Home / Files / strm.py

strm.py — removed (pre-split URL builders & show fetchers)

Removed — module no longer exists

The strm.py module was deleted in the 3-package split. Its responsibilities were absorbed into the shared runtime:

Provider show aggregation + _nav_* tagging now happens in library_sync.py's per-provider sync branches (_sync_movies / _sync_series).
The plugin:// URL synthesis is folded into _insert_movie / _insert_episode in library_sync.py.
Provider catalogue scrapers ship in easyplaytv-service/src/easyplaytv_service/vod_sync/providers/.

The detail below is retained purely as historical record of the pre-split architecture — none of the functions documented exist anymore.

Path removed Lines removed Type URL synthesis + provider-listing aggregator (deleted) ← called by none — module deleted → calls none — module deleted

Role

Two responsibilities (the module name strm is historical — the addon does not write .strm files on disk; all episode and movie files live as plugin:// URLs in PLUGIN_SOURCE_PATH rows in the Kodi DB):

Show aggregation — call every provider's list_shows for every unified genre and return a single flat list with provider, _genre_id, _nav_mode, and the provider-specific _nav_* fields baked in.
URL synthesis — given one of those show dicts (or a single episode dict from list_episodes), build a plugin://plugin.video.easyplaytv/?action=resolve_vod&provider=…&episode_id=…&… URL that Kodi's plugin entry point can route to the right resolver.

Module-level constants (lines 11–15)

STRM_MOVIES_DIR / STRM_SERIES_DIR / STATE_FILE defined but unused

STRM_MOVIES_DIR and STRM_SERIES_DIR are defined at module load but never referenced by any function in this module — the addon does not write .strm files on disk; all episode and movie file rows live as plugin://… URLs under PLUGIN_SOURCE_PATH in the Kodi DB. STATE_FILE points at strm_state.json, but the live state file the sync engine reads is library_sync_state.json + series_sync_state.json (see library_sync.py). The three constants exist purely for historical traceability — safe to delete them in a refactor.

STRM_DIR (line 12) is still used — through _plugin_lib at line 17 it sets up sys.path so the bare-name from genres import … at line 23 resolves whether the module is imported from inside Kodi or run standalone.

Provider import dispatcher

_import_vod(provider) → module | None · line 28

Lazy-imports the right *_vod module for a provider name. The try/except pattern handles two import conventions — Kodi's resources.lib.bbc_vod vs. the same module imported bare (bbc_vod) for tests run outside Kodi.

Films / series fetchers

_fetch_movies() → list[dict] · line 67

The movie catalog aggregator. Iterates UNIFIED_GENRES for the films genre only, then per-provider per-genre call lists the films with provider-appropriate pagination:

Provider	Pagination
BBC	Page-based loop until `cur_page >= total_pages`
My5	Offset-based with 100-row limit, stops when `offset >= total_entries`
ITVX	Single call — ITVX paginates internally and returns the full category payload
C4	Single call with `limit=9999` — C4's API caps at ~10k rows
STV	Filters to `show.standalone == True` only (single-film programmes)

Each show dict gets s["provider"] = "" added so downstream code knows where it came from.

_fetch_series_shows() → list[dict] · line 284

The series equivalent — iterates every unified genre except films. More importantly, it tags each show with _nav_mode (either "playable" for standalone/movie-like shows or "series" for shows with multiple episodes) and the provider-specific _nav_* fields that _sync_series uses to fetch seasons/episodes:

Provider	Series mode fields	Playable mode condition
BBC	`_nav_pid`	`show.get("standalone") and show.get("episode_id")`
My5	`_nav_f_name`	`show.get("standalone")`
ITVX	`_nav_prog_id`, `_nav_slug`	`show.get("standalone") and show.get("episode_id")`
C4	`_nav_f_name`	always series (C4 doesn't tag standalone at API level)
STV	`_nav_f_name`	`show.get("standalone")`

The _nav_* pre-pass meant the sync layer didn't have to know each provider's id-field naming.

Movie URL builder

_build_strm_url(show) → str | None · line 119

Builds a play URL for a movie show. Returns None if the show isn't standalone (it's actually a series — movies that are multi-episode specials shouldn't be treated as a single-play movie). For providers where every film needs an episode lookup (C4, STV), this function does a live season-list call to find the single episode:

BBC: resolve_vod&provider=bbc&episode_id=&title=…
My5: resolve_vod&provider=my5&episode_id=&standalone=1&title=…
ITVX: resolve_vod&provider=itvx&episode_id=…&title=…&playlist_url=
C4: calls c4_vod.list_seasons(f_name), requires exactly 1 episode, builds URL with both programme_id and asset_id
STV: iterates stv_vod.list_seasons(), then iterates the result for the first video_id it finds with a DRM flag

Note: All URLs are URL-encoded via safe_filename(show["title"]) for the title parameter.

Known bug — STV branch (line 173)

The STV branch references episodes at line 173 without defining it — looks like a refactor removed episodes = vod.list_episodes(...) but missed the if not episodes: return None guard. The code will throw a NameError on first STV movie. In practice nobody's noticed because STV's "films" catalogue is mostly empty in the production library. Worth fixing; tag a one-liner episodes = vod.list_episodes(f_name, series_guid=series_guid) back in.

Episode URL builder

_build_episode_strm_url(provider, episode) → str | None · line 201

Builds a play URL for a single episode within a series. Always includes the season + episode numbers in the URL via the _se() helper (lines 216-222) — even though the resolver doesn't strictly need them, the plugin entry-point reads them onto the listitem's setInfo so Kodi caches consistent SxxExx metadata for the played file. Without them, the displayed 1x01. prefix disappears after playback because the player's metadata overwrites the folder's listitem metadata (see the inline comment at line 205).

Provider	Extra params beyond `episode_id`, `title`, `season`, `episode`
BBC	`show_title` (extracted via `_bbc_ep_title`)
My5	`standalone=0/1`, `show_f_name`, `f_name` (if different)
ITVX	`playlist_url` (URL-encoded)
C4	`asset_id`
STV	`drm_enabled=0/1`

_bbc_ep_title(ep, en_int) → str · line 190

Helper for BBC episode titles. BBC programmes often have a "Series N:" subtitle pattern; this strips that prefix and falls back to the bare subtitle or just "Episode N". Used only inside _build_episode_strm_url for BBC.

Cross-file interactions

Importing safe_filename for title normalisation across all URL builders.
Calling c4_vod.list_seasons() and stv_vod.list_seasons() inline within _build_strm_url — these were sync-time HTTP fetches, so the cost was folded into the slower of the two paths (movie catalog fetch).
Returning flat lists that the sync layer filtered by _nav_mode == "series" and provider.