maurane
c5466a4171
adapt to new naming of parsing methods, remove unused paths in parser module, all module import logwatcher.config instead of importing logwatcher, rename test functions
141 lines
4.1 KiB
Python
141 lines
4.1 KiB
Python
import re
|
|
import warnings
|
|
from pathlib import Path
|
|
|
|
import pytest
|
|
|
|
from logwatcher.classifier import N2_PATTERNS, classify_log_entries
|
|
from logwatcher.config import FIXTURE_PATH
|
|
from logwatcher.parser import parse_file
|
|
|
|
|
|
def _pattern_name_to_filename(name: str) -> str:
|
|
"""
|
|
Returns the fixture corresponding to the error code name
|
|
"""
|
|
return f"CR_{name.lower()}.txt"
|
|
|
|
|
|
def _read_fixture(path: Path) -> str:
|
|
"""
|
|
Read a file with fallback Windows-1252.
|
|
"""
|
|
try:
|
|
content = path.read_text(encoding="utf-8")
|
|
except UnicodeDecodeError:
|
|
content = path.read_text(encoding="windows-1252")
|
|
if "\ufffd" not in content:
|
|
return content
|
|
return path.read_text(encoding="windows-1252") # TODO: improve coverage score
|
|
|
|
|
|
@pytest.mark.parametrize("name, pattern", N2_PATTERNS.items())
|
|
def test_pattern_matches_fixture(valid_log_dir, name, pattern):
|
|
"""
|
|
Assert each N2 pattern must match at least one line in its fixture.
|
|
"""
|
|
|
|
fixture = valid_log_dir / _pattern_name_to_filename(name)
|
|
if not fixture.exists():
|
|
msg = f"No fixture named '{fixture}'"
|
|
warnings.warn(msg)
|
|
pytest.skip(msg)
|
|
|
|
lignes = [l for l in _read_fixture(fixture).splitlines() if l.strip()]
|
|
|
|
assert lignes, f"empty fixture file : {fixture}"
|
|
|
|
match_count = 0
|
|
|
|
for ligne in lignes:
|
|
match = re.search(pattern, ligne)
|
|
if match:
|
|
match_count += 1
|
|
|
|
assert match_count > 0, (
|
|
f"{name} : no match in file {fixture.name}\npattern: {pattern}"
|
|
)
|
|
|
|
|
|
def test_patterns_do_not_overlap():
|
|
"""
|
|
Assert that a log line can match only one N2 pattern.
|
|
"""
|
|
fixtures_dir = FIXTURE_PATH
|
|
for fixture in fixtures_dir.glob("CR_*.txt"):
|
|
line_list = [
|
|
line for line in _read_fixture(fixture).splitlines() if line.strip()
|
|
] # TODO: improve coverage score
|
|
for line in line_list: # TODO: improve coverage score
|
|
matching = [ # TODO: improve coverage score
|
|
name
|
|
for name, raw in N2_PATTERNS.items()
|
|
if re.compile(raw).search(line)
|
|
]
|
|
assert len(matching) <= 1, ( # TODO: improve coverage score
|
|
f"Chevauchement détecté sur {fixture.name} : "
|
|
f"{matching} match line : {line[:80]}"
|
|
)
|
|
|
|
|
|
def test_no_orphan_fixtures():
|
|
"""
|
|
Assert that a fixture exist only if its associated pattern exists.
|
|
Aucun fichier de fixture ne doit exister sans pattern associé.
|
|
"""
|
|
expected = {f"CR_{name.lower()}.txt" for name in N2_PATTERNS}
|
|
actual = {f.name for f in FIXTURE_PATH.glob("CR_*.txt")}
|
|
orphan = actual - expected
|
|
assert not orphan, f"Fixtures orphelines (sans pattern) : {sorted(orphan)}"
|
|
|
|
|
|
def test_classify_log_entries_all_relevant(valid_log_dir: Path):
|
|
"""
|
|
Must return an empty irrelevant log entry list
|
|
"""
|
|
log_file = valid_log_dir / "only_relevant_logs.txt"
|
|
log_entries = parse_file(log_file)
|
|
|
|
# all entries are relevant
|
|
relevant, irrelevant = classify_log_entries(log_entries)
|
|
|
|
assert relevant and not irrelevant
|
|
|
|
|
|
def test_classify_log_entries_none_relevant(valid_log_dir: Path):
|
|
"""
|
|
Must return an empty relevant log entry list
|
|
"""
|
|
log_file = valid_log_dir / "no_relevant_logs.txt"
|
|
log_entries = parse_file(log_file)
|
|
|
|
# all entries are relevant
|
|
relevant, irrelevant = classify_log_entries(log_entries)
|
|
|
|
assert not relevant and irrelevant
|
|
|
|
|
|
def test_classify_log_entries_mixed(valid_log_dir: Path):
|
|
"""
|
|
Test classification on relevant and irrelevant log entry list
|
|
"""
|
|
log_file = valid_log_dir / "mixed_logs.txt"
|
|
log_entries = parse_file(log_file)
|
|
|
|
relevant, irrelevant = classify_log_entries(log_entries)
|
|
|
|
assert len(relevant) and len(irrelevant)
|
|
assert len(relevant) + len(irrelevant) == len(log_entries)
|
|
|
|
|
|
def test_classify_log_entries_empty_log_entries(invalid_log_dir: Path):
|
|
"""
|
|
Test classification on an empty log entry list
|
|
"""
|
|
empty_file = invalid_log_dir / "empty_file.txt"
|
|
log_entries = parse_file(empty_file)
|
|
|
|
relevant, irrelevant = classify_log_entries(log_entries)
|
|
|
|
assert not relevant and not irrelevant
|