from __future__ import annotations from collections import defaultdict from .logs import THREAT_ACTIONS, is_utm_event from .models import LogEvent def correlate_source_ips(events: list[LogEvent], *, limit: int = 20) -> list[dict[str, object]]: grouped: dict[str, list[LogEvent]] = defaultdict(list) for event in events: if event.src_ip: grouped[event.src_ip].append(event) correlations = [] for source_ip, source_events in grouped.items(): streams = sorted({event.fields.get("fgai_stream", "local_syslog") for event in source_events}) if len(streams) < 2: continue threat_events = sum(event.action in THREAT_ACTIONS or is_utm_event(event) for event in source_events) samples = [ { "stream": event.fields.get("fgai_stream", "local_syslog"), "timestamp": event.fields.get("eventtime", event.fields.get("timestamp", "")), "type": event.fields.get("type", ""), "subtype": event.subtype, "action": event.action, "severity": event.severity, "destination": event.dst_ip or "", "service": event.fields.get("service", ""), } for event in source_events[:20] ] correlations.append({"source_ip": source_ip, "streams": streams, "events": len(source_events), "security_events": threat_events, "samples": samples}) return sorted(correlations, key=lambda item: (int(item["security_events"]), int(item["events"])), reverse=True)[:limit]