#!/usr/bin/env python3 """Строит недельную сводку по CSV с метриками.""" from __future__ import annotations import argparse import csv import json import os import statistics import sys import tempfile from datetime import datetime, timedelta, timezone from pathlib import Path def parse_date(value: str): text = value.strip() if not text: return None if text.endswith("Z"): text = text[:-1] + "+00:00" parsed = None try: parsed = datetime.fromisoformat(text) except ValueError: pass if parsed is None: for fmt in ("%Y-%m-%d %H:%M:%S", "%Y-%m-%d", "%d.%m.%Y"): try: parsed = datetime.strptime(text, fmt) break except ValueError: continue if parsed is not None and parsed.tzinfo is not None: parsed = parsed.astimezone(timezone.utc).replace(tzinfo=None) return parsed def summarize(rows: list[dict], fields: list[str]) -> dict: result = {} for field in fields: values = [] for row in rows: try: values.append(float(row[field])) except (KeyError, TypeError, ValueError): continue if values: result[field] = { "count": len(values), "min": min(values), "max": max(values), "avg": statistics.mean(values), } return result def main() -> int: parser = argparse.ArgumentParser(description="Недельная сводка по CSV") parser.add_argument("--input", required=True, type=Path) parser.add_argument("--date-column", default="") parser.add_argument("--days", type=int, default=7) parser.add_argument("--format", choices=("text", "markdown"), default="text") parser.add_argument("--output", type=Path) parser.add_argument("--json", action="store_true") args = parser.parse_args() try: handle = args.input.open("r", encoding="utf-8", newline="") except OSError as error: print(f"ОШИБКА: {error}", file=sys.stderr) return 2 with handle: reader = csv.DictReader(handle) fields = reader.fieldnames or [] rows = list(reader) if not fields: print("В CSV нет заголовка", file=sys.stderr) return 2 date_field = args.date_column if not date_field: for candidate in ("date", "timestamp", "time"): if candidate in fields: date_field = candidate break if not date_field or date_field not in fields: print("Не найдена колонка с датой, укажите --date-column", file=sys.stderr) return 2 dated = [] for row in rows: stamp = parse_date(str(row.get(date_field, ""))) if stamp is not None: dated.append((stamp, row)) if not dated: print("Не удалось разобрать даты", file=sys.stderr) return 2 latest = max(stamp for stamp, _ in dated) current_start = latest - timedelta(days=args.days) previous_start = current_start - timedelta(days=args.days) current_rows = [row for stamp, row in dated if current_start <= stamp <= latest] previous_rows = [row for stamp, row in dated if previous_start <= stamp < current_start] metric_fields = [field for field in fields if field != date_field] current = summarize(current_rows, metric_fields) previous = summarize(previous_rows, metric_fields) comparison = [] for field, stats in current.items(): old = previous.get(field) delta = None if old and old["avg"] != 0: delta = (stats["avg"] - old["avg"]) * 100 / old["avg"] comparison.append( { "metric": field, "current_avg": round(stats["avg"], 3), "previous_avg": round(old["avg"], 3) if old else None, "delta_percent": round(delta, 1) if delta is not None else None, "min": round(stats["min"], 3), "max": round(stats["max"], 3), } ) payload = { "period_end": latest.isoformat(), "current_rows": len(current_rows), "previous_rows": len(previous_rows), "metrics": comparison, } if args.json: document = json.dumps(payload, ensure_ascii=False, indent=2) + "\n" elif args.format == "markdown": lines = [ "# Недельная сводка", "", f"Конец периода: {latest.strftime('%Y-%m-%d %H:%M')}.", f"Строк: текущий период {len(current_rows)}, предыдущий {len(previous_rows)}.", "", "| Метрика | Среднее | Было | Изменение | Мин | Макс |", "| --- | ---: | ---: | ---: | ---: | ---: |", ] for row in comparison: old = "-" if row["previous_avg"] is None else f"{row['previous_avg']:.3f}" delta = "-" if row["delta_percent"] is None else f"{row['delta_percent']:+.1f}%" lines.append( f"| {row['metric']} | {row['current_avg']:.3f} | {old} | {delta} | " f"{row['min']:.3f} | {row['max']:.3f} |" ) document = "\n".join(lines) + "\n" else: lines = [f"Конец периода: {latest.strftime('%Y-%m-%d %H:%M')}"] lines.append(f"Строк: текущий {len(current_rows)}, предыдущий {len(previous_rows)}") for row in comparison: old = "нет данных" if row["previous_avg"] is None else f"{row['previous_avg']:.3f}" delta = "" if row["delta_percent"] is None else f", изменение {row['delta_percent']:+.1f}%" lines.append( f"{row['metric']}: avg {row['current_avg']:.3f}, было {old}{delta}, " f"мин {row['min']:.3f}, макс {row['max']:.3f}" ) document = "\n".join(lines) + "\n" if args.output: target = args.output.expanduser() if not target.parent.is_dir(): print(f"Каталог не найден: {target.parent}", file=sys.stderr) return 2 descriptor, temporary = tempfile.mkstemp(dir=target.parent, prefix=".summary-", suffix=".txt") try: with os.fdopen(descriptor, "w", encoding="utf-8") as output: output.write(document) os.replace(temporary, target) except OSError as error: print(f"ОШИБКА: {error}", file=sys.stderr) return 2 print(f"Сводка записана: {target}") else: sys.stdout.write(document) return 0 if __name__ == "__main__": raise SystemExit(main())