parsehub 2.3.15__tar.gz → 2.3.17__tar.gz

{parsehub-2.3.15/src/parsehub.egg-info → parsehub-2.3.17}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: parsehub
-Version: 2.3.15
+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-2.3.15 → parsehub-2.3.17}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "parsehub"
-version = "2.3.15"
+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-2.3.15 → 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-2.3.15 → parsehub-2.3.17}/src/parsehub/parsers/parser/instagram.py

@@ -1,5 +1,7 @@
 import re
 
+from loguru import logger
+
 from ...provider_api.instagram import (
     InstagramAPI,
     InstagramAPIError,
@@ -19,6 +21,7 @@ from ...types import (
 )
 from ...utils.helpers import SecretCookie
 from ..base.base import BaseParser, FallbackStage
+from ..base.ytdlp import YtParser, YtVideoParseResult
 
 
 class InstagramParser(BaseParser):
@@ -34,29 +37,57 @@ class InstagramParser(BaseParser):
 
         post = await self._parse(shortcode)
 
-        width, height = post.width, post.height
-
-        match post.typename:
+        match post.type:
             case InstagramMediaType.SIDECAR:
                 media = [
-                    VideoRef(url=i.video_url, thumb_url=i.display_url, width=i.width, height=i.height)
-                    if i.is_video and i.video_url
-                    else ImageRef(url=i.display_url, width=i.width, height=i.height)
-                    for i in post.get_sidecar_nodes()
+                    VideoRef(
+                        url=item.video_url,
+                        thumb_url=item.display_url,
+                        duration=int(item.duration),
+                        width=item.width,
+                        height=item.height,
+                    )
+                    if item.is_video
+                    else ImageRef(url=item.display_url, width=item.width, height=item.height)
+                    for item in post.media
                 ]
                 return MultimediaParseResult(media=media, title=post.title, content=post.caption)
             case InstagramMediaType.IMAGE:
+                item = post.media[0]
                 return ImageParseResult(
-                    photo=[ImageRef(url=post.url, width=width, height=height)], title=post.title, content=post.caption
+                    photo=[ImageRef(url=item.display_url, width=item.width, height=item.height)],
+                    title=post.title,
+                    content=post.caption,
                 )
             case InstagramMediaType.VIDEO:
+                item = post.media[0]
+                try:
+                    info = await InstagramYtParse(proxy=self.proxy)._parse(f"https://www.instagram.com/p/{shortcode}/")
+                except ParseError as e:
+                    logger.warning(f"Instagram yt-dlp 解析失败, 使用直链视频: {e}")
+                else:
+                    return YtVideoParseResult(
+                        dl=info,
+                        title=post.title,
+                        content=post.caption,
+                        video=VideoRef(
+                            # 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(item.duration or info.duration),
+                        ),
+                    )
+                if not item.video_url:
+                    raise ParseError("未获取到视频地址")
                 return VideoParseResult(
                     video=VideoRef(
-                        url=post.video_url or post.url,
-                        thumb_url=post.url,
-                        duration=int(post.video_duration or 0),
-                        width=width,
-                        height=height,
+                        url=item.video_url,
+                        thumb_url=item.display_url,
+                        duration=int(item.duration),
+                        width=item.width,
+                        height=item.height,
                     ),
                     title=post.title,
                     content=post.caption,
@@ -80,14 +111,10 @@ class InstagramParser(BaseParser):
             return await api.get_post(shortcode)
         except ContentUnavailableError:
             raise
-        except Exception as e:
-            if isinstance(e, InstagramRestrictedError):
-                text = str(e)
-            elif isinstance(e, InstagramAPIError):
-                text = f"无法获取帖子内容: {e}"
-            else:
-                text = f"无法获取帖子内容(可能为私人内容): {e}"
-            raise ParseError(text) from e
+        except InstagramRestrictedError as e:
+            raise ParseError(str(e)) from e
+        except InstagramAPIError as e:
+            raise ParseError(f"无法获取帖子内容: {e}") from e
 
     @staticmethod
     def get_short_code(url: str) -> str | None:
@@ -96,4 +123,8 @@ class InstagramParser(BaseParser):
         return shortcode.group(2).split("/")[0] if shortcode else None
 
 
+class InstagramYtParse(YtParser, register=False):
+    pass
+
+
 __all__ = ["InstagramParser"]

{parsehub-2.3.15 → 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-2.3.17/src/parsehub/provider_api/instagram.py

