scrapeMM 0.4.1__tar.gz → 0.4.2__tar.gz

{scrapemm-0.4.1/scrapeMM.egg-info → scrapemm-0.4.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: scrapeMM
-Version: 0.4.1
+Version: 0.4.2
 Summary: Flexible multimodal scraper for social media and the open web.
 Author-email: Mark Rothermel <mark.rothermel@tu-darmstadt.de>
 License-Expression: Apache-2.0

{scrapemm-0.4.1 → scrapemm-0.4.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "scrapeMM"
-version = "0.4.1"
+version = "0.4.2"
 description = "Flexible multimodal scraper for social media and the open web."
 readme = "README.md"
 requires-python = ">=3.10"
@@ -33,3 +33,9 @@ namespaces = false  # Disable implicit namespaces
 
 [tool.setuptools.package-data]
 "*" = ["*.txt"]  # Include all .txt files contained in any (sub)package in the build
+
+[tool.pytest.ini_options]
+testpaths = [
+    "testing/",
+]
+asyncio_default_fixture_loop_scope = "session"

{scrapemm-0.4.1 → scrapemm-0.4.2/scrapeMM.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: scrapeMM
-Version: 0.4.1
+Version: 0.4.2
 Summary: Flexible multimodal scraper for social media and the open web.
 Author-email: Mark Rothermel <mark.rothermel@tu-darmstadt.de>
 License-Expression: Apache-2.0

{scrapemm-0.4.1 → scrapemm-0.4.2}/scrapemm/common/scraping_response.py

@@ -8,7 +8,7 @@ from ezmm import MultimodalSequence
 class ScrapingResponse:
     url: str  # The input URL that was scraped
     content: Optional[MultimodalSequence | str]  # The retrieved content. None if retrieval failed.
-    errors: dict[str, Optional[Exception]] = None  # The exceptions raised during retrieval for each method
+    errors: dict[str, Optional[Exception]] | None = None  # The exceptions raised during retrieval for each method
     method: Optional[str] = None  # The successful method used to retrieve the content
 
     @property

{scrapemm-0.4.1 → scrapemm-0.4.2}/scrapemm/integrations/base.py

@@ -15,7 +15,7 @@ class RetrievalIntegration(ABC):
 
     name: str
     domains: list[str]  # The domains supported by this integration
-    connected: bool = None
+    connected: bool | None = None
 
     @abstractmethod
     async def _connect(self):

{scrapemm-0.4.1 → scrapemm-0.4.2}/scrapemm/integrations/bluesky.py

@@ -42,7 +42,12 @@ class Bluesky(RetrievalIntegration):
 
         return result
 
-    async def _retrieve_post(self, url: str, session: aiohttp.ClientSession = None, max_video_size: int = None) -> Optional[MultimodalSequence]:
+    async def _retrieve_post(
+        self,
+        url: str,
+        session: aiohttp.ClientSession | None = None,
+        max_video_size: int | None = None
+    ) -> Optional[MultimodalSequence]:
         """Retrieve a post from the given Bluesky URL."""
         uri = self._construct_uri(url)
         if not uri:

{scrapemm-0.4.1 → scrapemm-0.4.2}/scrapemm/integrations/fb.py

@@ -15,7 +15,7 @@ VIDEO_URL_REGEX = r"facebook\.com/\d+/videos/\d+/?"
 
 class Facebook(RetrievalIntegration):
     name = "Facebook"
-    domains = ["facebook.com", "www.facebook.com"]
+    domains = ["facebook.com", "fb.watch"]
 
     async def _connect(self):
         self.api_available = False  # TODO
@@ -52,9 +52,12 @@ class Facebook(RetrievalIntegration):
 
     def _is_video_url(self, url: str) -> bool:
         """Checks if the URL is a Facebook video URL."""
-        # video URLS are in the format: https://www.facebook.com/watch?v=VIDEO_ID
+        # video URLS are in the format: https://www.facebook.com/watch?v=VIDEO_ID or fb.watch/...
         # or Reels: https://www.facebook.com/reel/REEL_ID
-        return "facebook.com/watch" in url or "facebook.com/reel" in url or bool(re.search(VIDEO_URL_REGEX, url))
+        return ("facebook.com/watch" in url
+                or "facebook.com/reel" in url
+                or bool(re.search(VIDEO_URL_REGEX, url))
+                or "fb.watch" in url)
 
     def _extract_video_id(self, url: str) -> str:
         """Extracts the video ID from a Facebook video URL."""

