Generalize date-numbered episode search

[rix1337]

What changed

• Parse Sonarr year + MM/DD numbering once into a validated shared episode_date.
• Centralize date query variants, title matching, exact-date validation, and Arr title canonicalization in shared utilities.
• Enable existing date searches for BY, DD, DJ, DL, DT, DW, FX, HE, HS, MB, NK, NX, RM, SF, SJ, SL, WD, and WX.
• Keep AL/AT disabled because their anime-specific numbering paths require numeric episodes; keep MX disabled because its API requires numeric season and episode; FF remains movie-only.
• Remove WWE/RAW/SmackDown-specific production logic while preserving both motivating WWE series through generic schedule-word and mixed-case handling.

Why

PR #400 established date-numbered searching in DL, but left source-specific date fields and WWE-specific aliases. This follow-up makes date numbering a shared search capability without changing any source endpoint, HTTP method, parameter name, payload shape, or response parser.

Impact

• Normal movie, season/episode, document, music, and feed paths remain unchanged when episode_date is absent.
• No config or hostname changes.
• Compatible TV sources reuse their existing IMDb/title lookup and candidate-result paths.

Verification

• uv run python -X utf8 pre-commit.py: passed; 149 tests.
• uv run ruff check .: passed.
• uv run ruff format --check .: passed.
• Latest main (c965cc5): 134 tests and Ruff passed in a detached worktree.
• Identical nine-case non-date matching corpus on latest main and this branch: movie acceptance/rejection, exact/wrong TV episode, season pack, movie-as-TV rejection, document result, and episode-range acceptance/rejection.
• Synthetic DL thread tests find and canonicalize both WWE RAW and WWE SmackDown date-numbered cases.
• Production scan contains no WWE or SmackDown literals outside the untouched obfuscated asset.
• Source transport diff audit found no endpoint/request/payload changes.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 0024036c13

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Strip query dates before canonicalizing release titles

When DL searches a date-numbered episode it iterates generated queries from date_numbering_search_strings(), including variants like Sample Show 2031.02.02/03/04, and passes that query into this helper. Because the prefix is built from the whole query, a match titled Sample.Show.2031.02.03.1080p-GRP can be returned as Sample.Show.2031.02.02.2031.02.03.1080p-GRP (or duplicate the same date), so the result name sent to Sonarr is malformed and may be rejected or misparsed.

Useful? React with 👍 / 👎.