"""Parse automation log files into structured events for the dashboard."""

import os
import re
from datetime import datetime, date
from pathlib import Path

# Patterns: (compiled_regex, event_type, optional extractor returning extra fields)
# Extractor receives the match object and returns a dict of extra fields.
# Order matters — first match wins.
PATTERNS = [
    # --- Email classification ---
    (re.compile(r"\[(\d+)/(\d+)\] Classifying: (.+?) \(from:"), "email_classifying",
     lambda m: {"progress": f"{m.group(1)}/{m.group(2)}", "subject": m.group(3)}),
    (re.compile(r"Cached: (.+?) -> (\w+)"), "email_cached",
     lambda m: {"subject": m.group(1), "category": m.group(2)}),
    (re.compile(r"→ (receipts|newsletters|work|personal|accounts)(?:\s*\((.+?)\))?"), "email_classified",
     lambda m: {"category": m.group(1), "label": m.group(2) or ""}),

    # --- Receipt extraction ---
    (re.compile(r"Would write:.*'vendor': '([^']+)'.*'amount': '([^']+)'"), "receipt_extracted",
     lambda m: {"vendor": m.group(1), "amount": m.group(2)}),
    (re.compile(r"Appended to CSV:.*vendor=([^,]+).*amount=([^,]+)"), "receipt_extracted",
     lambda m: {"vendor": m.group(1).strip(), "amount": m.group(2).strip()}),

    # --- Cron / automation completions ---
    (re.compile(r"Done! Stats: \{"), "cron_complete", lambda m: {}),

    # --- Container health / stack restarts ---
    (re.compile(r"Container (\S+) on (\S+) restarted"), "container_restarted",
     lambda m: {"container": m.group(1), "endpoint": m.group(2)}),
    (re.compile(r"LLM says (SAFE|UNSAFE) for (\S+)"), "restart_analysis",
     lambda m: {"decision": m.group(1), "container": m.group(2)}),
    (re.compile(r"[Uu]nhealthy.*?(\S+)\s+on\s+(\S+)"), "container_unhealthy",
     lambda m: {"container": m.group(1), "endpoint": m.group(2)}),
    (re.compile(r"[Uu]nhealthy"), "container_unhealthy", lambda m: {}),
    (re.compile(r"Stack-restart check complete"), "stack_healthy", lambda m: {}),

    # --- Backups ---
    (re.compile(r"Backup Validation: (OK|FAIL)"), "backup_result",
     lambda m: {"status": m.group(1)}),
    (re.compile(r"Backup Report"), "backup_result", lambda m: {"status": "report"}),

    # --- Config drift ---
    (re.compile(r"Detected (\d+) drifts? across (\d+) services?"), "drift_found",
     lambda m: {"drifts": m.group(1), "services": m.group(2)}),
    (re.compile(r"No drifts found"), "drift_clean", lambda m: {}),

    # --- Disk predictor ---
    (re.compile(r"WARNING.*volume.* (\d+) days"), "disk_warning",
     lambda m: {"days": m.group(1)}),
    (re.compile(r"Total filesystems: (\d+)"), "disk_scan_complete",
     lambda m: {"count": m.group(1)}),

    # --- Changelog / PR review ---
    (re.compile(r"Generated changelog with (\d+) commits"), "changelog_generated",
     lambda m: {"commits": m.group(1)}),
    (re.compile(r"(\d+) new commits since"), "changelog_commits",
     lambda m: {"count": m.group(1)}),
    (re.compile(r"Posted review comment on PR #(\d+)"), "pr_reviewed",
     lambda m: {"pr": m.group(1)}),

    # --- Catch-all patterns (lower priority) ---
    (re.compile(r"ERROR|CRITICAL"), "error", lambda m: {}),
    (re.compile(r"Starting .+ check|Starting .+ organizer"), "start", lambda m: {}),
    (re.compile(r"emails? downloaded|backup: \d+ total"), "backup_progress", lambda m: {}),
]

