parsehub 2.3.16__tar.gz → 2.3.17__tar.gz

{parsehub-2.3.16/src/parsehub.egg-info → parsehub-2.3.17}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: parsehub
-Version: 2.3.16
+Version: 2.3.17
 Summary: 轻量、异步、开箱即用的社交媒体聚合解析库
 Author-email: 梓澪 <zilingmio@gmail.com>
 License: MIT

{parsehub-2.3.16 → parsehub-2.3.17}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "parsehub"
-version = "2.3.16"
+version = "2.3.17"
 description = "轻量、异步、开箱即用的社交媒体聚合解析库"
 readme = "README.md"
 requires-python = ">=3.12.0"

{parsehub-2.3.16 → parsehub-2.3.17}/src/parsehub/parsers/base/ytdlp.py

@@ -411,7 +411,7 @@ class YtParser(BaseParser, register=False):
         title = dl["title"]
         duration = dl.get("duration", 0)
         thumbnail = dl.get("thumbnail")
-        description = dl["description"]
+        description = dl.get("description", "")
         width = dl.get("width", 0)
         height = dl.get("height", 0)
         return YtVideoInfo(

{parsehub-2.3.16 → parsehub-2.3.17}/src/parsehub/parsers/parser/instagram.py

@@ -71,11 +71,12 @@ class InstagramParser(BaseParser):
                         title=post.title,
                         content=post.caption,
                         video=VideoRef(
-                            url=info.url,
+                            # yt-dlp 的 info.url 是帖子页，provider 的 video_url 才是媒体直链
+                            url=item.video_url or info.url,
                             thumb_url=item.display_url or info.thumbnail,
                             width=info.width,
                             height=info.height,
-                            duration=int(info.duration or item.duration),
+                            duration=int(item.duration or info.duration),
                         ),
                     )
                 if not item.video_url:

{parsehub-2.3.16 → parsehub-2.3.17}/src/parsehub/parsers/parser/zuiyou.py

@@ -15,7 +15,7 @@ class ZuiYouParser(BaseParser):
         return MultimediaParseResult(
             content=zy.content,
             media=[
-                VideoRef(url=i.url, thumb_url=i.thumb_url)
+                VideoRef(url=i.url, thumb_url=i.thumb_url, duration=i.duration)
                 if i.type == MediaType.VIDEO
                 else ImageRef(url=i.url, thumb_url=i.thumb_url)
                 for i in zy.media

{parsehub-2.3.16 → parsehub-2.3.17}/src/parsehub/provider_api/instagram.py

@@ -116,11 +116,54 @@ class InstagramAPI:
                 cookies=self.cookie or None,
             ) as session:
                 if self.cookie:
-                    return await self._get_cookie_post(session, shortcode)
-                return await self._get_logged_out_post(session, shortcode)
+                    post = await self._get_cookie_post(session, shortcode)
+                else:
+                    post = await self._get_logged_out_post(session, shortcode)
+                return await self._fill_missing_video_fields(session, post, shortcode)
         except requests.RequestsError as exc:
             raise InstagramAPIError(f"请求 Instagram 失败: {exc}") from exc
 
+    async def _fill_missing_video_fields(
+        self,
+        session: requests.AsyncSession,
+        post: InstagramPost,
+        shortcode: str,
+    ) -> InstagramPost:
+        """用 embed 补齐 v1 / GraphQL 缺失的视频直链和时长。"""
+        if not any(
+            (post.type is InstagramMediaType.VIDEO or item.is_video)
+            and (not item.video_url or item.duration <= 0)
+            for item in post.media
+        ):
+            return post
+
+        try:
+            node = await self._fetch_embed(session, shortcode)
+        except requests.RequestsError:
+            return post
+        if node is None:
+            return post
+
+        try:
+            embed = InstagramPost.from_graph(node)
+        except InstagramAPIError:
+            return post
+
+        for index, item in enumerate(post.media):
+            if post.type is not InstagramMediaType.VIDEO and not item.is_video:
+                continue
+            source = embed.media[index] if index < len(embed.media) else None
+            if source is None and len(post.media) == 1 and len(embed.media) == 1:
+                source = embed.media[0]
+            if source is None:
+                continue
+            if not item.video_url and source.video_url:
+                item.video_url = source.video_url
+                item.is_video = True
+            if item.duration <= 0 and source.duration > 0:
+                item.duration = source.duration
+        return post
+
     async def _get_logged_out_post(self, session: requests.AsyncSession, shortcode: str) -> InstagramPost:
         detail = "未返回帖子数据"
         media_id = str(self.shortcode_to_media_id(shortcode))

