Web: replacing the jeep-sqlite engine with @sqlite.org/sqlite-wasm + OPFS (complete implementation with tests and a data migration; offering it upstream)
Nadie ha tomado este issue todavía.
Evaluación
- Dificultad
- 5/5
- Tiempo estimado
- Más de una semana
- Aptitud para principiantes
- 25/100
- Tipo de issue
- Nueva funcionalidad
- Claridad
- Bastante claro
- Estado de actividad
- Estancado
- Stack tecnológico
- sqlite, typescript, wasm
Línea de trabajo
Comienza con la rama feat/web-sqlite-wasm y revisa la implementación web junto con definitions.ts. Ejecuta las 259 pruebas de navegador de Vitest en Chromium real para ambos niveles de persistencia y, después, inspecciona la migración y la integración existente con CI. Se considera terminado cuando haya paridad de métodos y eventos, seguridad de la migración, versiones mínimas de navegador documentadas y pruebas superadas sin cambios en los wrappers nativos.
Escrito por el modelo de indexación a partir del texto del issue.
Descripción
Summary. The web implementation forwards every call to jeep-sqlite (sql.js in-memory, whole-database images in a localforage IndexedDB store). jeep-sqlite's last publish is 2.8.0 (August 2024), the ^2.7.2 pin here has not moved since June 2024, and the web layer has no test coverage in CI. The official @sqlite.org/sqlite-wasm now offers an OPFS VFS (opfs-sahpool) that needs no COOP/COEP headers and no SharedArrayBuffer, which makes a proper, durable web engine possible for exactly the deployments this plugin serves.
We maintain a Capacitor reading app that depends on this plugin, and we have built and are running a complete replacement of the web implementation. We would like to contribute it.
What it is.
- All SQL runs in a dedicated worker on
@sqlite.org/sqlite-wasm. Databases are durable OPFS files viaopfs-sahpool: no COOP/COEP, no SharedArrayBuffer, works on ordinary hosting. - Automatic fallback tier for browsers without OPFS sync access handles: the same engine with whole-image IndexedDB persistence (jeep's durability model), so FTS5 and behavior are identical on both tiers. Floors are measured and documented, including the syntax floor below which the wasm bundle cannot parse at all.
- A one-time, transparent migration of existing users' data out of the jeep store: raw IndexedDB read,
PRAGMA integrity_checkon every database, at-most-once by design, never overwrites, tolerates jeep'sbackup-*keys and never-saved placeholders. Interrupted migrations cannot corrupt anything. - Full method parity for all 36 web-supported methods and all five web events, driven through the unmodified
SQLiteConnection/SQLiteDBConnectionwrappers. Read-only connections become real on web.definitions.tssignatures are untouched, so there is zero impact on iOS, Android or Electron. - 259 tests in vitest browser mode (real Chromium) parameterized over both tiers, wired into the existing CI build.
- The worker ships as a classic script, so Firefox 111-113 (OPFS but no module workers) still works. Docs are swept: no element to define, no extra packages, no
sql-wasm.wasmcopy step.
Defects found in the current lineage while porting (fixed on our branch; we are happy to file the pieces that affect other platforms as standalone issues): the soft-delete rewrite performs a REAL delete when the table name is double-quoted; an unbracketed WHERE lets AND bind over OR and re-marks already-deleted rows; execute() bypasses soft-delete entirely; foreign-key enforcement was never enabled on web (PRAGMA foreign_keys defaults OFF), so declared constraints were silently ignored; and the foreign-key cascade logic misses referencing tables and deeper levels (that one also affects Electron).
The one breaking change, clearly marked in the branch as fix(web)!: foreign keys are now enforced on web connections. Apps whose data violates their declared schema will begin receiving constraint errors. We flag it for your versioning call.
What we ask. Is there interest in a PR? The branch is kklem0/sqlite:feat/web-sqlite-wasm, a reviewed series of conventional commits that each build. We are happy to split or reshape it however review works best for you. In the meantime we are publishing the fork as sssf-capacitor-sqlite so our app can ship, with the explicit goal of retiring it into upstream.
- Lenguaje dominante
- Swift
- Estrellas
- 661
- Forks
- 158
- Métricas de merge de PR
- Sin PR fusionados en 30 d
Preparar el entorno
Primeros pasos
- Lee el issue completo y luego la guía de contribución del proyecto.
- Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
- Haz un fork del repositorio y trabaja en una rama.
- Abre un pull request que haga referencia al número del issue.
Más de capacitor-community/sqlite
-
Dificultad 3/5 1-2 días Aptitud para principiantes 72/100
capacitor-community/sqlite#700 · 1 comentario · 1 reacción ·
-
bug/fix needs: triage
Dificultad 3/5 1-2 días Aptitud para principiantes 68/100
capacitor-community/sqlite#690 ·
-
feature needs: triage
Dificultad 3/5 1-2 días Aptitud para principiantes 55/100
capacitor-community/sqlite#689 ·
-
bug/fix needs: triage
Dificultad 3/5 1-2 días Aptitud para principiantes 45/100
capacitor-community/sqlite#685 ·
-
bug/fix needs: triage
Dificultad 3/5 1-2 días Aptitud para principiantes 48/100
capacitor-community/sqlite#671 ·
Todos los issues de capacitor-community/sqlite
Issues similares
-
bug
Dificultad 2/5 1-3 horas Aptitud para principiantes 84/100
EtanHey/brainlayer#984 ·
Los mantenedores suelen responder en 1 día
-
enhancement
Dificultad 2/5 1-3 horas Aptitud para principiantes 88/100
hoiekim/turnip-ios#197 ·
Los mantenedores suelen responder en 1 día
-
clawsweeper:needs-maintainer-review clawsweeper:needs-product-decision clawsweeper:no-new-fix-pr clawsweeper:source-repro impact:other issue-rating: 🦞 diamond lobster P2
Dificultad 2/5 1-3 horas Aptitud para principiantes 84/100
steipete/CodexBar#4071 · 1 comentario · 1 reacción ·
Los mantenedores suelen responder en 1 día
-
Dificultad 2/5 1-3 horas Aptitud para principiantes 65/100
gingerbeardman/Katana#15 ·
-
Dificultad 1/5 Menos de una hora Aptitud para principiantes 92/100
Los mantenedores suelen responder en 1 día