Fetching contributors…
Cannot retrieve contributors at this time
64 lines (44 sloc) 2.08 KB

Returning data from Cog


If you are using ruby or python to write your commands check out cog-rb and pycog. They both provide tools to make outputting data a bit easier

Returning data from a command in Cog is pretty easy. The simplest method is to just print a string. In ruby, puts("Hello, World"). Then Cog will print, "Hello, World".

For more granular control there are a number of output markers you can specify to let Cog know what you want done with your output. The two most common are JSON and COG_TEMPLATE:.

  • JSON - If you precede your response with the string "JSON" and a newline, Cog will attempt to parse your response as json.

    For example printing this to stdout:


would result in Cog returning:

{ "foo": "bar" }


Other than log messages, no other output can accompany json. Make sure any text following the JSON marker is valid json.

  • COG_TEMPLATE: - This is used to specify a template to render. Note: Unlike JSON there is no newline.

    For example printing this to stdout:

COG_TEMPLATE: foo_template

would tell cog to render the 'foo_template' with the json '{"foo":"bar"}'.

Additionally there are several other output markers for logging to the relay console. Log messages have the prefix COGCMD_ followed by the log level. Available log levels are: DEBUG, INFO, WARN, ERR. For example, if we wanted to log an error to the console we would just print, COGCMD_ERR: And error has occured from our command.


Note that log messages may be interleaved with other output. So you could for example puts("some text\nCOGCMD_DEBUG: A debug statement\n some more text") and Cog will happily print "some text some more text" to the user while logging, "A debug statement".

For the full reference of output tags see :doc:`/references/command_output_tags`.