sqlite-http

An HTTP interface for SQLite. Plays very nicely with Litestream for backup / replication.

Install

cargo install sqlite-http

> sqlite-http --help
An HTTP interface for SQLite.

Usage: sqlite-http [OPTIONS] --host <HOST> --db-path <DB_PATH>

Options:
      --host <HOST>
          The host to bind to
      --db-path <DB_PATH>
          The path to the database file
      --collect-metadata
          Enable metadata collection about incoming queries
      --disable-wal-mode
          Disable Write-Ahead-Logging mode. Enabled by default
      --disable-foreign-keys
          Don't enforce foreign key constraints. Enforced by default
      --load-extension <LOAD_EXTENSION>
          Load an extension from the provided path. This flag can be used multiple times in one invocation to load multiple extensions
  -h, --help
          Print help
  -V, --version
          Print version

Examples

Basic

sqlite-http --host 127.0.0.1:8080 --db-path ./example.db

Only receieve sqlite-http's logs

SQLITE_HTTP_LOG=off,sqlite_http=trace sqlite-http --host 127.0.0.1:8080 --db-path ./example.db

Replicate with litestream (existing config assumes a running local MinIO instance)

litestream replicate -config ./etc/litestream.yml

Restore with litestream (existing config assumes a running local MinIO instance)

litestream restore -config ./etc/litestream.yml -replica "S3 Backup" ./example.db

Usage

Send this

{
    "sql" "INSERT INTO my_table (str_col, bool_col, blob_col, real_col) VALUES (?, ?, ?, ?)",
    "args": ["column 1", true, "03", 4]
}

Get this

{
    "rows": []
}

Send this

{
    "sql" "SELECT * FROM my_table",
    "args": []
}

Get this

{
    rows: [
        ["column 1", true, "03", 4]
    ]
}

Collecting Metadata

Using the --collect-metadata flag creates a __metadata_query table that may be useful in debugging and performance monitoring. Its schema is as follows, and it can be queried as any other table.