{parsehub-2.3.16 → parsehub-2.3.17}/src/parsehub/provider_api/zuiyou.py

@@ -15,6 +15,7 @@ class Media:
     url: str
     thumb_url: str
     type: MediaType
+    duration: int = 0
 
 
 @dataclass
@@ -34,8 +35,15 @@ class ZuiYouPost:
             id_ = img["id"]
             img_url = list(img["urls"].values())[-1]["urls"][-1]
             if img.get("video"):
-                video_url = videos[str(id_)]["url"]
-                media.append(Media(url=video_url, thumb_url=img_url, type=MediaType.VIDEO))
+                video = videos[str(id_)]
+                media.append(
+                    Media(
+                        url=video["url"],
+                        thumb_url=img_url,
+                        type=MediaType.VIDEO,
+                        duration=int(video.get("dur") or 0),
+                    )
+                )
                 continue
             media.append(Media(url=img_url, thumb_url=img_url, type=MediaType.PHOTO))
 

{parsehub-2.3.16 → parsehub-2.3.17/src/parsehub.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: parsehub
-Version: 2.3.16
+Version: 2.3.17
 Summary: 轻量、异步、开箱即用的社交媒体聚合解析库
 Author-email: 梓澪 <zilingmio@gmail.com>
 License: MIT

{parsehub-2.3.16 → parsehub-2.3.17}/src/parsehub.egg-info/SOURCES.txt

@@ -80,4 +80,6 @@ test/test_bilibili_offline.py
 test/test_cli.py
 test/test_cli_config.py
 test/test_core_offline.py
-test/test_downloader.py
+test/test_downloader.py
+test/test_instagram_duration_offline.py
+test/test_zuiyou_duration_offline.py

parsehub-2.3.17/test/test_instagram_duration_offline.py

@@ -0,0 +1,134 @@
+import unittest
+from unittest.mock import AsyncMock, patch
+
+from parsehub.parsers.base.ytdlp import YtParser, YtVideoInfo
+from parsehub.parsers.parser.instagram import InstagramParser
+from parsehub.provider_api.instagram import InstagramAPI, InstagramMedia, InstagramMediaType, InstagramPost
+
+
+class TestInstagramParserDuration(unittest.IsolatedAsyncioTestCase):
+    async def test_video_prefers_provider_url_and_duration(self):
+        post = InstagramPost(
+            shortcode="abc",
+            type=InstagramMediaType.VIDEO,
+            media=[
+                InstagramMedia(
+                    is_video=True,
+                    display_url="https://cdn.example/thumb.jpg",
+                    video_url="https://cdn.example/v.mp4",
+                    duration=18.018,
+                    width=1920,
+                    height=1080,
+                )
+            ],
+        )
+        info = YtVideoInfo(
+            title="t",
+            description="",
+            thumbnail=None,
+            url="https://www.instagram.com/p/abc/",
+            info_json={},
+            duration=0,
+            width=1920,
+            height=1080,
+        )
+        parser = InstagramParser()
+
+        with (
+            patch.object(InstagramParser, "_parse", AsyncMock(return_value=post)),
+            patch(
+                "parsehub.parsers.parser.instagram.InstagramYtParse._parse",
+                AsyncMock(return_value=info),
+            ),
+        ):
+            result = await parser._do_parse("https://www.instagram.com/reel/abc/")
+
+        self.assertEqual(result.media.url, "https://cdn.example/v.mp4")
+        self.assertEqual(result.media.duration, 18)
+
+
+class TestInstagramAPIDuration(unittest.IsolatedAsyncioTestCase):
+    @staticmethod
+    def _embed_node() -> dict:
+        return {
+            "__typename": "GraphVideo",
+            "shortcode": "abc",
+            "display_url": "https://cdn.example/thumb.jpg",
+            "video_url": "https://cdn.example/v.mp4",
+            "video_duration": 30.933,
+            "dimensions": {"width": 1080, "height": 1920},
+        }
+
+    async def test_fills_duration_from_embed(self):
+        api = InstagramAPI()
+        post = InstagramPost(
+            shortcode="abc",
+            type=InstagramMediaType.VIDEO,
+            media=[
+                InstagramMedia(
+                    is_video=True,
+                    display_url="https://cdn.example/thumb.jpg",
+                    video_url="https://cdn.example/v.mp4",
+                    duration=0,
+                )
+            ],
+        )
+
+        with patch.object(InstagramAPI, "_fetch_embed", AsyncMock(return_value=self._embed_node())):
+            result = await api._fill_missing_video_fields(object(), post, "abc")
+
+        self.assertEqual(result.media[0].duration, 30.933)
+
+    async def test_fills_missing_video_url_from_embed(self):
+        api = InstagramAPI()
+        post = InstagramPost(
+            shortcode="abc",
+            type=InstagramMediaType.VIDEO,
+            media=[
+                InstagramMedia(
+                    is_video=False,
+                    display_url="https://cdn.example/thumb.jpg",
+                    duration=0,
+                )
+            ],
+        )
+
+        with patch.object(InstagramAPI, "_fetch_embed", AsyncMock(return_value=self._embed_node())):
+            result = await api._fill_missing_video_fields(object(), post, "abc")
+
+        self.assertEqual(result.media[0].video_url, "https://cdn.example/v.mp4")
+        self.assertTrue(result.media[0].is_video)
+        self.assertEqual(result.media[0].duration, 30.933)
+
+    async def test_skips_embed_when_complete(self):
+        api = InstagramAPI()
+        post = InstagramPost(
+            shortcode="abc",
+            type=InstagramMediaType.VIDEO,
+            media=[
+                InstagramMedia(
+                    is_video=True,
+                    display_url="https://cdn.example/thumb.jpg",
+                    video_url="https://cdn.example/v.mp4",
+                    duration=18,
+                )
+            ],
+        )
+
+        with patch.object(InstagramAPI, "_fetch_embed", AsyncMock()) as fetch:
+            await api._fill_missing_video_fields(object(), post, "abc")
+
+        fetch.assert_not_awaited()
+
+
+class _FakeYtParser(YtParser, register=False):
+    async def _extract_info(self, url: str) -> dict:
+        return {"title": "t", "duration": 10, "webpage_url": url}
+
+
+class TestYtParserDescription(unittest.IsolatedAsyncioTestCase):
+    async def test_missing_description_does_not_raise(self):
+        info = await _FakeYtParser()._parse("https://example.com/v")
+
+        self.assertEqual(info.description, "")
+        self.assertEqual(info.duration, 10)

