feat(dashboard): add forward "what's next" signal + fix upcoming floor

The daily-file bill queue is a backward snapshot, so "upcoming" could only
resurface the latest past day (the 08-31 snapshot read as today's queue) and
there was no "what is scheduled next." leginfo bill status/history are
backward-only, so the genuine future-dated signal is the chambers' "upcoming
events" calendars.

- api.py: floor the "upcoming" view at *today* (not the DB max date) when no
  as_of is pinned; surface next_legislative_day + scheduled_events (next 7
  days) on /bills. Add _today_iso / _next_legislative_day / _scheduled_events,
  all degrading to a recess marker / [] (never 5xx).
- relational.py: schema v2 adds the scheduled_events table +
  store_scheduled_events (idempotent per-day refresh), load_scheduled_events,
  next_legislative_day.
- parsers.py: parse_events_page + classify_event_kind for the chamber events
  calendar; skip empty-title rows (decompose the <time> before the fallback).
- ingest.py + cli.py: --store-events lane (cache-first, recess-safe) that
  parses the calendar HTML into scheduled_events.
- sources.py + config.yaml: register assembly_events / senate_events
  (kind: events, empty_is_ok: true).
- run_ingest.sh: run the events lane after the daily-file lane as non-fatal
  (skippable via CALI_SCRAPER_SKIP_EVENTS=1); daily-file exit still gates cron.
- frontend: "Next legislative day" line + "What's next · next 7 days" band
  with honest recess empty states.
- DASHBOARD.md: document the forward signal + known limitations (leginfo
  backward-only gap, recess empty state, 08-31 snapshot is analysis-only).
- tests: 5 new API tests + 18 parse-events tests; 170 pass, 92% coverage.
This commit is contained in:
kfj001 committed 2026-09-06 19:50:02 -07:00
1 parent 1919d43bf0
commit 03eeb6bb62
16 files changed
+941 -70

No files matched your search

+41 -4
View File
@@ -75,16 +75,42 @@ list:
marker and a "Floated: ..." reason line
- Zero-scored bills remain in their date-grouped order below
## Forward signal: "what's next" (scheduled_events, schema v2)
The daily-file bill queue is a *backward* snapshot — leginfo's bill status and
bill history are past-action logs, so they cannot answer "what is scheduled
next." The only genuine future-dated signal is the chambers' "upcoming events"
calendars (`assembly.ca.gov/events/upcoming`, `senate.ca.gov/calendar`). These
are fetched (`python3 -m scraper ingest --store-events`), parsed into the
`scheduled_events` table (relational schema v2), and surfaced on `/bills` as two
new envelope keys:
- `next_legislative_day` — `{date, kind, chamber, title, source}` (the earliest
event on/after `as_of`), or `{date: null, recess: true}` when nothing is
scheduled.
- `scheduled_events` — the next 7 days of chamber events.
The "Upcoming" filter is a *forward* view: with `upcoming=1` and no explicit
`as_of`, the bill queue floors at **today**, so a stale/older daily-file
snapshot can no longer present itself as "upcoming." Omitting `upcoming`
(the "All" / history view) still returns every row.
The dashboard renders a "Next legislative day" line in the freshness band and a
"What's next · next 7 days" band. A recess (empty calendar) is an *honest*
empty state, not an error.
## Tests
```bash
python3 -m pytest tests/test_dashboard_ranking.py tests/test_dashboard_api.py
python3 -m pytest tests/test_dashboard_ranking.py tests/test_dashboard_api.py \
tests/test_parse_events.py
```
32 tests (12 ranking + 20 API), all passing. Pure ranking logic is tested in
isolation; API handlers are tested with no sockets via the
55 tests (12 ranking + 25 API + 18 parse-events), all passing. Pure ranking
logic is tested in isolation; API handlers are tested with no sockets via the
`handler(req) -> (status, body)` pattern; a live `ThreadingHTTPServer` smoke test
confirms static + API endpoints serve.
confirms static + API endpoints serve. The event parser is tested against
inline fixtures plus the saved `samples/assembly_events_*.html` pages.
## DASHBOARD.md
@@ -110,3 +136,14 @@ cannot be auto-approved).
no roster data is supplied.
- Dev auth is a JSON-file local account, not production auth.
- ZIP -> district uses a documented demo dataset until roster ETL lands.
- **Forward signal gap:** leginfo's bill status/history are *backward-only*
(past actions + status narrative, no future dates), so there is no per-bill
"next step" label. The dashboard's forward answer comes solely from the
chamber events calendars (`scheduled_events`).
- **Recess empty state:** while the Assembly (and at times the Senate) is in
final recess, `/events/upcoming` legitimately returns **0 events**. The
"What's next" band and `next_legislative_day` then show an honest "recess /
unknown" state rather than fabricating a date. This is expected, not a
parsing failure.
- The 2026-08-31 daily-file snapshot is analysis-only; it is *not* re-surfaced
as "upcoming" because the upcoming view floors at today.