feat: migrate transform to 3-layer architecture with per-layer schemas

Remove raw/ layer — staging models now read landing JSON directly.
Rename all model schemas from padelnomics.* to staging.*/foundation.*/serving.*.
Web app queries updated to serving.planner_defaults via SERVING_DUCKDB_PATH.
Supervisor gets daily sleep interval between pipeline runs.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Deeman
2026-02-22 19:04:40 +01:00
parent 53e9bbd66b
commit 2db66efe77
19 changed files with 306 additions and 301 deletions

View File

@@ -5,86 +5,121 @@ Fetches raw data from external sources to the local landing zone. The pipeline t
## Running
```bash
# One-shot (most recent data only)
# Run all extractors sequentially
LANDING_DIR=data/landing uv run extract
# First-time full backfill (add your own backfill entry point)
LANDING_DIR=data/landing uv run python -m padelnomics_extract.execute
# Run a single extractor
LANDING_DIR=data/landing uv run extract-overpass
LANDING_DIR=data/landing uv run extract-eurostat
LANDING_DIR=data/landing uv run extract-playtomic-tenants
LANDING_DIR=data/landing uv run extract-playtomic-availability
```
## Architecture: one file per source
Each data source lives in its own module with a dedicated CLI entry point:
```
src/padelnomics_extract/
├── __init__.py
├── _shared.py # LANDING_DIR, logger, run_extractor() wrapper
├── utils.py # SQLite state tracking, atomic I/O helpers
├── overpass.py # OSM padel courts via Overpass API
├── eurostat.py # Eurostat city demographics (urb_cpop1, ilc_di03)
├── playtomic_tenants.py # Playtomic venue listings (tenant search)
├── playtomic_availability.py # Playtomic booking slots (next-day availability)
└── all.py # Runs all extractors sequentially
```
### Adding a new extractor
1. Create `my_source.py` following the pattern:
```python
from ._shared import run_extractor, setup_logging
from .utils import landing_path, write_gzip_atomic
logger = setup_logging("padelnomics.extract.my_source")
EXTRACTOR_NAME = "my_source"
def extract(landing_dir, year_month, conn, session):
"""Returns {"files_written": N, "bytes_written": N, ...}."""
year, month = year_month.split("/")
dest_dir = landing_path(landing_dir, "my_source", year, month)
# ... fetch data, write to dest_dir ...
return {"files_written": 1, "files_skipped": 0, "bytes_written": n}
def main():
run_extractor(EXTRACTOR_NAME, extract)
```
2. Add entry point to `pyproject.toml`:
```toml
extract-my-source = "padelnomics_extract.my_source:main"
```
3. Import in `all.py` and add to `EXTRACTORS` list.
4. Add a staging model in `transform/sqlmesh_padelnomics/models/staging/`.
## Design: filesystem as state
The landing zone is an append-only store of raw files. Each file is named by its content fingerprint (etag or SHA256 hash), so:
The landing zone is an append-only store of raw files:
- **Idempotency**: running twice writes nothing if the source hasn't changed
- **Debugging**: every historical raw file is preserved — reprocess any window by re-running transforms
- **Debugging**: every historical raw file is preserved
- **Safety**: extraction never mutates existing files, only appends new ones
### Etag-based dedup (preferred)
### Etag-based dedup (Eurostat)
When the source provides an `ETag` header, use it as the filename:
When the source provides an `ETag` header, store it in a sibling `.etag` file.
On the next request, send `If-None-Match` — 304 means skip.
```
data/landing/padelnomics/{year}/{month:02d}/{etag}.csv.gz
```
### Content-addressed (Overpass, Playtomic)
The file existing on disk means the content matches the server's current version. No content download needed.
### Hash-based dedup (fallback)
When the source has no etag (static files that update in-place), download the content and use its SHA256 prefix as the filename:
```
data/landing/padelnomics/{year}/{date}_{sha256[:8]}.csv.gz
```
Two runs that produce identical content produce the same hash → same filename → skip.
Files named by date or content. `write_gzip_atomic()` writes to a `.tmp` sibling
then renames — never leaves partial files on crash.
## State tracking
Every run writes one row to `data/landing/.state.sqlite`. Query it to answer operational questions:
Every run writes one row to `data/landing/.state.sqlite`:
```bash
# When did extraction last succeed?
sqlite3 data/landing/.state.sqlite \
"SELECT extractor, started_at, status, files_written, files_skipped, cursor_value
"SELECT extractor, started_at, status, files_written, cursor_value
FROM extraction_runs ORDER BY run_id DESC LIMIT 10"
# Did anything fail in the last 7 days?
sqlite3 data/landing/.state.sqlite \
"SELECT * FROM extraction_runs WHERE status = 'failed'
AND started_at > datetime('now', '-7 days')"
```
State table schema:
| Column | Type | Description |
|--------|------|-------------|
| `run_id` | INTEGER | Auto-increment primary key |
| `extractor` | TEXT | Extractor name (e.g. `padelnomics`) |
| `extractor` | TEXT | Extractor name (e.g. `overpass`, `eurostat`) |
| `started_at` | TEXT | ISO 8601 UTC timestamp |
| `finished_at` | TEXT | ISO 8601 UTC timestamp, NULL if still running |
| `finished_at` | TEXT | ISO 8601 UTC timestamp |
| `status` | TEXT | `running``success` or `failed` |
| `files_written` | INTEGER | New files written this run |
| `files_skipped` | INTEGER | Files already present (content unchanged) |
| `files_skipped` | INTEGER | Files already present |
| `bytes_written` | INTEGER | Compressed bytes written |
| `cursor_value` | TEXT | Last successful cursor (date, etag, page, etc.) |
| `error_message` | TEXT | Exception message if status = `failed` |
## Adding a new extractor
1. Add a function in `execute.py` following the same pattern as `extract_file_by_etag()` or `extract_file_by_hash()`
2. Call it from `extract_dataset()` with its own `extractor` name in `start_run()`
3. Store files under a new subdirectory: `landing_path(LANDING_DIR, "my_new_source", year)`
4. Add a new SQLMesh `raw/` model that reads from the new subdirectory glob
| `cursor_value` | TEXT | Resume cursor (date, index, etc.) |
| `error_message` | TEXT | Exception message if failed |
## Landing zone structure
```
data/landing/
├── .state.sqlite # extraction run history
── padelnomics/ # one subdirectory per source
└── {year}/
└── {month:02d}/
└── {etag}.csv.gz # immutable, content-addressed files
├── .state.sqlite
── overpass/{year}/{month}/courts.json.gz
├── eurostat/{year}/{month}/urb_cpop1.json.gz
├── eurostat/{year}/{month}/ilc_di03.json.gz
├── playtomic/{year}/{month}/tenants.json.gz
└── playtomic/{year}/{month}/availability_{date}.json.gz
```
## Data sources
| Source | Module | Schedule | Notes |
|--------|--------|----------|-------|
| Overpass API | `overpass.py` | Daily | OSM padel courts, ~5K nodes |
| Eurostat | `eurostat.py` | Daily (304 most runs) | urb_cpop1, ilc_di03 — etag dedup |
| Playtomic tenants | `playtomic_tenants.py` | Daily | ~8K venues, bounded pagination |
| Playtomic availability | `playtomic_availability.py` | Daily | Next-day slots, ~4.5h runtime |