{scrapemm-0.4.1 → scrapemm-0.4.2}/scrapemm/integrations/instagram.py

@@ -28,7 +28,10 @@ class Instagram(RetrievalIntegration):
             return await self._get_video(url, **kwargs)
         elif self._is_photo_url(url):
             # /p/ URLs can also be reels, so try both
-            content = await self._get_video(url, **kwargs)
+            try:
+                content = await self._get_video(url, **kwargs)
+            except Exception:
+                content = None
             if content and content.has_videos():
                 return content
             else:
@@ -51,8 +54,7 @@ class Instagram(RetrievalIntegration):
 
     async def _get_photo(self, url: str, **kwargs) -> MultimodalSequence | None:
         """Retrieves content from an Instagram photo URL (can also be a reel)."""
-        logger.warning("❌ Native Instagram photo download not yet supported. Use Decodo for that.")
-        return None
+        raise NotImplementedError("Native Instagram photo download not yet supported. Use Decodo for that.")
 
     async def _get_user_profile(self, url: str, **kwargs) -> MultimodalSequence | None:
         """Retrieves content from an Instagram user profile URL."""

{scrapemm-0.4.1 → scrapemm-0.4.2}/scrapemm/integrations/telegram.py

@@ -31,11 +31,12 @@ class Telegram(RetrievalIntegration):
         if api_id and api_hash and bot_token:
             self.client = TelegramClient(self.session_path, api_id, api_hash)
             try:
-                self.client.start(bot_token=bot_token)
+                await self.client.start(bot_token=bot_token)  # Returns a coroutine b/c event loop exists already
             except sqlite3.OperationalError:  # Database is locked from an interrupted previous session
                 # Remove the database file and try again
                 journal_path = Path(self.session_path + ".session-journal")
                 journal_path.unlink(missing_ok=True)
+                await self.client.start(bot_token=bot_token)  # Returns a coroutine b/c event loop exists already
             self.connected = True
             logger.info("✅ Successfully connected to Telegram.")
         else:
@@ -56,8 +57,12 @@ class Telegram(RetrievalIntegration):
 
         # Get the message
         channel = await self.client.get_entity(channel_name)
-        message = await self.client.get_messages(channel, ids=post_id)
+        if not channel:
+            return None
 
+        assert isinstance(channel, Channel), f"Can only retrieve posts from Telegram channels, but got {type(channel).__name__}."
+
+        message = await self.client.get_messages(channel, ids=post_id)
         if not message:
             return None
 

{scrapemm-0.4.1 → scrapemm-0.4.2}/scrapemm/integrations/ytdlp.py

@@ -10,7 +10,7 @@ import aiohttp
 from ezmm import MultimodalSequence, download_image, Video, Image
 from yt_dlp import YoutubeDL
 
-from scrapemm.common.exceptions import IPBannedError
+from scrapemm.common.exceptions import IPBannedError, ContentNotFoundError
 
 logger = logging.getLogger("scrapeMM")
 
@@ -23,8 +23,8 @@ logger_yt_dlp.addHandler(logging.StreamHandler(sys.stdout))
 async def download_video_with_ytdlp(
         url: str,
         session: aiohttp.ClientSession,
-        max_video_size: int = None,
-        cookie_file: str = None
+        max_video_size: int | None = None,
+        cookie_file: str | None = None
 ) -> tuple[Optional[Video], Optional[Image], Optional[dict[str, Any]]]:
     """Downloads a video and (if not available or exceeds max. duration) its thumbnail, and the metadata using yt-dlp.
     @param max_video_size: Maximum video size in bytes. If the video is larger, the download will be aborted."""
@@ -79,6 +79,10 @@ async def download_video_with_ytdlp(
         if "The following content is not available on this app" in str(e):
             logger.warning(f"You should update yt-dlp to re-enable YouTube downloads.")
             raise e
+        elif "Video unavailable" in str(e):
+            raise ContentNotFoundError(f"Video is unavailable.")
+        elif "Cannot parse data; please report this issue" in str(e):
+            raise RuntimeError(f"yt-dlp is unable to parse the received video metadata.")
         else:
             raise RuntimeError(f"Could not download video with yt-dlp: {e}")
 

