Big changes

This commit is contained in:
Justin Visser 2026-02-21 17:33:05 +01:00
parent 4240ad38e2
commit e3f0d63fec
99 changed files with 8804 additions and 1731 deletions

View file

@ -0,0 +1,29 @@
# API Contract
## Envelope Invariant
All API responses under `/api/v1` use one of these envelopes:
- Success: `{"data": ..., "meta": {"request_id": "..."}}`
- Error: `{"error": {"code": "...", "message": "...", "details": ...}, "meta": {"request_id": "..."}}`
`meta.request_id` must be present for success and error responses.
## Publications Semantics
- `GET /api/v1/publications` supports `mode=all|unread|latest`.
- `mode=new` is currently accepted as a compatibility alias for `latest`.
- Pagination controls: `page`, `page_size` (with backward-compatible `limit`/`offset` support).
- Pagination fields in response: `page`, `page_size`, `has_prev`, `has_next`, `total_count`.
- `unread` represents read-state (`is_read=false`).
- `latest` represents discovery-state (`first seen in the latest completed run`).
Publication payloads expose:
- `pub_url` (canonical scholar detail URL)
- `doi` (normalized DOI)
- `pdf_url` (resolved OA PDF when available)
## Scholar Portability
- `GET /api/v1/scholars/export` exports tracked scholars and scholar-publication link state.
- `POST /api/v1/scholars/import` imports that payload while preserving global publication deduplication.

