Map studio/site into Plex Collection tags (with dedupe) and cover performer role thumbs

[Copilot]

This updates the in-progress TPDB→Plex mapping fix so studio/site is also emitted as a Plex Collection tag in both match and full metadata payloads. It preserves existing studio behavior and avoids duplicate collection entries when studio already appears in series/collection/franchise data.

• Collection mapping update

  • Updated provider/mappers/tpdb_to_plex.py to include _get_studio(scene) in _get_collections(scene).
  • Reuses existing seen-based dedupe, so studio is only added once even if already present in incoming collection fields.
  • Because both map_scene_to_match(...) and map_scene_to_metadata(...) consume _get_collections(scene), the behavior is consistent across both endpoints.

• Studio field compatibility

  • Kept current studio field population unchanged in match/metadata output.
  • Change is additive: studio remains in studio and is now also available in Collection for Plex browsing workflows.

• Test coverage additions

  • Extended tests/test_tpdb_enrichment.py to assert:
    • studio is present in Collection for metadata output,
    • studio is deduped when already present in collection-like fields,
    • performer images map to Role[].thumb,
    • performers without images still map cleanly without thumb data.

def _get_collections(scene: dict[str, Any]) -> list[dict[str, str]]:
    candidates = []
    for key in ("collections", "collection", "series", "franchise", "franchises"):
        ...
    studio = _get_studio(scene)
    if studio:
        candidates.append(studio)
    ...

Original prompt

Update the in-progress fix for adeze/plex-tpdb-scanner to also map the studio/site name into Plex collections.

Additional requirement from the user:

• Map the Studio tag to Collections because that is more functional for their Plex usage.

Repository context already identified:

• TPDB-to-Plex mapping lives in provider/mappers/tpdb_to_plex.py
• _get_studio(scene) extracts the studio/site name
• _get_collections(scene) extracts collection/series/franchise names
• map_scene_to_match(...) and map_scene_to_metadata(...) currently set studio separately and populate Collection from _get_collections(scene) only

Please implement the following in the PR:

1. Ensure the studio/site name is included in Plex Collection output for both match results and full metadata.
2. Avoid duplicates if the studio/site name is already present in collections.
3. Preserve current studio field behavior unless there is a strong compatibility reason not to.
4. Update or add tests to cover the studio-to-collection mapping behavior.
5. Keep the implementation simple and consistent with the existing Python style in the repo.

This should be added to the existing PR/task work rather than creating a separate unrelated change.

The following is the prior conversation context from the user's chat exploration (may be truncated):

User: how do i rebuild the container via docker compose?

Assistant: From the repo root, run:

docker compose up -d --build

That works because docker-compose.yml uses build: . for the tpdb-provider service, so Compose will rebuild the image from the local Dockerfile and recreate the container.

If you want to force a clean rebuild, use:

docker compose build --no-cache
docker compose up -d

Or in one go:

docker compose up -d --build --force-recreate

A few repo-specific notes:

• docker-compose.yml is the file to use for local rebuilds.
• docker-compose.portainer.yml uses a prebuilt image (ghcr.io/mystrock/plex-tpdb-scanner:latest), so that one does not rebuild from your local source.
• You’ll also need TPDB_API_KEY set in your shell or a .env file before starting.

Typical flow:

git clone https://github.com/adeze/plex-tpdb-scanner.git
cd plex-tpdb-scanner
export TPDB_API_KEY=your_api_key_here
docker compose up -d --build

To verify it came back up:

docker compose ps
docker compose logs -f

If you want, I can also show you the best docker compose workflow for Apple Silicon specifically, including when to use --no-cache, down, and volume cleanup.

User: do i need to turn off the current running container? how?

Assistant: Yes — usually you should stop and remove the currently running container before bringing up the rebuilt one, though docker compose up -d --build will often recreate it for you automatically.

Simplest safe approach from the repo directory:

docker compose down
docker compose up -d --build

What these do:

• docker compose down stops and removes the existing container, network, and Compose-managed resources.
• docker compose up -d --build rebuilds the image and starts a fresh container in the background.

If you want minimal steps, this is often enough by itself:

docker compose up -d --build

Compose will detect the config/image change and replace the running container.

If you specifically want to stop just the running service first:

docker compose stop
docker compose rm -f
docker compose up -d --build

Useful checks:

docker compose ps
docker compose logs -f tpdb-provider

If the old container was started outside Compose, remove it manually:

docker stop tpdb-provider
docker rm tpdb-provider

Recommended for your case: use this sequence:

cd plex-tpdb-scanner
docker compose down
docker compose up -d --build

If you want, I can give you the exact command sequence for macOS zsh with your TPDB_API_KEY included.

User: i think there's an error since the logs show:

INFO:     172.67.149.54:31411 - "GET /library/metadata/vrmassaged-fit-redhead-is-craving-your-dick-pt2/images?X-Plex-Country=US&X-Plex-Language=en-US HTTP/1.1" 404 Not Found

2026-06-04 07:09:19,450 - provider.main - INFO - >>> REQUEST: GET /library/metadata/vrmassaged-fit-redhead-is-craving-your-dick-pt2/images

2026-06-04 07:09:19,451 - provider.main - INFO - <<< RESPONSE: GET /library/metadata/vrmassaged-fit-redhead-is-craving-your-dick-pt2/images -> 404

INFO:     172.67.149.54:56456 - "GET /library/metadata/vrmassaged-fit-redhead-is-craving-your-dick-pt2/images?X-Plex-Country=US&X-Plex-Language=en-US HTTP/1.1" 404 Not Found

shouldn't it be looking for the uuid or something?

Assistant: Yes — I think you found a real bug.

Short version: Plex is requesting:

GET /library/metadata/<slug>/images

but this repo only implements:

GET /library/metadata/{rating_key}
POST /library/metadata/matches
GET /

So the 404...

This pull request was created from Copilot chat.

[adeze]

ok can you ensure the performers images are being downloaded too ?

[Copilot]

Pull request overview

This PR updates the TPDB→Plex mapping to also include the TPDB studio/site name in Plex Collection tags (while preserving the existing studio field), and adds/updates tests to verify the new behavior and deduplication.

Changes:

• Include studio/site name in _get_collections(...) so both match results and full metadata emit it via Collection.
• Deduplicate studio vs existing collection/series/franchise names.
• Update/add tests to assert studio-to-collection mapping and dedupe behavior.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
provider/mappers/tpdb_to_plex.py	Appends extracted studio/site name into the collections candidate set (deduped) so it flows into Collection for match + metadata.
tests/test_tpdb_enrichment.py	Updates existing metadata collection expectations and adds a new test covering studio→collection dedupe in match mapping.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.