{scrapemm-0.4.1 → scrapemm-0.4.2}/scrapemm/retrieval.py

@@ -20,7 +20,7 @@ async def retrieve(
         remove_urls: bool = False,
         show_progress: bool = True,
         actions: list[dict] = None,
-        methods: list[str] = None,
+        methods: list[str] | list[list[str]] = None,
         format: str = "multimodal_sequence",
         max_video_size: int | None = None,
 ) -> ScrapingResponse | list[ScrapingResponse]:
@@ -39,24 +39,24 @@ async def retrieve(
         - "integrations" (API integrations for Twitter, Instagram, etc.)
         - "firecrawl" (Firecrawl scraping service)
         - "decodo" (Decodo Web Scraping API)
-        You can specify any subset in any order, e.g., ["decodo", "firecrawl"] or ["integrations"].
+        You can specify any subset in any order, e.g., ["decodo", "firecrawl"] or ["integrations"]. If provided
+        a list of strings, that order of methods will be applied to all submitted URLs. In contrast, if provided
+        a list of lists, each list will be applied to the corresponding URL in the batch.
     :param format: The format of the output. Available formats:
         - "multimodal_sequence" (MultimodalSequence containing parsed and downloaded media from the page)
         - "html" (string containing the raw HTML code of the page, not compatible with 'integrations' method)
     :param max_video_size: Maximum size of videos to download, in MB. If None, no limit is applied.
     """
     if methods is None:
-        methods = METHODS
+        methods = METHODS.copy()  # Use copy to avoid modifying the original list
     else:
-        assert len(methods) >= 1
-        for method in methods:
-            assert method in METHODS
+        assert isinstance(methods, list) and len(methods) >= 1, "'methods' must be a non-empty list."
 
     assert isinstance(urls, (str, list)), "'urls' must be a string or a list of strings."
 
     # Ensure URLs are string or list
     single_url = isinstance(urls, str)
-    urls_to_retrieve = [urls] if single_url else urls
+    urls_to_retrieve: list[str] = [urls] if single_url else urls
 
     if len(urls_to_retrieve) == 0:
         return []
@@ -64,15 +64,24 @@ async def retrieve(
     if actions:
         raise NotImplementedError("Actions are not supported yet.")
 
-    async with aiohttp.ClientSession() as session:
-        # Remove duplicates
-        urls_unique = set(urls_to_retrieve)
+    # Build per-URL method dict according to the provided 'methods'
+    if isinstance(methods[0], str):
+        # methods: list[str] → apply same order to all URLs
+        url_to_methods = {url: methods[:] for url in urls_to_retrieve}
+    elif isinstance(methods[0], list):
+        # methods: list[list[str]] → each inner list corresponds to the URL at the same index
+        url_to_methods = dict(zip(urls_to_retrieve, methods))
+    else:
+        raise AssertionError("'methods' must be either a list[str] or a list[list[str]]")
+
+    urls_unique = set(urls_to_retrieve)
 
+    async with aiohttp.ClientSession() as session:
         # Retrieve URLs concurrently
-        tasks = [_retrieve_single(url, remove_urls, session, methods, actions,
+        tasks = [_retrieve_single(url, remove_urls, session, url_to_methods[url], actions,
                                   format, max_video_size) for url in
                  urls_unique]
-        results = await run_with_semaphore(tasks, limit=20, show_progress=show_progress and len(urls_to_retrieve) > 1,
+        results = await run_with_semaphore(tasks, limit=20, show_progress=show_progress and len(urls_unique) > 1,
                                            progress_description="Retrieving URLs...")
 
