fix(皮皮搞笑): 提取帖子图集,补齐 imgs.urls 原图直链

This commit is contained in:
ucmao
2026-09-19 09:31:38 +08:00
parent a65b188153
commit 280cebf6ca
4 changed files with 121 additions and 5 deletions
+14 -5
View File
@@ -7,9 +7,10 @@
## 1. 平台特征与支持能力
* **平台标识**:`皮皮搞笑`
* **支持媒体类型**:无水印视频 (MP4) / 帖子标题 / 封面
* **支持媒体类型**:无水印视频 (MP4) / 图集 / 帖子文案 / 封面 / 作者
* **常见链接形态**:
* 移动分享页:`https://h5.pipigx.com/pp/post/815491325984?pid=815491325984&type=post`
* 空间/新版分享页:`https://h5.ippzone.com/spacey/post/897800342766?pid=897800342766&type=post`
* **Cookie 依赖**:无需 Cookie。
---
@@ -17,13 +18,21 @@
## 2. 核心逆向流程
1. **ID 提取**:从 Query 参数或 Path 中提取 `pid` / `post_id`。
2. **H5 数据接口**:
* 接口:`POST https://api.pipigx.com/ppapi/cell/detail`
* 或直接在 H5 落地页 HTML 中提取内嵌 JSON。
3. **视频直链提取**:从 `post.video.video_download_url` 或 `post.video.video_url` 中提取 MP4 直链。
2. **H5 分享接口**:
* 接口:`POST https://h5.pipigx.com/ppapi/share/fetch_content`
* 请求体:`{"mid": null, "pid": <pid>, "type": "post"}`(JSON,表单方式会返回"请求参数错误")
* 响应:`data.post`,其中 `content` 为文案、`imgs` 为图片/视频素材、`videos` 为视频直链映射。
3. **视频直链提取**:`post.videos[str(imgs[0].id)].url`。
4. **图集提取**:`post.imgs` 中 `video` 不为真的条目即配图,直链取
`imgs[i].urls.origin.urls[0]`(原图),缺失时依次回落到 `540` / `360` 变体,
最后回落 `https://file.ippzone.com/img/view/id/{imgs[i].id}`。
该 URL 目录下没有文件名,是按图片 ID 动态返回的。
> 视频帖的 `imgs[0]` 是视频本身(`video: 1`,且没有 `urls`),不要当成配图返回。
---
## 3. 测试与验证
* **单元测试**:`python -m unittest tests/test_pipigaoxiao_parser.py`
* **样本验证**:`python tests/manual_verify_parsers.py --platform 皮皮搞笑`
+27
View File
@@ -64,6 +64,33 @@ class PipigaoxiaoParser(BaseParser):
logger.warning(f"Failed to parse Pipigaoxiao description: {e}")
return None
def get_image_list(self):
"""提取帖子图集;imgs 里 video=1 的是视频本身,不是配图。"""
try:
if not self.data: return []
imgs = self.data.get('data', {}).get('post', {}).get('imgs') or []
image_list = []
for img in imgs:
if not isinstance(img, dict) or img.get('video'):
continue
# 优先原图,其次 540/360 压缩图,最后回落到按图片 ID 拼的地址
urls = img.get('urls') or {}
candidates = []
for key in ('origin', '540', '360'):
variant = urls.get(key) or {}
candidates.extend(variant.get('urls') or [])
url = next((u for u in candidates if u), None)
if not url and img.get('id'):
url = f"https://file.ippzone.com/img/view/id/{img['id']}"
if url:
image_list.append(url)
return image_list
except Exception as e:
logger.warning(f"Failed to parse Pipigaoxiao image list: {e}")
return []
def get_cover_photo_url(self):
try:
if not self.data: return ""
+14
View File
@@ -580,6 +580,20 @@
"note": "2026-08-28 已验证",
"pattern": "基础标准样例"
},
{
"platform": "皮皮搞笑",
"url": "https://h5.ippzone.com/spacey/post/897800342766?pid=897800342766&type=post",
"media_types": [
"images"
],
"expected_fields": [
"images",
"cover",
"author"
],
"note": "2026-09-19 用户反馈修复: 空间/新版分享页 8 图图集",
"pattern": "spacey 图集"
},
{
"platform": "皮皮虾",
"url": "https://h5.pipix.com/s/lggjR-ynbpI/",
+66
View File
@@ -0,0 +1,66 @@
import unittest
from unittest.mock import patch
from src.parsers.pipigaoxiao_parser import PipigaoxiaoParser
def _image_node(img_id, video=0, live=0, origin=True):
node = {
"id": img_id,
"w": 1080,
"h": 1441,
"video": video,
"live": live,
"urls": {},
}
if origin:
node["urls"] = {
"origin": {"urls": [f"http://bd-file.ippzone.com/img/view/id/{img_id}/sz/src",
f"http://file.ippzone.com/img/view/id/{img_id}/sz/src"]},
"540": {"urls": [f"http://file.ippzone.com/img/view/id/{img_id}/sz/540"]},
}
return node
class PipigaoxiaoParserTest(unittest.TestCase):
def _parser_with(self, post):
with patch.object(PipigaoxiaoParser, "fetch_html_data", return_value={"data": {"post": post}}):
return PipigaoxiaoParser("https://h5.ippzone.com/spacey/post/897800342766")
def test_image_post_returns_all_origin_images(self):
parser = self._parser_with({
"content": "",
"videos": None,
"imgs": [_image_node(2538988134), _image_node(2538988128), _image_node(2538988129)],
})
self.assertEqual(parser.get_image_list(), [
"http://bd-file.ippzone.com/img/view/id/2538988134/sz/src",
"http://bd-file.ippzone.com/img/view/id/2538988128/sz/src",
"http://bd-file.ippzone.com/img/view/id/2538988129/sz/src",
])
self.assertIsNone(parser.get_real_video_url())
def test_video_post_does_not_report_images(self):
parser = self._parser_with({
"content": "350。。。小钱钱。。晚上继续喝",
"videos": {"2405682288": {"url": "http://video01.szsttkj.com/a/b/c"}},
"imgs": [_image_node(2405682288, video=1, origin=False)],
})
self.assertEqual(parser.get_image_list(), [])
self.assertEqual(parser.get_real_video_url(), "http://video01.szsttkj.com/a/b/c")
def test_falls_back_to_id_url_when_variants_missing(self):
parser = self._parser_with({"imgs": [_image_node(2538988134, origin=False)]})
self.assertEqual(parser.get_image_list(),
["https://file.ippzone.com/img/view/id/2538988134"])
def test_empty_data_is_safe(self):
parser = self._parser_with({})
self.assertEqual(parser.get_image_list(), [])
if __name__ == "__main__":
unittest.main()