go build -ldflags="-s -w" -buildmode=c-shared -o postgresql.so

SQLite extension to replicate data from PostgreSQL

Installation

Download postgresql extension from the releases page. Here’s a great article that explains how to install the SQLite extension.

Compiling from source

• Go 1.24+ and CGO_ENABLED=1 is required.

go build -ldflags="-s -w" -buildmode=c-shared -o postgresql.so

• Use .so extension for Linux, .dylib for MacOS and .dll for Windows

Basic usage

Prepare PostgreSQL

1. Modify postgresql.conf:

• Set wal_level = logical.
• Adjust max_replication_slots and max_wal_senders according to the number of subscribers and replication slots needed.

2. Modify pg_hba.conf:

Add an entry to allow the replication user to connect from the subscriber’s IP address. For example:

host    replication     rep_user        subscriber_ip/32        md5

3. Create a Replication User:

• Create a user with replication privileges:

CREATE ROLE rep_user WITH REPLICATION LOGIN PASSWORD 'secret';

4. Create a Publication:

• Define which tables or all tables in a database should be replicated:

CREATE PUBLICATION my_publication FOR TABLE table1, table2;
-- or for all tables in the current database:
CREATE PUBLICATION my_publication FOR ALL TABLES;

5. Restart PostgreSQL.

Prepare SQLite

1. Convert PostgreSQL databse to SQLite

go install github.com/litesql/postgresql/cmd/pg2sqlite@latest

pg2sqlite [postgresql_url] example.db

2. Loading the extension

sqlite3 example.db

# Load the extension
.load ./postgresql

# check version (optional)
SELECT pg_info();

3. Start replication to sqlite

• Create a slot (if necessary)

SELECT pg_create_slot('postgres://rep_user:secret@127.0.0.1:5432/postgres', 'my_slot');

• Insert data into pg_sub virtual table to start replication.

INSERT INTO temp.pg_sub(connect, slot, publication) VALUES('postgres://rep_user:secret@127.0.0.1:5432/postgres', 'my_slot', 'my_publication');

Configuring

You can configure replication by passing parameters to the VIRTUAL TABLE.

Param	Description	Default
use_namespace	Keep schema/namespace (otherwise always use main database)	false
position_tracker_table	Table to store replication position checkpoints	pg_pub_stat
timeout	Timeout in milliseconds	10000
logger	Log errors to “stdout, stderr or file:/path/to/log.txt”