test_coding_standards.py

# encoding: utf-8

'''A module for coding standards tests.

These are tests that are not functional- or unit-testing any particular piece
of CKAN code, but are checking coding standards. For example: checking that
there are no errors in the Sphinx build, that there are no PEP8 problems,
etc.

'''

import io
import os
import os.path
import re
import subprocess


def test_building_the_docs():
    '''There should be no warnings or errors when building the Sphinx docs.

    This test unfortunately does take quite a long time to run - rebuilding the
    docs from scratch just takes a long time.

    This test will also fail is build_sphinx exits with non-zero status.

    '''
    try:
        output = subprocess.check_output(
            ['python',
             'setup.py',
             'build_sphinx',
             '--all-files',
             '--fresh-env'],
            stderr=subprocess.STDOUT)
    except subprocess.CalledProcessError as err:
        assert False, (
            "Building the docs failed with return code: {code}".format(
                code=err.returncode))
    output_lines = output.split('\n')

    errors = [line for line in output_lines if 'ERROR' in line]
    if errors:
        assert False, ("Don't add any errors to the Sphinx build: "
                       "{errors}".format(errors=errors))

    warnings = [line for line in output_lines if 'WARNING' in line]

    # Some warnings have been around for a long time and aren't easy to fix.
    # These are allowed, but no more should be added.
    allowed_warnings = [
        'WARNING: duplicate label ckan.auth.create_user_via_web',
        'WARNING: duplicate label ckan.auth.create_unowned_dataset',
        'WARNING: duplicate label ckan.auth.user_create_groups',
        'WARNING: duplicate label ckan.auth.anon_create_dataset',
        'WARNING: duplicate label ckan.auth.user_delete_organizations',
        'WARNING: duplicate label ckan.auth.create_user_via_api',
        'WARNING: duplicate label ckan.auth.create_dataset_if_not_in_organization',
        'WARNING: duplicate label ckan.auth.user_delete_groups',
        'WARNING: duplicate label ckan.auth.user_create_organizations',
        'WARNING: duplicate label ckan.auth.roles_that_cascade_to_sub_groups'
    ]

    # Remove the allowed warnings from the list of collected warnings.
    # Be sure to only remove one warning for each allowed warning.
    warnings_to_remove = []
    for allowed_warning in allowed_warnings:
        for warning in warnings:
            if allowed_warning in warning:
                warnings_to_remove.append(warning)
                break
    new_warnings = [warning for warning in warnings
                    if warning not in warnings_to_remove]

    if new_warnings:
        assert False, ("Don't add any new warnings to the Sphinx build: "
                       "{warnings}".format(warnings=new_warnings))


def test_source_files_specify_encoding():
    '''
    Test that *.py files have a PEP 263 UTF-8 encoding specification.

    Empty files and files that only contain comments are ignored.
    '''
    root_dir = os.path.join(os.path.dirname(__file__), '..', '..')
    test_dirs = ['ckan', 'ckanext']
    ignored_dirs = ['ckan/include']
    pattern = re.compile(r'#.*?coding[:=][ \t]*utf-?8')
    decode_errors = []
    no_specification = []

    def check_file(filename):
        try:
            with io.open(filename, encoding='utf-8') as f:
                for line in f:
                    line = line.strip()
                    if pattern.match(line):
                        # Pattern found
                        return
                    elif line and not line.startswith('#'):
                        # File contains non-empty non-comment line
                        no_specification.append(os.path.relpath(filename,
                                                root_dir))
                        return
        except UnicodeDecodeError:
            decode_errors.append(filename)

    for test_dir in test_dirs:
        base_dir = os.path.join(root_dir, test_dir)
        for root, dirnames, filenames in os.walk(base_dir):
            dirnames[:] = [d for d in dirnames if not
                           os.path.relpath(os.path.join(root, d), root_dir)
                           in ignored_dirs]
            for filename in filenames:
                if not filename.endswith('.py'):
                    continue
                check_file(os.path.join(root, filename))

    msgs = []
    if no_specification:
        msgs.append('The following files are missing an encoding '
                    + 'specification: {}'.format(no_specification))
    if decode_errors:
        msgs.append('The following files are not valid UTF-8: {}'.format(
                    decode_errors))
    if msgs:
        assert False, '\n\n'.join(msgs)