mostlyright.weather.earnings.ledger

mostlyright.weather.earnings.ledger

Append-only transcript + fact parquet ledger (Phase 27, 27-04).

Two append-only parquet ledgers, one per canonical schema:

TranscriptLedger — schema.earnings_transcript.v1 segment rows. FactLedger — schema.earnings_fact.v1 counted-mention rows.

Audio is NEVER written (D-27.9). The on-disk parquet schema is derived STRICTLY from the canonical schema’s COLUMNS — text + derived-fact columns only. There is NO audio bytes / audio-path column, and a caller-supplied audio key is DROPPED (never persisted), asserted structurally in the test suite. The audio is a transient ingest artifact; only the transcript text and the derived facts cross into the ledger.

Path layout (honors MOSTLYRIGHT_CACHE_DIR):

<cache-root>/earnings/transcripts/<TICKER>/<CALL_ID>.parquet
<cache-root>/earnings/facts/<TICKER>/<CALL_ID>.parquet

Append semantics: a filelock-guarded read-modify-write merge (mirrors the CWOP / satellite cache tiers) — an append concatenates onto any rows already persisted for (ticker, call_id) and rewrites the partition atomically via .tmp + os.replace. This is deliberately a standalone island: it does NOT import the parity-coupled weather/cache.py (the earnings ledger has no current-month-skip / LST semantics and must never touch the parity firewall).

Classes

FactLedger([root])	Append-only schema.earnings_fact.v1 counted-mention ledger (no audio).
TranscriptLedger([root])	Append-only schema.earnings_transcript.v1 segment ledger (no audio).

class mostlyright.weather.earnings.ledger.FactLedger(root=None)

Bases: _ParquetLedger

Append-only schema.earnings_fact.v1 counted-mention ledger (no audio).

• Parameters: root (Path | str | None)

class mostlyright.weather.earnings.ledger.TranscriptLedger(root=None)

Bases: _ParquetLedger

Append-only schema.earnings_transcript.v1 segment ledger (no audio).

• Parameters: root (Path | str | None)