         # Reconstruct output list
@@ -87,17 +96,24 @@ async def _retrieve_single(
         url: str,
         remove_urls: bool,
         session: aiohttp.ClientSession,
-        methods: list[str],
+        methods: list[str] = None,
         actions: list[dict] = None,
         format: str = "multimodal_sequence",
         max_video_size: int | None = None,
 ) -> ScrapingResponse:
     logger.debug(f"Retrieving {url}")
 
+    if methods is None:
+        methods = METHODS.copy()
+
     try:
         # Ensure URL is a string
         url = str(url)
 
+        # Validate methods
+        for method in methods:
+            assert method in METHODS, f"Unknown method '{method}'. Allowed: {METHODS}"
+
         # Ensure compatibility with methods
         if format == "html" and "integrations" in methods:
             methods.remove("integrations")

{scrapemm-0.4.1 → scrapemm-0.4.2}/scripts/example.py

@@ -2,10 +2,9 @@ from scrapemm import retrieve
 import asyncio
 
 if __name__ == "__main__":
-    url = "https://www.tiktok.com/@realdonaldtrump/video/7433870905635409198"
+    url = "https://www.awesomescreenshot.com/image/43774028?key=0de2147873b73ed468d07aeb93512c51"
     result = asyncio.run(retrieve(url))
     if result.errors:
         print(result.errors)
     else:
         print(result.content)
-        print(str(result.content))

scrapemm-0.4.2/testing/test_retrieval.py

@@ -0,0 +1,100 @@
+import pytest
+from ezmm import MultimodalSequence
+from scrapemm.common import ScrapingResponse
+
+from scrapemm import retrieve
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize("url", [
+    "https://www.vishvasnews.com/viral/fact-check-upsc-has-not-reduced-the-maximum-age-limit-for-ias-and-ips-exams/",
+    "https://health.medicaldialogues.in/fact-check/brain-health-fact-check/fact-check-is-sprite-the-best-remedy-for-headaches-in-the-world-140368",
+    "https://www.washingtonpost.com/politics/2024/05/15/bidens-false-claim-that-inflation-was-9-percent-when-he-took-office/",
+    "https://assamese.factcrescendo.com/viral-claim-that-the-video-shows-the-incident-from-uttar-pradesh-and-the-youth-on-the-bike-and-the-youth-being-beaten-and-taken-away-by-the-police-are-the-same-youth-named-abdul-is-false/",
+    "https://factuel.afp.com/doc.afp.com.43ZN7NP",
+    "https://leadstories.com/365cb414b83e29d26fecae374d55c743a3eac4c7.png",
+    "https://leadstories.com/assets_c/2025/08/193f14f06dd6f15b89bf8050e553ad7fb1be6530-thumb-900xauto-3165872.png"
+])
+@pytest.mark.parametrize("method", ["firecrawl", "decodo"])
+async def test_generic_retrieval(url, method):
+    result = await retrieve(url, methods=[method])
+    assert isinstance(result, ScrapingResponse)
+    print(result)
+    assert result
+    content = result.content
+    assert isinstance(content, MultimodalSequence)
+    assert content.has_images()
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize("url", [
