| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566 |
- """M2 离线测试：用 5 个真实 capture 验证 parse_detail_response 字段映射。"""
- from __future__ import annotations
- import json
- from pathlib import Path
- import pytest
- from creation_knowledge.integrations.crawler import (
- parse_content_id,
- parse_detail_response,
- )
- FIXTURES = Path(__file__).parent / "fixtures"
- # content_id -> (期望作者, 标题里应出现的子串)
- EXPECTED = {
- "67e4bdf50000000006028a59": ("海狸教自媒体运营", "短视频脚本"),
- "698481e1000000000a02a7c1": ("Irvin是个编剧（接稿版）", "叙事结构"),
- "67e2e39b0000000003028ff0": ("拾意", "剧本创作"),
- "699308fa0000000016009697": ("方圆的增长飞轮", "选题"),
- "680659e8000000001a007a11": ("故事设计原理拆解学习"[:2], "故事设计"),
- }
- def _load(content_id: str) -> dict:
- return json.loads((FIXTURES / f"xhs_case_{content_id}.json").read_text("utf-8"))
- @pytest.mark.parametrize("content_id", list(EXPECTED))
- def test_parse_detail_fields(content_id: str):
- # 重抓的 xhs_case_*.json 即原始响应 {code,msg,data}，无 response 包裹
- response = _load(content_id)
- post = parse_detail_response(response, fallback_content_id=content_id)
- assert post.id == f"xhs_{content_id}"
- assert post.platform == "xiaohongshu"
- assert post.content_id == content_id
- assert content_id in post.url
- assert post.title, "title 不应为空"
- _, title_sub = EXPECTED[content_id]
- assert title_sub in post.title
- assert post.raw.get("code") == 0
- def test_authors():
- for cid, (author, _) in EXPECTED.items():
- post = parse_detail_response(_load(cid), fallback_content_id=cid)
- if cid in ("67e4bdf50000000006028a59", "698481e1000000000a02a7c1",
- "67e2e39b0000000003028ff0", "699308fa0000000016009697"):
- assert post.author_name == author
- def test_shiyi_posts_text_sparse():
- """拾意两条正文几乎只有话题串——M3 多模态提取存在的理由。"""
- for cid in ("67e2e39b0000000003028ff0", "680659e8000000001a007a11"):
- post = parse_detail_response(_load(cid), fallback_content_id=cid)
- stripped = post.body_text.replace("#", "").replace("话题", "").strip()
- assert len(stripped) < 40, f"{cid} body_text 应很稀疏: {post.body_text!r}"
- def test_parse_content_id_from_url():
- url = ("https://www.xiaohongshu.com/explore/67e4bdf50000000006028a59"
- "?xsec_token=ABC=&xsec_source=pc_like")
- assert parse_content_id(url) == "67e4bdf50000000006028a59"
- assert parse_content_id("67e4bdf50000000006028a59") == "67e4bdf50000000006028a59"
|
