multicorn

Overview

Version

Install

You can install multicorn directly. First, make sure the PGDG repository is added and enabled:

pig repo add pgdg -u          # Add PGDG repo and update cache

Install the extension using pig or apt/yum/dnf:

pig install multicorn;          # Install for current active PG version

pig ext install -y multicorn -v 18  # PG 18
pig ext install -y multicorn -v 17  # PG 17
pig ext install -y multicorn -v 16  # PG 16
pig ext install -y multicorn -v 15  # PG 15
pig ext install -y multicorn -v 14  # PG 14

dnf install -y multicorn2_18       # PG 18
dnf install -y multicorn2_17       # PG 17
dnf install -y multicorn2_16       # PG 16
dnf install -y multicorn2_15       # PG 15
dnf install -y multicorn2_14       # PG 14

apt install -y postgresql-18-multicorn   # PG 18
apt install -y postgresql-17-multicorn   # PG 17
apt install -y postgresql-16-multicorn   # PG 16
apt install -y postgresql-15-multicorn   # PG 15
apt install -y postgresql-14-multicorn   # PG 14

Create Extension:

CREATE EXTENSION multicorn;

Usage

Sources: README, CHANGELOG

Multicorn2 allows you to write Foreign Data Wrappers in Python. You implement a Python class that inherits from multicorn.ForeignDataWrapper, and Multicorn handles bridging it to PostgreSQL’s FDW interface. Version 3.2 is tested with PostgreSQL 14-18 and Python 3.9-3.13, though upstream recommends Python 3.10-3.12 for near-term distro compatibility.

Define a Python FDW Class

Create a Python module (e.g., myfdw.py) accessible to the PostgreSQL process:

from multicorn import ForeignDataWrapper

class MyFDW(ForeignDataWrapper):
    def __init__(self, options, columns):
        super().__init__(options, columns)
        self.options = options
        self.columns = columns

    def execute(self, quals, columns):
        """Yield rows as dictionaries. quals contains WHERE pushdown info."""
        yield {"id": 1, "name": "example"}

    def insert(self, new_values):
        """Handle INSERT operations."""
        pass

    def update(self, old_values, new_values):
        """Handle UPDATE operations."""
        pass

    def delete(self, old_values):
        """Handle DELETE operations."""
        pass

Create Server and Foreign Table

CREATE EXTENSION multicorn;

CREATE SERVER multicorn_srv FOREIGN DATA WRAPPER multicorn
  OPTIONS (wrapper 'myfdw.MyFDW');

CREATE FOREIGN TABLE my_table (
  id integer,
  name text
)
SERVER multicorn_srv
OPTIONS (
  option1 'value1'
);

SELECT * FROM my_table;

The wrapper option specifies the fully qualified Python class name. Any additional options are passed to the class constructor’s options parameter.

Built-in FDW Examples

Multicorn ships with several example FDW implementations that can be used directly or as reference:

• CsvFdw – read CSV files
• ProcessFdw – execute system commands and parse output
• GCalFdw – access Google Calendar
• ImapFdw – query IMAP mailboxes
• RssFdw – read RSS/Atom feeds

CREATE SERVER csv_srv FOREIGN DATA WRAPPER multicorn
  OPTIONS (wrapper 'multicorn.csvfdw.CsvFdw');

CREATE FOREIGN TABLE csvtest (
  col1 text,
  col2 text
)
SERVER csv_srv
OPTIONS (
  filename '/tmp/data.csv',
  skip_header '1',
  delimiter ','
);

Version Notes

Multicorn 3.2 adds basic OFFSET/LIMIT pushdown and LDAP paging support, and fixes LDAP right-parenthesis escaping. Upstream 3.1 added PostgreSQL 18 and Python 3.13 support while dropping PostgreSQL versions before 14.

Caveats

Multicorn2 and PL/Python are incompatible in the same PostgreSQL database on Python 3.12 due to CPython limitations. They can be installed on the same system, but avoid enabling both in one database.