0.7, py-1.6.0, pluggy-0.4.0 -- /home/igncampa/Repos/venvs/pgcli/bin/python3.6
cachedir: .cache
rootdir: /home/igncampa/Repos/pgcli/tests, inifile: pytest.ini
collecting ... collected 6 items

test_naive_completion.py::test_empty_string_completion PASSED
test_naive_completion.py::test_select_keyword_completion PASSED
test_naive_completion.py::test_function_name_completion PASSED
test_naive_completion.py::test_column_name_completion PASSED
test_naive_completion.py::test_paths_completion FAILED
test_naive_completion.py::test_alter_well_known_keywords_completion PASSED

=================================== FAILURES ===================================
____________________________ test_paths_completion _____________________________

completer = <pgcli.pgcompleter.PGCompleter object at 0x7f7aab17a2b0>
complete_event = <Mock id='140164833192144'>

    def test_paths_completion(completer, complete_event):
        text = '\i '
        position = len(text)
        result = set(completer.get_completions(
            Document(text=text, cursor_position=position),
            complete_event,
            smart_completion=True))
>       assert result > set([Completion(text="setup.py", start_position=0)])
E       AssertionError: assert {Completion(text='metadata.py', start_position=0), Completion(text='conftest.py', start_position=0), Completion(text='...=0), Completion(text='test_prioritization.py', start_position=0), Completion(text='pytest.ini', start_position=0), ...} > {Completion(text='setup.py', start_position=0)}
E        +  where {Completion(text='setup.py', start_position=0)} = set([Completion(text='setup.py', start_position=0)])

complete_event = <Mock id='140164833192144'>
completer  = <pgcli.pgcompleter.PGCompleter object at 0x7f7aab17a2b0>
position   = 3
result     = {Completion(text='metadata.py', start_position=0), Completion(text='conftest.py', start_position=0), Completion(text='...=0), Completion(text='test_prioritization.py', start_position=0), Completion(text='pytest.ini', start_position=0), ...}
text       = '\\i '

test_naive_completion.py:65: AssertionError
====================== 1 failed, 5 passed in 0.12 seconds ======================

Thanks for the help.

[j-bennet]

After changing pg_hba.conf, postgresql service needs a restart.

[j-bennet]

@igncampa I got it, you're running tests from the wrong directory. To run tests:

cd /home/igncampa/Repos/pgcli
py.test -sv

to run behave tests:

cd /home/igncampa/Repos/pgcli/tests
behave

That test_paths_completion tests actual paths, not mocked. It is not very well designed, sorry.

[igncampa]

Thank you, @j-bennet. You were right on both cases!

Looking forward to contributing some code!