| 1 | """热榜 / 搜索数据采集模块。 |
| 2 | |
| 3 | 仅负责数据落盘（JSONL），不下载媒体本体。用户拿到结果后可挑感兴趣的链接再丢进下载器。 |
| 4 | """ |
| 5 | |
| 6 | from __future__ import annotations |
| 7 | |
| 8 | import asyncio |
| 9 | import json |
| 10 | from datetime import datetime |
| 11 | from pathlib import Path |
| 12 | from typing import TYPE_CHECKING, Any, Dict, List, Optional |
| 13 | |
| 14 | import aiofiles |
| 15 | |
| 16 | from utils.logger import setup_logger |
| 17 | |
| 18 | if TYPE_CHECKING: # pragma: no cover |
| 19 | from core.api_client import DouyinAPIClient |
| 20 | |
| 21 | logger = setup_logger("Discovery") |
| 22 | |
| 23 | |
| 24 | async def _write_jsonl(path: Path, items: List[Dict[str, Any]]) -> None: |
| 25 | path.parent.mkdir(parents=True, exist_ok=True) |
| 26 | async with aiofiles.open(path, "w", encoding="utf-8") as f: |
| 27 | for item in items: |
| 28 | await f.write(json.dumps(item, ensure_ascii=False)) |
| 29 | await f.write("\n") |
| 30 | |
| 31 | |
| 32 | async def dump_hot_board( |
| 33 | api_client: "DouyinAPIClient", |
| 34 | output_dir: Path, |
| 35 | *, |
| 36 | limit: int = 0, |
| 37 | ) -> Dict[str, Any]: |
| 38 | """抓取抖音热搜榜并写入 output_dir/hot_board/{ts}.jsonl。 |
| 39 | |
| 40 | Args: |
| 41 | limit: 上限（0=全部） |
| 42 | Returns: |
| 43 | dict(items, path) |
| 44 | """ |
| 45 | page = await api_client.get_hot_search_board() |
| 46 | items = list(page.get("items") or []) |
| 47 | if limit and limit > 0: |
| 48 | items = items[:limit] |
| 49 | |
| 50 | ts = datetime.now().strftime("%Y%m%d_%H%M%S") |
| 51 | out_path = output_dir / "hot_board" / f"{ts}.jsonl" |
| 52 | await _write_jsonl(out_path, items) |
| 53 | logger.info("Hot board snapshot saved: %s items -> %s", len(items), out_path) |
| 54 | return {"items": items, "path": str(out_path), "count": len(items)} |
| 55 | |
| 56 | |
| 57 | async def search_and_dump( |
| 58 | api_client: "DouyinAPIClient", |
| 59 | keyword: str, |
| 60 | output_dir: Path, |
| 61 | *, |
| 62 | max_items: int = 50, |
| 63 | page_size: int = 10, |
| 64 | sort_type: int = 0, |
| 65 | publish_time: int = 0, |
| 66 | rate_limiter: Optional[Any] = None, |
| 67 | ) -> Dict[str, Any]: |
| 68 | """搜索作品并将结果写入 output_dir/search/{keyword}_{ts}.jsonl。 |
| 69 | |
| 70 | Args: |
| 71 | max_items: 最多累计条数（0=不限，建议设置以防失控） |
| 72 | """ |
| 73 | accumulated: List[Dict[str, Any]] = [] |
| 74 | offset = 0 |
| 75 | seen_ids: set = set() |
| 76 | |
| 77 | while True: |
| 78 | if rate_limiter is not None: |
| 79 | try: |
| 80 | await rate_limiter.acquire() |
| 81 | except Exception: # noqa: BLE001 |
| 82 | pass |
| 83 | |
| 84 | page = await api_client.search_aweme( |
| 85 | keyword, |
| 86 | offset=offset, |
| 87 | count=page_size, |
| 88 | sort_type=sort_type, |
| 89 | publish_time=publish_time, |
| 90 | ) |
| 91 | items = page.get("items") or [] |
| 92 | if not items: |
| 93 | break |
| 94 | |
| 95 | for item in items: |
| 96 | if not isinstance(item, dict): |
| 97 | continue |
| 98 | aweme_id = str(item.get("aweme_id") or "") |
| 99 | if aweme_id in seen_ids: |
| 100 | continue |
| 101 | if aweme_id: |
| 102 | seen_ids.add(aweme_id) |
| 103 | accumulated.append(item) |
| 104 | if 0 < max_items <= len(accumulated): |
| 105 | break |
| 106 | |
| 107 | if 0 < max_items <= len(accumulated): |
| 108 | break |
| 109 | if not page.get("has_more"): |
| 110 | break |
| 111 | next_offset = int(page.get("max_cursor") or 0) |
| 112 | if next_offset == offset: |
| 113 | break |
| 114 | offset = next_offset |
| 115 | await asyncio.sleep(0.1) |
| 116 | |
| 117 | ts = datetime.now().strftime("%Y%m%d_%H%M%S") |
| 118 | safe_keyword = "".join(c if c.isalnum() else "_" for c in keyword)[:40] or "query" |
| 119 | out_path = output_dir / "search" / f"{safe_keyword}_{ts}.jsonl" |
| 120 | await _write_jsonl(out_path, accumulated) |
| 121 | logger.info("Search '%s' saved: %s items -> %s", keyword, len(accumulated), out_path) |
| 122 | return { |
| 123 | "keyword": keyword, |
| 124 | "items": accumulated, |
| 125 | "count": len(accumulated), |
| 126 | "path": str(out_path), |
| 127 | } |
| 128 |