@@ -0,0 +1,399 @@
+from __future__ import annotations
+
+import json
+import re
+from dataclasses import dataclass
+from enum import StrEnum
+from typing import Any
+from urllib.parse import urlparse
+
+from curl_cffi import requests
+
+from ..types import ContentUnavailableError
+
+
+class InstagramAPIError(RuntimeError):
+    """Instagram 接口请求或响应解析失败。"""
+
+
+class InstagramRestrictedError(InstagramAPIError):
+    """Instagram 内容受限。"""
+
+
+class InstagramMediaType(StrEnum):
+    IMAGE = "GraphImage"
+    VIDEO = "GraphVideo"
+    SIDECAR = "GraphSidecar"
+
+
+@dataclass(slots=True, kw_only=True)
+class InstagramMedia:
+    is_video: bool
+    display_url: str
+    video_url: str = ""
+    width: int = 0
+    height: int = 0
+    duration: float = 0
+
+
+@dataclass(slots=True, kw_only=True)
+class InstagramPost:
+    shortcode: str
+    type: InstagramMediaType
+    caption: str = ""
+    title: str = ""
+    media: list[InstagramMedia]
+
+    @classmethod
+    def from_v1(cls, item: dict[str, Any]) -> InstagramPost:
+        media_type = _v1_type(item)
+        if media_type is InstagramMediaType.SIDECAR:
+            media_items = item.get("carousel_media") or []
+            if not media_items:
+                raise InstagramAPIError("未返回帖子数据")
+        else:
+            media_items = [item]
+        if not all(isinstance(media_item, dict) for media_item in media_items):
+            raise InstagramAPIError("Instagram 帖子数据格式无效")
+        return cls(
+            shortcode=str(item.get("code") or item.get("shortcode") or ""),
+            type=media_type,
+            caption=_caption_text(item.get("caption")),
+            title=str(item.get("title") or ""),
+            media=[_v1_media(media_item) for media_item in media_items],
+        )
+
+    @classmethod
+    def from_graph(cls, node: dict[str, Any]) -> InstagramPost:
+        media_type = _graph_type(node)
+        if media_type is InstagramMediaType.SIDECAR:
+            edges = (node.get("edge_sidecar_to_children") or {}).get("edges") or []
+            if not edges:
+                raise InstagramAPIError("未返回帖子数据")
+            media = [_graph_media(edge.get("node") or {}) for edge in edges]
+        else:
+            media = [_graph_media(node)]
+
+        caption_edges = (node.get("edge_media_to_caption") or {}).get("edges") or []
+        caption = caption_edges[0].get("node", {}).get("text", "") if caption_edges else ""
+        return cls(
+            shortcode=str(node.get("shortcode") or node.get("code") or ""),
+            type=media_type,
+            caption=str(caption or node.get("caption") or ""),
+            title=str(node.get("title") or ""),
+            media=media,
+        )
+
+
+class InstagramAPI:
+    HOME_URL = "https://www.instagram.com/"
+    GRAPHQL_URL = "https://www.instagram.com/api/graphql"
+    EMBED_URL = "https://www.instagram.com/{kind}/{shortcode}/embed/captioned/"
+    MEDIA_INFO_URL = "https://www.instagram.com/api/v1/media/{media_id}/info/"
+    RULING_URL = "https://www.instagram.com/api/v1/web/get_ruling_for_content/"
+    APP_ID = "936619743392459"
+    ASBD_ID = "359341"
+    DOC_ID = "27130156389949648"
+    MOBILE_USER_AGENT = "Mozilla/5.0 (iPhone; CPU iPhone OS 16_0 like Mac OS X) AppleWebKit/605.1.15"
+
+    def __init__(
+        self,
+        *,
+        proxy: str | None = None,
+        cookie: dict[str, str] | None = None,
+        timeout: float = 30,
+    ):
+        self.proxy = proxy
+        self.cookie = cookie or {}
+        self.timeout = timeout
+
+    async def get_post(self, shortcode: str) -> InstagramPost:
+        try:
+            async with requests.AsyncSession(
+                impersonate="chrome",
+                proxy=self.proxy,
+                timeout=self.timeout,
+                cookies=self.cookie or None,
+            ) as session:
+                if self.cookie:
+                    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))
+
+        try:
+            item = await self._fetch_graphql(session, shortcode, media_id)
+            if item is not None:
+                return InstagramPost.from_v1(item)
+        except InstagramAPIError as exc:
+            detail = str(exc)
+        except requests.RequestsError as exc:
+            detail = f"请求 Instagram 失败: {exc}"
+
+        try:
+            node = await self._fetch_embed(session, shortcode)
+            if node is not None:
+                return InstagramPost.from_graph(node)
+        except InstagramAPIError as exc:
+            detail = str(exc)
+        except requests.RequestsError as exc:
+            detail = f"请求 Instagram Embed 失败: {exc}"