+    "https://www.zeit.de/politik/deutschland/2025-07/spionage-iran-festnahme-anschlag-juden-berlin-daenemark",
+    "https://factnameh.com/fa/fact-checks/2025-04-16-araghchi-witkoff-fake-photo",
+    "https://www.thip.media/health-news-fact-check/fact-check-can-a-kalava-on-the-wrist-prevent-paralysis/74724/",
+])
+@pytest.mark.parametrize("method", ["firecrawl", "decodo"])
+async def test_html_retrieval(url, method):
+    result = await retrieve(url, format="html", methods=[method])
+    assert isinstance(result, ScrapingResponse)
+    content = result.content
+    print(content)
+    assert content
+    assert isinstance(content, str)
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize("url, max_video_size, download_expected", [
+    ("https://www.facebook.com/reel/1089214926521000", None, True),
+    ("https://www.facebook.com/reel/1089214926521000", 128_000_000, True),
+    ("https://www.facebook.com/reel/1089214926521000", 1_000_000, False),
+    ("https://www.youtube.com/shorts/cE0zgN6pYOc", None, True),
+    ("https://www.youtube.com/shorts/cE0zgN6pYOc", 4_000_000, True),
+    ("https://www.youtube.com/shorts/cE0zgN6pYOc", 3_000_000, False),
+])
+async def test_max_video_size(url, max_video_size, download_expected):
+    result = await retrieve(url, max_video_size=max_video_size)
+    assert isinstance(result, ScrapingResponse)
+    content = result.content
+    assert isinstance(content, MultimodalSequence)
+    assert content.has_videos() == download_expected
+    if max_video_size and content.has_videos():
+        video = content.videos[0]
+        assert video.size <= max_video_size
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize("urls, methods", [
+    ([
+         "https://www.youtube.com/shorts/cE0zgN6pYOc"
+     ], [
+         "integrations"
+     ]),
+    ([
+         "https://www.facebook.com/reel/1089214926521000",
+         "https://www.zeit.de/politik/deutschland/2025-07/spionage-iran-festnahme-anschlag-juden-berlin-daenemark",
+     ], [
+         ["integrations"],
+         ["firecrawl"]
+     ]),
+    ([
+         "https://theconversation.com/dandelions-are-a-lifeline-for-bees-on-the-brink-we-should-learn-to-love-them-204504",
+         "https://yussus.wixsite.com/newsspoilers/post/leaked-durex-to-launch-reversible-condom-to-circumvent-single-use-plastic-law",
+         "https://www.bbc.co.uk/iplayer/episode/b09zg78h/question-time-2018-28062018#t=19m38s",
+     ], [
+         "decodo",
+     ]),
+    ([
+         "https://factuel.afp.com/doc.afp.com.43ZN7NP",
+         "https://x.com/realDonaldTrump"
+     ],
+     None
+    ),
+])
+async def test_methods(urls: list[str], methods: list[str] | list[list[str]] | None):
+    results = await retrieve(urls, methods=methods)
+    assert results
+    if methods:
+        if isinstance(methods[0], str):
+            methods = [methods] * len(urls)
+        for result, method_list in zip(results, methods):
+            assert result.method in method_list

