DuckDB.FSharp

Thin F#-friendly layer for DuckDB, inspired by Npgsql.FSharp.

Provides a functional, type-safe, and ergonomic API for working with DuckDB databases in F#, including connection management, parameterized queries, and flexible reader mappings.

Installation

dotnet add package DuckDB.FSharp

Quick Start

DuckDB connection strings are simple paths to a database file or :memory: for an in-memory database.

open DuckDB.FSharp

let connectionString = ":memory:"  // or "mydb.db" for a file-based database

connectionString
|> Sql.connect
|> Sql.query "SELECT 1 AS value"
|> Sql.execute (fun read -> read.int "value")

Usage Examples

Reading results as a list of records

type User = {
    Id: int
    Username: string
    Email: string option
}

let getUsers (connectionString: string) : List<User> =
    connectionString
    |> Sql.connect
    |> Sql.query "SELECT user_id, username, email FROM users"
    |> Sql.execute (fun read ->
        {
            Id = read.int "user_id"
            Username = read.text "username"
            Email = read.textOrNone "email"
        })

Asynchronous execution

let getUsersAsync (connectionString: string) : Async<List<User>> =
    connectionString
    |> Sql.connect
    |> Sql.query "SELECT user_id, username, email FROM users"
    |> Sql.executeAsync (fun read ->
        {
            Id = read.int "user_id"
            Username = read.text "username"
            Email = read.textOrNone "email"
        })

Parameterized queries

let getActiveUsers (connectionString: string) (isActive: bool) =
    connectionString
    |> Sql.connect
    |> Sql.query "SELECT * FROM users WHERE is_active = @isActive"
    |> Sql.parameters [ "@isActive", Sql.bool isActive ]
    |> Sql.execute (fun read ->
        {
            Id = read.int "user_id"
            Username = read.text "username"
            Email = read.textOrNone "email"
        })

Reading a single row (scalar values)

let getUserCount (connectionString: string) : int64 =
    connectionString
    |> Sql.connect
    |> Sql.query "SELECT COUNT(*) AS count FROM users"
    |> Sql.executeRow (fun read -> read.int64 "count")

Transactions (multiple statements)

connectionString
|> Sql.connect
|> Sql.executeTransaction [
    "INSERT INTO users (username, email) VALUES (@name1, @email1), (@name2, @email2)", [
        [ "@name1", Sql.text "alice" ; "@email1", Sql.text "alice@example.com" ]
        [ "@name2", Sql.text "bob"   ; "@email2", Sql.textOrNone None ]
    ]
    "UPDATE settings SET version = @version", [
        [ "@version", Sql.int 2 ]
    ]
]

Features

• Pure functional pipeline style (inspired by Npgsql.FSharp)
• Parameterized queries with type-safe parameter helpers
• Support for nullable columns via textOrNone, intOrNone, etc.
• Synchronous and asynchronous execution
• Transaction support
• Comprehensive unit tests covering connections, queries, parameters, readers, and edge cases

Status

Contributing

Contributions are welcome! Feel free to open issues or PRs on GitHub.

License

MIT License (see LICENSE file)

Support & Freelance

If DuckDB.FSharp is useful in your projects:

• ⭐ Star the repo to show support
• Consider sponsoring on GitHub ❤️

I'm also available for freelance F# work, especially data/analytics projects with DuckDB. Feel free to reach out on X: @typesmar