SQL Storage implementation for DataScript
Add this to deps.edn
:
io.github.tonsky/datascript-storage-sql {:mvn/version "2.0.0"}
Create storage by passing in javax.sql.DataSource
and :dbtype
option:
(def datasource (doto (org.sqlite.SQLiteDataSource.) (.setUrl "jdbc:sqlite:target/db.sqlite"))) (def storage (storage-sql/make datasource {:dbtype :sqlite}))
You can also pass optional options:
(storage-sql/make datasource {:dbtype :sqlite :batch-size 1000 :table "datascript" :ddl "create table if not exists datascript (addr INTEGER primary key, content TEXT)" :freeze-str pr-str :thaw-str clojure.edn/read-string})
Or use binary serialization:
(storage-sql/make datasource {:dbtype :sqlite :freeze-bytes (fn ^bytes [obj] (with-open [out (ByteArrayOutputStream.)] (t/write (t/writer out :msgpack) obj) (.toByteArray out))) :thaw-bytes (fn [^bytes b] (t/read (t/reader (ByteArrayInputStream. b) :msgpack)))})
After that, use it as any other storage:
(d/create-conn schema {:storage storage})
or
Currently supported :dbtype
-s:
:h2
:mysql
:postgresql
:sqlite
If your JDBC driver only provides you with DataSource
and you want to add some basic pooling on top, use storage-sql/pool
:
(def datasource
(doto (SQLiteDataSource.)
(.setUrl "jdbc:sqlite:target/db.sqlite")))
(def pooled-datasource
(storage-sql/pool datasource
{:max-conn 10
:max-idle-conn 4}))
(def storage
(storage-sql/make pooled-datasource
{:dbtype :sqlite}))
pool
takes non-pooled DataSource
and returns new DataSource
that pools connections for you.
If you used pool to create storage, you can close it this way:
(storage-sql/close storage)
Copyright © 2023 Nikita Prokopov
Licensed under MIT License.
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4