Go SQL driver for DuckDB

The DuckDB driver conforms to the built-in database/sql interface.

Notice on v1.4.0

Version 1.4.0 changed the DuckDB decimal representation from float64 to a new Decimal type, which is much more precise. If you are upgrading to v1.4.0 and are using DuckDBs decimals, please make sure to update your code to make use of the new Decimal type.

Installation

go get github.com/marcboeker/go-duckdb

go-duckdb uses CGO to make calls to DuckDB. You must build your binaries with CGO_ENABLED=1.

Usage

go-duckdb hooks into the database/sql interface provided by the Go stdlib. To open a connection, simply specify the driver type as duckdb:

db, err := sql.Open("duckdb", "")

This creates an in-memory instance of DuckDB. If you would like to store the data on the filesystem, you need to specify the path where to store the database:

db, err := sql.Open("duckdb", "/path/to/foo.db")

If you want to set specific config options for DuckDB, you can add them as query style parameters in the form of name=value to the DSN, like:

db, err := sql.Open("duckdb", "/path/to/foo.db?access_mode=read_only&threads=4")

Alternatively, you can also use sql.OpenDB when you want to perform some initialization before the connection is created and returned from the connection pool on call to db.Conn. Here's an example that installs and loads the JSON extension for each connection:

connector, err := duckdb.NewConnector("/path/to/foo.db?access_mode=read_only&threads=4", func(execer driver.Execer) error {
  bootQueries := []string{
    "INSTALL 'json'",
    "LOAD 'json'",
  }

  for _, qry := range bootQueries {
    _, err = execer.Exec(qry, nil)
    if err != nil {
      return err
    }
  }
  return nil
})
if err != nil {
  return nil, err
}

db := sql.OpenDB(connector)
db.SetMaxOpenConns(poolsize)
...

Please refer to the database/sql GoDoc for further usage instructions.

DuckDB Appender API

If you want to use the DuckDB Appender API, you can obtain a new Appender by supplying a DuckDB connection to NewAppenderFromConn().

connector, err := duckdb.NewConnector("test.db", nil)
if err != nil {
  ...
}
conn, err := connector.Connect(context.Background())
if err != nil {
  ...
}
defer conn.Close()

// Retrieve appender from connection (note that you have to create the table 'test' beforehand).
appender, err := NewAppenderFromConn(conn, "", "test")
if err != nil {
  ...
}
defer appender.Close()

err = appender.AppendRow(...)
if err != nil {
  ...
}

// Optional, if you want to access the appended rows immediately.
err = appender.Flush()
if err != nil {
  ...
}

DuckDB Apache Arrow Interface

If you want to use the DuckDB Arrow Interface, you can obtain a new Arrow by supplying a DuckDB connection to NewArrowFromConn().

connector, err := duckdb.NewConnector("", nil)
if err != nil {
  ...
}
conn, err := connector.Connect(context.Background())
if err != nil {
  ...
}
defer conn.Close()

// Retrieve Arrow from connection.
ar, err := duckdb.NewArrowFromConn(conn)
if err != nil {
  ...
}

rdr, err := ar.QueryContext(context.Background(), "SELECT * FROM generate_series(1, 10)")
if err != nil {
  ...
}
defer rdr.Release()

for rdr.Next() {
  // Process records.
}

Linking DuckDB

By default, go-duckdb statically links DuckDB into your binary. Statically linking DuckDB adds around 30 MB to your binary size. On Linux (Intel) and macOS (Intel and ARM), go-duckdb bundles pre-compiled static libraries for fast builds.

Alternatively, you can dynamically link DuckDB by passing -tags=duckdb_use_lib to go build. You must have a copy of libduckdb available on your system (.so on Linux or .dylib on macOS), which you can download from the DuckDB releases page. For example:

# On Linux
CGO_ENABLED=1 CGO_LDFLAGS="-L/path/to/libs" go build -tags=duckdb_use_lib main.go
LD_LIBRARY_PATH=/path/to/libs ./main

# On macOS
CGO_ENABLED=1 CGO_LDFLAGS="-L/path/to/libs" go build -tags=duckdb_use_lib main.go
DYLD_LIBRARY_PATH=/path/to/libs ./main

elefeint / go-duckdb