pglite-tools
A selection of tools for working with PGlite databases, including pg_dump.
Install with:
npm install @electric-sql/pglite-tools
pgDump
pg_dump is a tool for dumping a PGlite database to a SQL file, this is a WASM build of pg_dump that can be used in a browser or other JavaScript environments. You can read more about pg_dump in the Postgres docs.
Note: pg_dump will execute DEALLOCATE ALL; after each dump. Since this is running on the same (single) connection, any prepared statements that you have made before running pg_dump will be affected.
Options
pg: A PGlite instance.
args: An array of arguments to pass to pg_dump - see pg_dump docs for more details.
fileName: The name of the file to write the dump to, defaults to dump.sql.
There are a number of arguments that are automatically added to the end of the command, these are:
--inserts - use inserts format for the output, this ensures that the dump can be restored by simply passing the output to pg.exec().
-j 1 - concurrency level, set to 1 as multithreading isn't supported.
-f /tmp/out.sql - the output file is always written to /tmp/out.sql in the virtual file system.
-U postgres - use the postgres user is hard coded.
Returns
- A
File object containing the dump.
Caveats
- After restoring a dump, you might want to set the same search path as the initial db.
Example
import { PGlite } from '@electric-sql/pglite'
import { pgDump } from '@electric-sql/pglite-tools/pg_dump'
const pg = await PGlite.create()
await pg.exec(`
CREATE TABLE test (
id SERIAL PRIMARY KEY,
name TEXT
);
`)
await pg.exec(`
INSERT INTO test (name) VALUES ('test');
`)
const initialSearchPath = (await pg1.query<{ search_path: string }>('SHOW SEARCH_PATH;')).rows[0].search_path
const dump = await pgDump({ pg })
const dumpContent = await dump.text()
const restoredPG = await PGlite.create()
await restoredPG.exec(dumpContent)
await restoredPG.exec(`SET search_path TO ${initialSearchPath};`);