2
0
mirror of https://github.com/catchorg/Catch2 synced 2026-02-23 04:12:15 +00:00
Files
Catch2/tests/TestScripts/testSectionFiltering.py
Martin Hořeňovský 3a0cf7e75f Support filtering on both sections and generators (#3069)
Not being able to filter generators to specific element has been regularly
causing problems. It was possible to use a dynamic section to run tests
for specific element in a generator, at least if the element had a nice
string representation, but the test case would still run once per element
in the generator.

With this change, it is possible to have the generator return only one
specific element, and do so based on the index, rather than the string
representation of the element. This enables simple debugging of tests
that fail for specific generator element.
2026-02-15 20:27:15 +01:00

190 lines
5.2 KiB
Python
Executable File

#!/usr/bin/env python3
# Copyright Catch2 Authors
# Distributed under the Boost Software License, Version 1.0.
# (See accompanying file LICENSE.txt or copy at
# https://www.boost.org/LICENSE_1_0.txt)
# SPDX-License-Identifier: BSL-1.0
"""
This test script verifies the behaviour of the legacy section filtering
using `-c`, `--section` CLI parameters.
This is done by having a hardcoded set of test filter + section filter
combinations, together with the expected number of assertions that will
be run inside the test for given filter combo.
"""
import os
import subprocess
import sys
from typing import Tuple, List
import xml.etree.ElementTree as ET
def make_old_filters(section_names: Tuple[str, ...]) -> List[str]:
final = []
for name in section_names:
final.append('--section')
final.append(name)
return final
def make_new_filters(path_filters: Tuple[str, ...]) -> List[str]:
final = []
for component in path_filters:
final.append('--path-filter')
final.append(component)
return final
def run_one_test(binary_path: str,
test_name: str,
section_names: Tuple[str, ...],
expected_assertions: int,
new_filters: bool):
cmd = [
binary_path,
'--reporter', 'xml',
test_name
]
if new_filters:
cmd.extend(make_new_filters(section_names))
else:
cmd.extend(make_old_filters(section_names))
try:
ret = subprocess.run(
cmd,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
check=True,
universal_newlines=True,
)
stdout = ret.stdout
except subprocess.SubprocessError as ex:
print('Could not run "{}"'.format(cmd))
print("Return code: {}".format(ex.returncode))
print("stdout: {}".format(ex.stdout))
print("stderr: {}".format(ex.stderr))
raise
try:
tree = ET.fromstring(stdout)
except ET.ParseError as ex:
print("Invalid XML: '{}'".format(ex))
raise
# Validate that we ran exactly 1 test case, and it passed
test_case_stats = tree.find('OverallResultsCases')
expected_testcases = {'successes' : '1', 'failures' : '0', 'expectedFailures': '0', 'skips': '0'}
assert test_case_stats.attrib == expected_testcases, f'We did not run single passing test case as expected. {test_name}: {test_case_stats.attrib}'
# Validate that we got exactly the expected number of passing assertions
expected_assertions = {'successes' : str(expected_assertions), 'failures' : '0', 'expectedFailures': '0', 'skips': '0'}
assertion_stats = tree.find('OverallResults')
assert assertion_stats.attrib == expected_assertions, f'"{test_name}": {assertion_stats.attrib} vs {expected_assertions}'
# Inputs taken from issue #3038
old_section_tests = {
'foo': (
((), 3),
(('A',), 3),
(('A', 'B'), 0),
(('A', 'B1'), 1),
(('A', 'B2'), 1),
(('A', 'B1', 'B2'), 1),
(('A', 'B2', 'XXXX'), 1),
),
'bar': (
((), 9),
(('A',), 9),
(('A', 'B1'), 3),
(('XXXX',), 2),
(('B1',), 2),
(('A', 'B1', 'B2'), 3),
),
'baz': (
((), 4),
(('A',), 1),
(('A', 'B'), 1),
(('A', 'XXXX'), 1),
(('B',), 3),
(('XXXX',), 0),
),
'qux': (
((), 12),
(('A',), 7),
(('B',), 9),
(('B', 'XXXX'), 9),
(('XXXX',), 6),
),
'corge': (
((), 12),
(('i=2',), 7),
(('i=3',), 7),
),
'grault': (
((), 12),
(('A',), 3),
(('B',), 9),
(('B', 'i=1'), 7),
(('B', 'XXXX'), 6),
),
}
new_filter_tests = {
'grault': (
((), 12),
(('c:B', 'g:1'), 3),
(('c:B', 'g:1', 'c:i=2'), 3),
# We filter the generator to different element than the dynamic section
(('c:B', 'g:1', 'c:i=1'), 2),
),
'garply': (
((), 3),
(('g:1',), 1),
(('g:*',), 3),
# Followup section filters do not change results (there are no sections underneath)
(('g:1', 'c:abcdef'), 1),
(('g:*', 'c:abcdef'), 3),
),
'waldo': (
((), 9),
(('g:1',), 3),
(('g:*',), 9),
(('g:1', 'g:2'), 1),
(('g:*', 'g:1'), 3),
(('g:0', 'g:*'), 3),
(('g:*', 'g:*'), 9),
(('g:*', 'g:1', 'c:abcdef'), 3),
),
'fred': (
((), 24),
# No sections entered, but we still hit assertion at the start of test
(('c:abcdef',), 1),
(('c:A', ), 12),
(('c:A', 'g:*', 'c:A1'), 6),
(('c:A', 'g:1', 'c:A1'), 2),
(('c:A', 'g:1'), 4),
# Section 'A3' does not exist
(('c:A', 'g:*', 'c:A3'), 3),
(('c:A', 'g:1', 'c:A3'), 1),
# We would like something like 'c:*', 'g:2', but we do not support globs
# for sections (yet?)
),
}
if len(sys.argv) != 2:
print("Wrong number of arguments, expected just the path to Catch2 SelfTest binary")
exit(1)
bin_path = os.path.abspath(sys.argv[1])
for test_filter, specs in old_section_tests.items():
for section_path, expected_assertions in specs:
run_one_test(bin_path, test_filter, section_path, expected_assertions, new_filters=False)
for test_filter, specs in new_filter_tests.items():
for section_path, expected_assertions in specs:
run_one_test(bin_path, test_filter, section_path, expected_assertions, new_filters=True)