# Timestamp pattern at the start of log lines
TS_PATTERN = re.compile(r"^(\d{4}-\d{2}-\d{2}[\sT_]\d{2}:\d{2}:\d{2})")


def parse_timestamp(line: str) -> datetime | None:
    """Extract timestamp from a log line."""
    m = TS_PATTERN.match(line)
    if m:
        ts_str = m.group(1).replace("_", " ").replace("T", " ")
        try:
            return datetime.strptime(ts_str, "%Y-%m-%d %H:%M:%S")
        except ValueError:
            pass
    return None


def classify_line(line: str) -> tuple[str, dict] | None:
    """Return (event_type, extra_fields) if line matches a known pattern, else None."""
    for pattern, event_type, extractor in PATTERNS:
        m = pattern.search(line)
        if m:
            try:
                extra = extractor(m)
            except Exception:
                extra = {}
            return event_type, extra
    return None


def get_recent_events(log_dir: str | Path, max_events: int = 50) -> list[dict]:
    """Parse today's events from all log files in log_dir."""
    log_dir = Path(log_dir)
    today = date.today().isoformat()
    events = []

    for log_file in log_dir.glob("*.log"):
        source = log_file.stem
        try:
            with open(log_file, "r", errors="replace") as f:
                for line in f:
                    line = line.strip()
                    if not line or today not in line:
                        continue
                    ts = parse_timestamp(line)
                    if ts is None or ts.date().isoformat() != today:
                        continue
                    result = classify_line(line)
                    if result:
                        event_type, extra = result
                        raw_msg = line[len(ts.isoformat().split("T")[0]) + 1:].strip().lstrip(",").strip()
                        event = {
                            "time": ts.strftime("%H:%M:%S"),
                            "timestamp": ts.isoformat(),
                            "type": event_type,
                            "source": source,
                            "raw": raw_msg,
                            **extra,
                        }
                        events.append(event)
        except (OSError, PermissionError):
            continue

    events.sort(key=lambda e: e["timestamp"], reverse=True)
    return events[:max_events]


def tail_logs(log_dir: str | Path) -> dict[str, int]:
    """Return current file positions (sizes) for SSE polling."""
    log_dir = Path(log_dir)
    positions = {}
    for log_file in log_dir.glob("*.log"):
        try:
            positions[str(log_file)] = log_file.stat().st_size
        except OSError:
            positions[str(log_file)] = 0
    return positions


def get_new_lines(log_dir: str | Path, positions: dict[str, int]) -> tuple[list[dict], dict[str, int]]:
    """Read new lines since last positions. Returns (new_events, updated_positions)."""
    log_dir = Path(log_dir)
    today = date.today().isoformat()
    new_events = []
    new_positions = dict(positions)

    for log_file in log_dir.glob("*.log"):
        path_str = str(log_file)
        old_pos = positions.get(path_str, 0)
        try:
            current_size = log_file.stat().st_size
        except OSError:
            continue

        if current_size <= old_pos:
            new_positions[path_str] = current_size
            continue

        source = log_file.stem
        try:
            with open(log_file, "r", errors="replace") as f:
                f.seek(old_pos)
                for line in f:
                    line = line.strip()
                    if not line or today not in line:
                        continue
                    ts = parse_timestamp(line)
                    if ts is None:
                        continue
                    result = classify_line(line)
                    if result:
                        event_type, extra = result
                        raw_msg = line[len(ts.isoformat().split("T")[0]) + 1:].strip().lstrip(",").strip()
                        new_events.append({
                            "time": ts.strftime("%H:%M:%S"),
                            "timestamp": ts.isoformat(),
                            "type": event_type,
                            "source": source,
                            "raw": raw_msg,
                            **extra,
                        })
            new_positions[path_str] = current_size
        except (OSError, PermissionError):
            continue

    new_events.sort(key=lambda e: e["timestamp"], reverse=True)
    return new_events, new_positions