View file

@ -0,0 +1,68 @@
# Environment Reference and Audit
This document is the source-of-truth audit for what remains configurable versus internal defaults.
## Decision Summary
- Keep in `.env.example`: all runtime and compose controls currently exposed by `app/settings.py` plus compose/bootstrap extras.
- Move to internal defaults only: none at this time.
- Internal fallback behavior still exists in code to keep local/test startup resilient when values are omitted.
## Configurable Variables (By Section)
### Compose + Database
`POSTGRES_DB`, `POSTGRES_USER`, `POSTGRES_PASSWORD`, `DATABASE_URL`, `TEST_DATABASE_URL`, `SCHOLARR_IMAGE`
### App Runtime + Networking
`APP_NAME`, `APP_HOST`, `APP_PORT`, `APP_HOST_PORT`, `APP_RELOAD`, `MIGRATE_ON_START`, `FRONTEND_ENABLED`, `FRONTEND_DIST_DIR`
### Database Pool
`DATABASE_POOL_MODE`, `DATABASE_POOL_SIZE`, `DATABASE_POOL_MAX_OVERFLOW`, `DATABASE_POOL_TIMEOUT_SECONDS`
### Frontend Dev Overrides
`FRONTEND_HOST_PORT`, `CHOKIDAR_USEPOLLING`, `VITE_DEV_API_PROXY_TARGET`
### Auth + Session
`SESSION_SECRET_KEY`, `SESSION_COOKIE_SECURE`, `LOGIN_RATE_LIMIT_ATTEMPTS`, `LOGIN_RATE_LIMIT_WINDOW_SECONDS`
### HTTP Security Headers + CSP
`SECURITY_HEADERS_ENABLED`, `SECURITY_X_CONTENT_TYPE_OPTIONS`, `SECURITY_X_FRAME_OPTIONS`, `SECURITY_REFERRER_POLICY`, `SECURITY_PERMISSIONS_POLICY`, `SECURITY_CROSS_ORIGIN_OPENER_POLICY`, `SECURITY_CROSS_ORIGIN_RESOURCE_POLICY`, `SECURITY_CSP_ENABLED`, `SECURITY_CSP_POLICY`, `SECURITY_CSP_DOCS_POLICY`, `SECURITY_CSP_REPORT_ONLY`, `SECURITY_STRICT_TRANSPORT_SECURITY_ENABLED`, `SECURITY_STRICT_TRANSPORT_SECURITY_MAX_AGE`, `SECURITY_STRICT_TRANSPORT_SECURITY_INCLUDE_SUBDOMAINS`, `SECURITY_STRICT_TRANSPORT_SECURITY_PRELOAD`
### Logging
`LOG_LEVEL`, `LOG_FORMAT`, `LOG_REQUESTS`, `LOG_UVICORN_ACCESS`, `LOG_REQUEST_SKIP_PATHS`, `LOG_REDACT_FIELDS`
### Scheduler + Ingestion Safety
`SCHEDULER_ENABLED`, `SCHEDULER_TICK_SECONDS`, `SCHEDULER_QUEUE_BATCH_SIZE`, `INGESTION_AUTOMATION_ALLOWED`, `INGESTION_MANUAL_RUN_ALLOWED`, `INGESTION_MIN_RUN_INTERVAL_MINUTES`, `INGESTION_MIN_REQUEST_DELAY_SECONDS`, `INGESTION_NETWORK_ERROR_RETRIES`, `INGESTION_RETRY_BACKOFF_SECONDS`, `INGESTION_MAX_PAGES_PER_SCHOLAR`, `INGESTION_PAGE_SIZE`, `INGESTION_ALERT_BLOCKED_FAILURE_THRESHOLD`, `INGESTION_ALERT_NETWORK_FAILURE_THRESHOLD`, `INGESTION_ALERT_RETRY_SCHEDULED_THRESHOLD`, `INGESTION_SAFETY_COOLDOWN_BLOCKED_SECONDS`, `INGESTION_SAFETY_COOLDOWN_NETWORK_SECONDS`, `INGESTION_CONTINUATION_QUEUE_ENABLED`, `INGESTION_CONTINUATION_BASE_DELAY_SECONDS`, `INGESTION_CONTINUATION_MAX_DELAY_SECONDS`, `INGESTION_CONTINUATION_MAX_ATTEMPTS`
### Scholar Images + Name Search Safety
`SCHOLAR_IMAGE_UPLOAD_DIR`, `SCHOLAR_IMAGE_UPLOAD_MAX_BYTES`, `SCHOLAR_NAME_SEARCH_ENABLED`, `SCHOLAR_NAME_SEARCH_CACHE_TTL_SECONDS`, `SCHOLAR_NAME_SEARCH_BLOCKED_CACHE_TTL_SECONDS`, `SCHOLAR_NAME_SEARCH_CACHE_MAX_ENTRIES`, `SCHOLAR_NAME_SEARCH_MIN_INTERVAL_SECONDS`, `SCHOLAR_NAME_SEARCH_INTERVAL_JITTER_SECONDS`, `SCHOLAR_NAME_SEARCH_COOLDOWN_BLOCK_THRESHOLD`, `SCHOLAR_NAME_SEARCH_COOLDOWN_SECONDS`, `SCHOLAR_NAME_SEARCH_ALERT_RETRY_COUNT_THRESHOLD`, `SCHOLAR_NAME_SEARCH_ALERT_COOLDOWN_REJECTIONS_THRESHOLD`
### OA Enrichment + PDF Resolution
`UNPAYWALL_ENABLED`, `UNPAYWALL_EMAIL`, `UNPAYWALL_TIMEOUT_SECONDS`, `UNPAYWALL_MIN_INTERVAL_SECONDS`, `UNPAYWALL_MAX_ITEMS_PER_REQUEST`, `UNPAYWALL_RETRY_COOLDOWN_SECONDS`, `UNPAYWALL_PDF_DISCOVERY_ENABLED`, `UNPAYWALL_PDF_DISCOVERY_MAX_CANDIDATES`, `UNPAYWALL_PDF_DISCOVERY_MAX_HTML_BYTES`, `PDF_AUTO_RETRY_INTERVAL_SECONDS`, `PDF_AUTO_RETRY_FIRST_INTERVAL_SECONDS`, `PDF_AUTO_RETRY_MAX_ATTEMPTS`, `CROSSREF_ENABLED`, `CROSSREF_MAX_ROWS`, `CROSSREF_TIMEOUT_SECONDS`, `CROSSREF_MIN_INTERVAL_SECONDS`, `CROSSREF_MAX_LOOKUPS_PER_REQUEST`
### Startup Bootstrap + DB Wait
`BOOTSTRAP_ADMIN_ON_START`, `BOOTSTRAP_ADMIN_EMAIL`, `BOOTSTRAP_ADMIN_PASSWORD`, `BOOTSTRAP_ADMIN_FORCE_PASSWORD`, `DB_WAIT_TIMEOUT_SECONDS`, `DB_WAIT_INTERVAL_SECONDS`
## Drift Guard
CI enforces env parity with:
```bash
python3 scripts/check_env_contract.py
```
The check fails when:
- `app/settings.py` references a variable missing from `.env.example`.
- `.env.example` contains an unknown key (outside the approved compose/bootstrap extras).
- `.env.example` contains duplicate keys.