scrapemm-0.4.1/testing/test_retrieval.py

@@ -1,163 +0,0 @@
-import pytest
-from ezmm import MultimodalSequence
-
-from scrapemm import retrieve
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://www.vishvasnews.com/viral/fact-check-upsc-has-not-reduced-the-maximum-age-limit-for-ias-and-ips-exams/",
-    "https://health.medicaldialogues.in/fact-check/brain-health-fact-check/fact-check-is-sprite-the-best-remedy-for-headaches-in-the-world-140368",
-    "https://www.washingtonpost.com/politics/2024/05/15/bidens-false-claim-that-inflation-was-9-percent-when-he-took-office/",
-    "https://assamese.factcrescendo.com/viral-claim-that-the-video-shows-the-incident-from-uttar-pradesh-and-the-youth-on-the-bike-and-the-youth-being-beaten-and-taken-away-by-the-police-are-the-same-youth-named-abdul-is-false/",
-    "https://factuel.afp.com/doc.afp.com.43ZN7NP",
-    "https://leadstories.com/365cb414b83e29d26fecae374d55c743a3eac4c7.png",
-    "https://leadstories.com/assets_c/2025/08/193f14f06dd6f15b89bf8050e553ad7fb1be6530-thumb-900xauto-3165872.png"
-])
-@pytest.mark.parametrize("method", ["firecrawl", "decodo"])
-async def test_generic_retrieval(url, method):
-    result = await retrieve(url, methods=[method])
-    print(result)
-    assert result
-    content = result.content
-    assert isinstance(content, MultimodalSequence)
-    assert content.has_images()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://www.zeit.de/politik/deutschland/2025-07/spionage-iran-festnahme-anschlag-juden-berlin-daenemark",
-    "https://factnameh.com/fa/fact-checks/2025-04-16-araghchi-witkoff-fake-photo",
-    "https://www.thip.media/health-news-fact-check/fact-check-can-a-kalava-on-the-wrist-prevent-paralysis/74724/",
-])
-@pytest.mark.parametrize("method", ["firecrawl", "decodo"])
-async def test_html_retrieval(url, method):
-    result = await retrieve(url, format="html", methods=[method])
-    content = result.content
-    print(content)
-    assert content
-    assert isinstance(content, str)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://t.me/durov/404",  # One image
-    "https://t.me/tglobaleye/16172",  # Multiple images
-    "https://t.me/tglobaleye/16178",  # Video and quote
-    "https://t.me/tglobaleye/6289",  # GIF (treated as video)
-    "https://t.me/tglobaleye/16192",  # Images and video
-])
-async def test_telegram(url):
-    result = await retrieve(url)
-    content = result.content
-    print(content)
-    assert content
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://www.tiktok.com/@realdonaldtrump/video/7433870905635409198",
-    "https://www.tiktok.com/@xxxx.xxxx5743/video/7521704371109793046"
-])
-async def test_tiktok(url):
-    result = await retrieve(url)
-    content = result.content
-    print(content)
-    assert content
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://x.com/PopBase/status/1938496291908030484",
-    "https://x.com/realDonaldTrump"
-])
-async def test_x(url):
-    result = await retrieve(url)
-    content = result.content
-    print(content)
-    assert content
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://www.instagram.com/p/CqJDbyOP839",
-])
-async def test_instagram_images(url):
-    result = await retrieve(url)
-    content = result.content
-    print(content)
-    assert content
-    assert content.has_images()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://www.instagram.com/reel/DKqPQqpTDW4",
-    "https://www.instagram.com/reel/C75nh7Lvo8F",
-    "https://www.instagram.com/p/DMuOe6th94D",  # yes, this is a video
-])
-async def test_instagram_videos(url):
-    result = await retrieve(url)
-    content = result.content
-    print(content)
-    assert content
-    assert content.has_videos()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://www.facebook.com/reel/1089214926521000",
-    "https://www.facebook.com/reel/3466446073497470",  # restricted for misinformation
-    "https://www.facebook.com/61561558177010/videos/1445957793080961/",
-    "https://www.facebook.com/watch/?v=1445957793080961",
-    "https://www.facebook.com/groups/1973976962823632/posts/3992825270938781/",  # restricted for misinformation, yt-dlp fails here
-])
-async def test_facebook_videos(url):
-    result = await retrieve(url)
-    content = result.content
-    print(content)
-    assert content
-    assert content.has_videos()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://www.facebook.com/photo/?fbid=1721085455188778&set=a.107961589834514&_rdc=1&_rdr",
-])
-async def test_facebook_images(url):
-    result = await retrieve(url)
-    content = result.content
-    print(content)
-    assert content
-    assert content.has_images()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url", [
-    "https://www.youtube.com/watch?v=A4dVOznX6Kk",
-    "https://www.youtube.com/shorts/cE0zgN6pYOc",
-])
-async def test_youtube(url):
-    result = await retrieve(url)
-    content = result.content
-    print(content)
-    assert content
-    assert content.has_videos()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("url, max_video_size, download_expected", [
-    ("https://www.facebook.com/reel/1089214926521000", None, True),
-    ("https://www.facebook.com/reel/1089214926521000", 128_000_000, True),
-    ("https://www.facebook.com/reel/1089214926521000", 1_000_000, False),
-    ("https://www.youtube.com/shorts/cE0zgN6pYOc", None, True),
-    ("https://www.youtube.com/shorts/cE0zgN6pYOc", 4_000_000, True),
-    ("https://www.youtube.com/shorts/cE0zgN6pYOc", 3_000_000, False),
-])
-async def test_max_video_size(url, max_video_size, download_expected):
-    result = await retrieve(url, max_video_size=max_video_size)
-    content = result.content
-    assert content.has_videos() == download_expected
-    if max_video_size and content.has_videos():
-        video = content.videos[0]
-        assert video.size <= max_video_size