Embed an SQLite database in your PostgreSQL table. AKA multitenancy has been solved.

Run a PostgreSQL 17 database that has pglite-fusion already installed.

docker run --network=host frectonz/pglite-fusion

pglite-fusion is also distributed with other PostgreSQL versions.

docker run --network=host frectonz/pglite-fusion:pg12

docker run --network=host frectonz/pglite-fusion:pg13

docker run --network=host frectonz/pglite-fusion:pg14

docker run --network=host frectonz/pglite-fusion:pg15

docker run --network=host frectonz/pglite-fusion:pg16

docker run --network=host frectonz/pglite-fusion:pg17

Connect to the PostgreSQL database using psql.

psql postgresql://postgres@localhost:5432/

Here's some demo usage.

-- Load PG extension
CREATE EXTENSION pglite_fusion;

-- Create a table with an SQLite column
CREATE TABLE people (
    name     TEXT NOT NULL,
    database SQLITE DEFAULT init_sqlite('CREATE TABLE todos (task TEXT)')
);

-- Insert a row into the people table
INSERT INTO people VALUES ('frectonz');

-- Create a todo for "frectonz"
UPDATE people
SET database = execute_sqlite(
    database,
    'INSERT INTO todos VALUES (''solve multitenancy'')'
)
WHERE name = 'frectonz';

-- Create a todo for "frectonz"
UPDATE people
SET database = execute_sqlite(
    database,
    'INSERT INTO todos VALUES (''buy milk'')'
)
WHERE name = 'frectonz';

-- Fetch frectonz's info
SELECT 
    name, 
    (
        SELECT json_agg(get_sqlite_text(sqlite_row, 0))
        FROM query_sqlite(
            database, 
            'SELECT * FROM todos'
        )
    ) AS todos
FROM 
    people 
WHERE 
    name = 'frectonz';

git clone git@github.com:frectonz/pglite-fusion.git 
cd pglite-fusion
nix develop
cargo pgrx init --pg13 download # this will take some time, since you are compiling postgres from source
cargo pgrx run # this will drop you into a psql repl, you can then follow the example shown above

Creates an empty SQLite database and returns it as a binary object. This can be used to initialize an empty SQLite database in a PostgreSQL column.

SELECT empty_sqlite();

Executes a SQL query on a SQLite database stored as a binary object and returns the result as a table of JSON-encoded rows. This function is useful for querying SQLite databases stored in PostgreSQL columns.

• sqlite: The SQLite database to query, stored as a binary object.
• query: The SQL query string to execute on the SQLite database.

SELECT * FROM query_sqlite(
    database, 
    'SELECT * FROM todos'
);

Executes a SQL statement (such as INSERT, UPDATE, or DELETE) on a SQLite database stored as a binary object. The updated SQLite database is returned as a binary object, allowing further operations on it.

• sqlite: The SQLite database to execute the SQL query on, stored as a binary object.
• query: The SQL statement to execute on the SQLite database.

UPDATE people
SET database = execute_sqlite(
    database,
    'INSERT INTO todos VALUES (''solve multitenancy'')'
)
WHERE name = 'frectonz';

Creates an SQLite database with an initialization query already applied on it. This can be used to initialize a SQLite database with the expected tables already created.

• query: The SQL statement to execute on the SQLite database.

CREATE TABLE people (
    name     TEXT NOT NULL,
    database SQLITE DEFAULT init_sqlite('CREATE TABLE todos (task TEXT)')
);

Extracts a text value from a specific column in a row returned by query_sqlite. Use this function to retrieve text values from query results.

• sqlite_row: A row from the results of query_sqlite.
• index: The index of the column to extract from the row.

SELECT get_sqlite_text(sqlite_row, 0) 
FROM query_sqlite(database, 'SELECT * FROM todos');

Extracts an integer value from a specific column in a row returned by query_sqlite. Use this function to retrieve integer values from query results.

• sqlite_row: A row from the results of query_sqlite.
• index: The index of the column to extract from the row.

SELECT get_sqlite_integer(sqlite_row, 1) 
FROM query_sqlite(database, 'SELECT * FROM todos');

Extracts a real (floating-point) value from a specific column in a row returned by query_sqlite. Use this function to retrieve real number values from query results.

• sqlite_row: A row from the results of query_sqlite.
• index: The index of the column to extract from the row.

SELECT get_sqlite_real(sqlite_row, 2) 
FROM query_sqlite(database, 'SELECT * FROM todos');

Checks whether a given SQLite database is valid. This function attempts to open the SQLite database and returns true if it is a valid database.

• sqlite: A SQLite database stored as a binary blob.

SELECT is_valid_sqlite(database);

Lists all table names present in the SQLite database.

• sqlite: A SQLite database stored as a binary blob.

SELECT table_name FROM list_sqlite_tables(database);

Retrieves the SQL schema of each table in the SQLite database, the CREATE TABLE statements for each table in the SQLite database.

• sqlite: A SQLite database stored as a binary blob.

SELECT schema_sql FROM sqlite_schema(database);

Loads a SQLite database from a file into PostgreSQL.

• path: The file path of the SQLite database to load.

SELECT import_sqlite_from_file('/tmp/mydb.sqlite');

Saves the SQLite database to a file on disk.

• sqlite: The SQLite database.
• path: The file path where the database should be saved.

SELECT export_sqlite_to_file(database, '/tmp/mydb.sqlite');

Runs the SQLite VACUUM command to optimize storage by reclaiming unused space.

• sqlite: The SQLite database.

SELECT vacuum_sqlite(database);

Counts the number of rows in a specified table within the SQLite database.

• sqlite: The SQLite database.
• table: The name of the table to count rows from.

SELECT count_sqlite_rows(database, 'users');