DBSheets is a simple in-memory query engine that allows you to retrieve filter, project and join data coming from MongoDB and PostgreSQL data sources.

In the `etc/config.yml' file you can configure multiple connections to databases that you will reuse within the "sheets".

example:

datasources:
    mongo1:
        url: mongodb://localhost
    postgres1:
        host: '127.0.0.1'
        database: 'foobar'
        user: foobar

In this example we define two databases, describe their connection details and assign them the keys mongo1 and postgres1.

The sheets are composed of two top items:

• variables: the sheet can be parametrized with variables that are evaluated at run time. The variables block will allow you to declare which variables the program will need before executing the sheet.
• steps: the steps of the sheet, describing the operations that need to be done

Example sheets can be found in the sheets/ directory. The steps can be subdivded between data sources and actions:

• mongodb: a MongoDB data source
• postgres: a PostgreSQL data source
• joined: a data source that joins two existing data sources
• public: a data source that takes an existing data source, perform some adjustments on field names, and exposes it

• comment: prints a comment in the terminal. For debug purposes
• setVar: sets a variable

The application can run in two modes:

• command line: the output will be presented in the shell and the program will exit once the operation is finished. This is the default mode.
• server: by issuing the -S flag, a web server will be activated on port 5000. The web server allows the execution of arbitrary sheets.

All options are available in the help screen of the application.

The simplest way to run DBSheets is using Docker.

docker run --rm -p 5000:5000 -v `pwd`/etc:/usr/src/app/etc -v `pwd`/sheets:/usr/src/app/sheets -ti theirish81/dbsheets [args]

where pwd, if you're not a *Nix user, is the current directory. [args] is a place holder for command line arguments. Leave it empty to see the help.

If you're a *Nix user, the script run.sh will do exactly this for you.