csvquery

Query CSV files using SQL.

Features

Interactive REPL to quickly query CSV files with on demand file loading.
MySQL-compatible server to query CSV files, powered by go-mysql-server.
Insertions of new rows (using INSERT INTO).

For more info about what subset of SQL is supported, check out the documentation of go-mysql-server.

Install

You can install csvquery by downloading one of the prebuilt binaries available in the releases page.

Or install it manually:

go get github.com/erizocosmico/csvquery/...

Usage

There are two way you can use csvquery: REPL or server.

REPL

You can start the REPL with the following command:

csvquery repl

If you want to add some CSV files as tables directly when you're starting the command, you can do so with the -f flag.

csvquery repl -f some/file.csv -f another/file.csv:custom_table_name

As you can see in the previous example, you can choose a custom name for the table with the syntax FILE_PATH:TABLE_NAME. If no custom name is provided, the table name will be the file name without the extension and with all characters that are not alphanumeric or underscore removed.

If you don't want to load the files on startup time, don't worry, you can load them later using load.

Once the REPL has started, you can use the following commands:

load <FILE_PATH>[ <TABLE_NAME>] to load a new file into the database with the given table name (or the default one if it's not provided).
help to show the usage.
quit or exit to exit the REPL.
<SQL QUERY> to execute a SQL query.

Server

The server, unlike the REPL, needs all tables loaded at the start of the command, so you must provide all the CSV files that you want to query as flags using the -f flag. You can see an example of how to do that in the REPL section.

These are all available options for the server:

-d, --dbname=   Database name. (default: csv)
-f, --file=     Add file as a table. You can use the flag in the format '/path/to/file:NAME' to give the file a
                specific table name. Otherwise, the file name without extension will be the table name with only
                alphanumeric characters and underscores.
-u, --user=     User name to access the server. (default: root)
-p, --password= Password to access the server.
-P, --port=     Port in which the server will listen. (default: 3306)
-h, --host=     Host name of the server. (default: 127.0.0.1)

For example, let's start the server with the test data contained in the _testdata directory of this repository.

csvquery server -f _testdata/ratings.csv -f _testdata/superheroes.csv

Now, let's connect to the server using the mysql client tool and execute a query:

$ mysql --host=127.0.0.1 --port=3306 -u root
mysql> SELECT COUNT(*) as num_ratings, id
    -> FROM superheroes s
    -> INNER JOIN ratings r ON s.id = r.superhero_id
    -> GROUP BY id;
+-------------+------+
| num_ratings | id   |
+-------------+------+
|           1 | sm   |
|           2 | bm   |
|           2 | ww   |
|           2 | ma   |
+-------------+------+
4 rows in set (0.00 sec)

Roadmap

Read-only queries with REPL and server.
Support insertions.
Support updates and deletes.
Support creating indexes.

LICENSE

MIT License, see LICENSE.

GeorgeMcMullen/csvquery