From a95b187edc2b795aa201027e47566ab0563aeb2e Mon Sep 17 00:00:00 2001
From: lrnselfreliance <33263708+lrnselfreliance@users.noreply.github.com>
Date: Sun, 11 Oct 2026 16:20:44 -0600
Subject: [PATCH] A yt-dlp format file is not the primary file when merged
 media is present (#817)

`get_primary_file` returned the first video/audio file in whatever order it
was given, so a group holding both `<stem>.mp4` and a leftover video-only
`<stem>.f313.webm` could make the leftover the primary file.  The Video then
pointed at a file without audio, and a download of it was deferred with
"Download did not contain audio stream" on every retry.

Skip yt-dlp per-format files (`.fNNN` stems) when a merged media file is in
the group.  A group of only format files still gets one as its primary.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
---
 wrolpi/files/lib.py           | 13 ++++++++++---
 wrolpi/files/test/test_lib.py | 17 +++++++++++++++++
 2 files changed, 27 insertions(+), 3 deletions(-)

diff --git a/wrolpi/files/lib.py b/wrolpi/files/lib.py
index 7a9aca64..68dc8d37 100644
--- a/wrolpi/files/lib.py
+++ b/wrolpi/files/lib.py
@@ -694,6 +694,10 @@ def _paths_to_files_dict(group: List[pathlib.Path]) -> List[dict]:
     return files
 
 
+# The stem of a yt-dlp per-format file, which is merged into the final media file.  Example:  The Title.f313.webm
+YTDLP_FORMAT_STEM = re.compile(r'\.f\d{2,4}$')
+
+
 def get_primary_file(files: Union[Tuple[pathlib.Path], Iterable[pathlib.Path]]) -> pathlib.Path:
     """Given a list of files, return the file that we can model or index.
 
@@ -720,13 +724,16 @@ def get_primary_file(files: Union[Tuple[pathlib.Path], Iterable[pathlib.Path]])
     # EPUB has high priority.
     mimetypes = {i[1] for i in file_mimetypes}
     has_epub = any(i.startswith('application/epub') for i in mimetypes)
+    # A yt-dlp per-format file (video-only or audio-only) is primary only when no merged media file is present.
+    has_merged_media = any(m.startswith(('video/', 'audio/')) and not YTDLP_FORMAT_STEM.search(f.stem)
+                           for f, m in file_mimetypes)
     for file, mimetype in file_mimetypes:
         # These are the files that can be modeled.
         if is_singlefile_file(file):
             return file
-        if mimetype.startswith('video/'):
-            return file
-        if mimetype.startswith('audio/'):
+        if mimetype.startswith(('video/', 'audio/')):
+            if has_merged_media and YTDLP_FORMAT_STEM.search(file.stem):
+                continue
             return file
         if mimetype.startswith('application/epub'):
             return file
diff --git a/wrolpi/files/test/test_lib.py b/wrolpi/files/test/test_lib.py
index 7072803e..b8028fc0 100644
--- a/wrolpi/files/test/test_lib.py
+++ b/wrolpi/files/test/test_lib.py
@@ -1,3 +1,4 @@
+import itertools
 import json
 import pathlib
 import shutil
@@ -995,6 +996,22 @@ def test_get_primary_file(test_directory, video_file, srt_file3, example_epub, e
         assert lib.get_primary_file([foo, bar])
 
 
+def test_get_primary_file_skips_ytdlp_format_files(test_directory):
+    """A yt-dlp per-format file (`<stem>.f313.webm`) is never the primary file when the merged media is present."""
+    merged = test_directory / 'a video.mp4'
+    video_only = test_directory / 'a video.f313.webm'
+    audio_only = test_directory / 'a video.f251.webm'
+    for path in (merged, video_only):
+        shutil.copy(PROJECT_DIR / 'test/big_buck_bunny_720p_1mb.mp4', path)
+    shutil.copy(PROJECT_DIR / 'test/big_buck_bunny.mp3', audio_only)
+
+    for files in itertools.permutations([merged, video_only, audio_only]):
+        assert lib.get_primary_file(list(files)) == merged
+
+    # Only format files remain (an unmerged download); one must still be the primary.
+    assert lib.get_primary_file([video_only, audio_only]) in (video_only, audio_only)
+
+
 @pytest.mark.asyncio
 async def test_refresh_files_no_groups(async_client, test_session, test_directory, make_files_structure,
                                        zip_file_factory, refresh_files):