Migrate the BookStream 001/002 batches (65 files)

[docxology]

• Series: BookStream
• Files: 65
• Estimated time: ~150 minutes
• Labels: good-first-issue, i16-translations

BookStream_001 still holds 46 legacy files across six stems ("Active Inference BookStream 001.010/012/02/06/08/1 ~ Governing Continuous Transformation…"). Each token matches exactly one of the item's 12 same-titled parts; the extra "Active Inference" prefix defeated the M2 auto-matcher. Two files (001.010, 001.08 transcripts) have no language suffix — inspect content (likely en). BookStream_002 holds 19 more across two stems: "002.0 ~ …Chapters 1, 2, 3, 6" → part FV7pW4p60VI (#2.01 overview) and "002.02 ~ …Chapters 4, 5, 7, 8" → part 3_5pTCguAv4 (#2.02 overview); the chapter list in each stem matches the part title exactly.

Scope (from the triage doc, Part A):

• Item BookStream/BookStream_001
  • 1 files, stem Active Inference BookStream 001.010 ~ Governing Continuous Transformation_transcript → target translations/eIZjx0miM9o.<bcp47>.srt
  • 11 files, stem Active Inference BookStream 001.012 ~ Governing Continuous Transformation → target translations/5IGzzm28qec.<bcp47>.srt
  • 11 files, stem Active Inference BookStream 001.02 ~ Governing Continuous Transformation → target translations/d8c0iFU8vms.<bcp47>.srt
  • 11 files, stem Active Inference BookStream 001.06 ~ Governing Continuous Transformation → target translations/0P_6ME2LpCw.<bcp47>.srt
  • 1 files, stem Active Inference BookStream 001.08 ~ Governing Continuous Transformation_transcript → target translations/yNZg5b63hb8.<bcp47>.srt
  • 11 files, stem Active Inference BookStream 001.1 ~ Governing Continuous Transformation → target translations/FB_93-zDqNo.<bcp47>.srt
• Item BookStream/BookStream_002
  • 8 files, stem Active Inference BookStream 002.0 ~ Parr, Pezzulo, Friston ~ Chapters 1, 2, 3, 6 → target translations/FV7pW4p60VI.<bcp47>.srt
  • 11 files, stem Active Inference BookStream 002.02 ~ Parr, Pezzulo, Friston ~ Chapters 4, 5, 7, 8 → target translations/3_5pTCguAv4.<bcp47>.srt

Read first: docs/volunteer-transcript-tasks.md —
sections "The 30-minute task (per file)" and "Exact command pattern".

Steps, per file:

1. Verify the guessed video_id in the item's metadata.json + a YouTube spot-check of the part.
2. Confirm the language by opening the .srt (normalize legacy codes per the migration map).
3. Move with the two-step git mv into translations/<video_id>.<bcp47>.srt.
4. Append the original repo-relative path to the destination item's metadata.json previous_paths.
5. One path-scoped commit per item: translations(<series>): migrate <N> legacy files (I16 triage).

Acceptance criteria:

• every listed file moved (or deleted, if byte-identical to an already-migrated target) with previous_paths recorded
• no listed file remains in the item's Translations/ folder; empty Translations/ and translations_tmp/ removed
• spot-checked target files open as valid subtitles and match the verified part/language