parsehub-2.3.17/test/test_zuiyou_duration_offline.py

@@ -0,0 +1,69 @@
+import unittest
+from unittest.mock import AsyncMock, patch
+
+from parsehub.parsers.parser.zuiyou import ZuiYouParser
+from parsehub.provider_api.zuiyou import Media, MediaType, ZuiYou, ZuiYouPost
+
+
+class TestZuiyouDuration(unittest.TestCase):
+    def test_parse_reads_video_dur(self):
+        payload = {
+            "data": {
+                "post": {
+                    "content": "x",
+                    "imgs": [
+                        {
+                            "id": 1,
+                            "urls": {"a": {"urls": ["https://cdn.example/thumb.jpg"]}},
+                            "video": True,
+                        }
+                    ],
+                    "videos": {"1": {"dur": 5, "url": "https://cdn.example/v.mp4"}},
+                }
+            }
+        }
+
+        post = ZuiYouPost.parse(payload)
+
+        self.assertEqual(post.media[0].type, MediaType.VIDEO)
+        self.assertEqual(post.media[0].duration, 5)
+        self.assertEqual(post.media[0].url, "https://cdn.example/v.mp4")
+
+    def test_photo_duration_defaults_to_zero(self):
+        payload = {
+            "data": {
+                "post": {
+                    "imgs": [
+                        {
+                            "id": 1,
+                            "urls": {"a": {"urls": ["https://cdn.example/p.jpg"]}},
+                        }
+                    ]
+                }
+            }
+        }
+
+        post = ZuiYouPost.parse(payload)
+
+        self.assertEqual(post.media[0].duration, 0)
+
+
+class TestZuiyouParserDuration(unittest.IsolatedAsyncioTestCase):
+    async def test_parser_sets_video_duration(self):
+        zy = ZuiYouPost(
+            content="x",
+            media=[
+                Media(
+                    url="https://cdn.example/v.mp4",
+                    thumb_url="https://cdn.example/thumb.jpg",
+                    type=MediaType.VIDEO,
+                    duration=19,
+                )
+            ],
+            raw={},
+        )
+
+        with patch.object(ZuiYou, "parse", AsyncMock(return_value=zy)):
+            result = await ZuiYouParser()._do_parse("https://s.izuiyou.com/s/abc")
+
+        self.assertEqual(result.media[0].duration, 19)