+
+        await self._check_ruling(session, media_id)
+        raise InstagramAPIError(detail)
+
+    async def _fetch_graphql(
+        self, session: requests.AsyncSession, shortcode: str, media_id: str
+    ) -> dict[str, Any] | None:
+        response = await session.get(self.HOME_URL)
+        if response.status_code != 200:
+            raise InstagramAPIError(f"HTTP {response.status_code}: {response.text[:300]}")
+
+        match = re.search(r'\["LSD",\[\],\{"token":"([^"]+)"', response.text)
+        if not match:
+            raise InstagramAPIError(f"未能获取 Instagram LSD token: {response.text[:300]}")
+        lsd = match.group(1)
+        csrf = session.cookies.get("csrftoken", "")
+        headers = {
+            "X-IG-App-ID": self.APP_ID,
+            "X-ASBD-ID": self.ASBD_ID,
+            "X-IG-WWW-Claim": "0",
+            "Origin": self.HOME_URL.rstrip("/"),
+            "Accept": "*/*",
+            "X-FB-Friendly-Name": "PolarisLoggedOutDesktopWWWPostRootContentQuery",
+            "X-CSRFToken": csrf,
+            "X-FB-LSD": lsd,
+            "X-Requested-With": "XMLHttpRequest",
+            "Referer": f"https://www.instagram.com/p/{shortcode}/",
+        }
+        data = {
+            "lsd": lsd,
+            "fb_api_caller_class": "RelayModern",
+            "fb_api_req_friendly_name": "PolarisLoggedOutDesktopWWWPostRootContentQuery",
+            "server_timestamps": "true",
+            "variables": json.dumps({"media_id": media_id}, separators=(",", ":")),
+            "doc_id": self.DOC_ID,
+        }
+        response = await session.post(self.GRAPHQL_URL, headers=headers, data=data)
+        if response.status_code != 200:
+            raise InstagramAPIError(f"HTTP {response.status_code}: {response.text[:300]}")
+        payload = _response_json(response)
+        graphql_data = payload.get("data")
+        media = graphql_data.get("xig_polaris_media") if isinstance(graphql_data, dict) else None
+        item = media.get("if_not_gated_logged_out") if isinstance(media, dict) else None
+        if isinstance(item, dict):
+            return item
+        errors = payload.get("errors")
+        if errors:
+            raise InstagramAPIError(str(errors[0])[:300])
+        return None
+
+    async def _fetch_embed(self, session: requests.AsyncSession, shortcode: str) -> dict[str, Any] | None:
+        for kind in ("p", "reel"):
+            response = await session.get(
+                self.EMBED_URL.format(kind=kind, shortcode=shortcode),
+                headers={"Accept": "text/html", "User-Agent": self.MOBILE_USER_AGENT},
+                allow_redirects=False,
+            )
+            if response.status_code != 200:
+                continue
+            node = self._extract_embed_media(response.text)
+            if node is not None:
+                return node
+        return None
+
+    @staticmethod
+    def _extract_embed_media(html: str) -> dict[str, Any] | None:
+        match = re.search(r'"init",\[\],\[(.*?)\]\],', html, re.DOTALL)
+        if not match:
+            return None
+        try:
+            payload = json.loads(f"[{match.group(1)}]")
+            context = payload[0].get("contextJSON")
+            context = json.loads(context) if isinstance(context, str) else context
+        except (IndexError, AttributeError, json.JSONDecodeError, TypeError):
+            return None
+        if not isinstance(context, dict):
+            return None
+        node = (context.get("gql_data") or {}).get("shortcode_media")
+        return node if isinstance(node, dict) else None
+
+    async def _check_ruling(self, session: requests.AsyncSession, media_id: str) -> None:
+        try:
+            response = await session.get(
+                self.RULING_URL,
+                params={"content_type": "MEDIA", "target_id": media_id},
+                headers={"X-IG-App-ID": self.APP_ID, "Accept-Language": "zh-CN,zh;q=0.9"},
+            )
+            body = _response_json(response)
+        except (InstagramAPIError, requests.RequestsError):
+            return
+
+        if body.get("message") == "Media cannot be found":
+            raise ContentUnavailableError("帖子不存在或已被删除")
+        if title := body.get("title"):
+            description = body.get("description")
+            raise InstagramRestrictedError(f"{title}: {description}" if description else str(title))
+
+    async def _get_cookie_post(self, session: requests.AsyncSession, shortcode: str) -> InstagramPost:
+        media_id = self.shortcode_to_media_id(shortcode)
+        response = await session.get(
