log_pattern_matcher.md

log pattern matcher

Language: Python · Sphere: programming · Category: Monitoring

What it does

Ordered regex log parser that extracts named groups into structured records.

Use to turn heterogeneous log lines into dicts: register named patterns (each a regex with named groups, e.g. apache/app/login formats), then parse a line or a batch. The first registered pattern that matches wins, and per-pattern match counts are tracked. Guarantees (proven by self-test): each format's named fields are extracted exactly; a line matching no pattern returns (None, None) rather than being force-fit; match counts are exact per pattern; ordering (first-wins) is respected; and an invalid regex is refused at registration, not at match time.

Guarantee

When it runs, log pattern matcher guarantees int(data['status']) + int(data['size']) == 2526; name == 'apache_common'; data == {'ip': '192.168.1.1', 'timestamp': '10/Oct/2023:13:55:36 +0000', 'method': 'GET', 'path': '/index.html', 'status': '200', 'size': '2326'} (proven by run).

Checkable constraints:

• name == 'apache_common'
• data == {'ip': '192.168.1.1', 'timestamp': '10/Oct/2023:13:55:36 +0000', 'method': 'GET', 'path': '/index.html', 'status': '200', 'size': '2326'}
• int(data['status']) + int(data['size']) == 2526
• name == 'app_log' and data['level'] == 'ERROR'
• data['message'] == 'Database connection failed'
• name == 'login_event'
• data == {'user': 'john', 'ip': '192.168.1.3'}
• results[7] == (None, None)

Verification evidence

• Green-run: ✓ passes (re-run under the extractor's gate)
• Constraint strength: recovery (truth-pinned)
• Independent oracle: — none yet (green-run candidate; not an axiom under the frozen ruler)
• Peer review: unreviewed

---

△ AURA Pattern Library — © Reality Optimizer