First, create your properties file. JDBCJSON breaks properties down by jobs. Each job has four parameters: driver, url, sql, and out. driver is an optional parameter for specifying the driver class to be registered. driver is only required for non-JDBC4 drivers. JDBC4 drivers should self register. url is the JDBC connection URL (see your driver documentation for more details.) sql is the SQL statement executed whose result is used to generate the JSON. out is the path for resulting JSON. Take the following example:

    people.driver = net.sourceforge.jtds.jdbc.Driver
    people.url = jdbc:jtds:sqlserver://myserver/db;user=sa;password=sa
    people.sql = SELECT * FROM people
    people.out = people.json

This properties file will create a job called people that will query a SQL Server database on localhost, selecting all the columns from people, and output the result to people.json. Multiple jobs may be specified in a single properties file.

Next, run JDBCJSON. It's always a good idea to specify the debug switch the first time you run a new properties file against JDBC. The debug switch will give you additional warnings about things like unsupported fields (BLOBs, for example.)

    java -jar jdbcjson.jar -d people.properties

advanced usage

Only the url parameter of a properties file is required. The url and out properties have default values, based on the job name. The sql parameter defaults to SELECT * FROM <jobname>. The out parameter defaults to <jobname>.json. Thus:

    mytable.url = jdbc:postgresql://localhost/mydb?user=postgres&password=postgres

Is the same as:

    mytable.url = jdbc:postgresql://localhost/mydb?user=postgres&password=postgres
    mytable.sql = SELECT * FROM mytable
    mytable.out = mytable.json