usql is a universal command-line interface for working with SQL databases.
usql provides a universal command-line interface for the following databases: PostgreSQL, MySQL, Oracle, SQLite, and Microsoft SQL Server.
The goal is to eventually have usql be a drop in replacement for PostgreSQL's psql command, with all the bells/whistles, but with the added benefit of working with more than one database.
Install in the usual Go way:
# install usql $ go get -u github.com/knq/usql # install with oracle support $ go get -u -tags oracle github.com/knq/usql# display command line arguments $ usql --help # connect to a postgres database $ usql pg://user:pass@localhost/dbname $ usql pgsql://user:pass@localhost/dbname $ usql postgres://user:pass@localhost:port/dbname # connect to a mysql database $ usql my://user:pass@localhost/dbname $ usql mysql://user:pass@localhost:port/dbname $ usql /var/run/mysqld/mysqld.sock # connect to a mssql (Microsoft SQL) database $ usql ms://user:pass@localhost/dbname $ usql mssql://user:pass@localhost:port/dbname # connect to a oracle database $ usql or://user:pass@localhost/dbname $ usql oracle://user:pass@localhost:port/dbname # connect to a sqlite file $ usql dbname.sqlite3 $ usql sq://path/to/dbname.sqlite3 $ usql sqlite3://path/to/dbname.sqlite3 $ usql file:/path/to/dbname.sqlite3The following is an example of connecting to xo's booktest example Oracle database, performing a query, and then connecting to the PostgreSQL, MySQL, Microsoft SQL Server, and SQLite3 databases and executing various queries.
A list of planned / in progress work:
- fix mysql "Error 1049:" messages, and other database error messages
- fix issue with readline on windows
- proper command processing (WIP)
- remaining
psqlcli parameters - the "j" commands (WIP)
- SQL completion (WIP)
- syntax highlighting (WIP)
- all \d* commands from
psql-- need to finish work extracting introspection code fromxo
Need to write scripts for packaging and build binaries for:
- Debian/Ubuntu (deb)
- MacOS X (.pkg)
- Windows (.msi)
- CentOS/RHEL (.rpm)
Additional:
- Submit upstream to Debian unstable (WIP)
- full test suite for databases, doing a minimal set of SELECT, INSERT, UPDATE, DELETE
Notes / thoughts / comments on adding support for various "databases":
- Google Spanner
- SAP HANA (cannot seem to run installer/get it running, support is already in usql, but not tested)
- CockroachDB (uses lib/pq wire protocol)
- ODBC (cross platform issues, similar to oracle with build support)
- Cassandra
- VoltDB
- MemSQL (uses mysql wire protocol)
- Atlassian JIRA JQL (why not? lol)