+            self.MEDIA_INFO_URL.format(media_id=media_id),
+            headers={
+                "X-IG-App-ID": self.APP_ID,
+                "X-ASBD-ID": self.ASBD_ID,
+                "X-IG-WWW-Claim": "0",
+                "X-CSRFToken": self.cookie.get("csrftoken", ""),
+                "Origin": self.HOME_URL.rstrip("/"),
+                "Referer": f"https://www.instagram.com/p/{shortcode}/",
+            },
+            allow_redirects=False,
+        )
+        location = response.headers.get("Location", "")
+        redirect_path = urlparse(location).path
+        if response.status_code in (301, 302, 303, 307, 308) and (
+            redirect_path.startswith("/accounts/login") or redirect_path == "/"
+        ):
+            raise InstagramAPIError("Cookie 已失效")
+        if response.status_code != 200:
+            raise InstagramAPIError(f"HTTP {response.status_code}: {response.text[:300]}")
+
+        payload = _response_json(response)
+        items = payload.get("items") or []
+        if not items or not isinstance(items[0], dict):
+            raise InstagramAPIError("未返回帖子数据")
+        return InstagramPost.from_v1(items[0])
+
+    @staticmethod
+    def shortcode_to_media_id(shortcode: str) -> int:
+        if len(shortcode) > 28:
+            shortcode = shortcode[:-28]
+        alphabet = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-_"
+        media_id = 0
+        for char in shortcode:
+            media_id = media_id * 64 + alphabet.index(char)
+        return media_id
+
+
+def _response_json(response: Any) -> dict[str, Any]:
+    try:
+        payload = response.json()
+    except (ValueError, requests.RequestsError) as exc:
+        raise InstagramAPIError(f"Instagram 返回无效 JSON: {response.text[:300]}") from exc
+    if not isinstance(payload, dict):
+        raise InstagramAPIError("Instagram 返回的数据格式无效")
+    return payload
+
+
+def _caption_text(caption: Any) -> str:
+    if isinstance(caption, dict):
+        caption = caption.get("text")
+    return str(caption or "")
+
+
+def _v1_type(item: dict[str, Any]) -> InstagramMediaType:
+    media_types = {
+        1: InstagramMediaType.IMAGE,
+        2: InstagramMediaType.VIDEO,
+        8: InstagramMediaType.SIDECAR,
+    }
+    media_type = item.get("media_type")
+    if isinstance(media_type, int) and media_type in media_types:
+        return media_types[media_type]
+    raise InstagramAPIError(f"不支持的 Instagram 媒体类型: {media_type}")
+
+
+def _v1_media(item: dict[str, Any]) -> InstagramMedia:
+    candidates = (item.get("image_versions2") or {}).get("candidates") or []
+    candidate = candidates[0] if candidates and isinstance(candidates[0], dict) else {}
+    video_versions = item.get("video_versions") or []
+    video = video_versions[0] if video_versions and isinstance(video_versions[0], dict) else {}
+    video_url = str(video.get("url") or "")
+    return InstagramMedia(
+        is_video=bool(video_url),
+        display_url=str(candidate.get("url") or item.get("display_url") or ""),
+        video_url=video_url,
+        width=int(item.get("original_width") or candidate.get("width") or 0),
+        height=int(item.get("original_height") or candidate.get("height") or 0),
+        duration=float(item.get("video_duration") or 0),
+    )
+
+
+def _graph_type(node: dict[str, Any]) -> InstagramMediaType:
+    typename = node.get("__typename")
+    if not isinstance(typename, str):
+        raise InstagramAPIError(f"不支持的 Instagram 媒体类型: {typename}")
+    try:
+        return InstagramMediaType(typename)
+    except ValueError as exc:
+        raise InstagramAPIError(f"不支持的 Instagram 媒体类型: {typename}") from exc
+
+
+def _graph_media(node: dict[str, Any]) -> InstagramMedia:
+    video_url = str(node.get("video_url") or "")
+    dimensions = node.get("dimensions") or {}
+    return InstagramMedia(
+        is_video=bool(video_url),
+        display_url=str(node.get("display_url") or node.get("display_src") or ""),
+        video_url=video_url,
+        width=int(dimensions.get("width") or 0),
+        height=int(dimensions.get("height") or 0),
+        duration=float(node.get("video_duration") or 0),
+    )
+
+
+__all__ = [
+    "InstagramAPI",
+    "InstagramAPIError",
+    "InstagramMedia",
+    "InstagramMediaType",
+    "InstagramPost",
+    "InstagramRestrictedError",
+]

{parsehub-2.3.15 → 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-2.3.15 → parsehub-2.3.17/src/parsehub.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: parsehub
-Version: 2.3.15
+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-2.3.15 → 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