jlafer-gooddata-cli

This package provides a CLI -- gdutil -- that I find useful when working with the GoodData API to run reports.

Installation

npm install

To add gdutil to your system path, you can optionally run npm link in the project folder. Note: it will likely require root or sudo access as it changes the system path.

Edit src/cfg.js to set BASE_GD_URL to the URL of your GoodData API. This will either be https://secure.gooddata.com/ for the standard GoodData cloud service or the white-label URL of your data provider (e.g., https://analytics.ytica.com/). Note that the trailing slash is required.

Command-line Interface

gdutil list
gdutil get

The commands above are for querying and displaying GoodData metadata. The data returned are needed for executing the gdutil report command.

Please run gdutil --help at the command-line to see commands and arguments.

The recommended usage is the following sequence of commands:

gdutil list -u <user> -p <pswd> -w <wrkspc_id> -t report

to list all reports and their ID; note the ID of the report you want to run; it will be used in the next command

gdutil get -u <user> -p <pswd> -w <wrkspc_id> --type report --object <obj_id>

to get the report definitions for your report; normally you want the latest definition, which is normally listed first and has the highest ID number

gdutil get -u <user> -p <pswd> -w <wrkspc_id> --type rptdefn --object <obj_id>

to get the report definition and list of attributeDisplayForms; the latter are uses of attributes and you will need any of those that you wish to use in dynamic filters when the report is executed; for list-type filters, you will also need to note the attribute ID (printed at the end of each line) and use it in the next step

gdutil get -u <user> -p <pswd> -w <wrkspc_id> --type attribForm --object <obj_id>

(optional) to get the list of values that can be used in a list filter; run this for each attributeDisplayForm upon which you will have a list filter; note the element ID(s) of any values (shown in brackets) that will be in the list of values

Armed with the required IDs noted above, you can now create a command-line to run the gdutil report command.

NOTE: the GoodData gray pages can also get this data, although often not as conveniently.

gdutil report

The report command is for executing and exporting a report, with support for dynamic filters. The resulting .csv file is placed in the /tmp folder.

All of the command-line arguments are straightforward, with the exception of the --filters (-f) argument. It takes a JSON object with the following format:

[{"type":<type>,"filterId":<id>,"attributeId":<id>,"values":[<id>,...]}]

The parameter is an array of filters. Each filter has the following properties:

• type is one of: list, floating, or interval
• filterId is the attributeDisplyForm ID
• attributeId is the attribute ID; only used with type=list
• values is an array of values or, when type=list, element IDs; normal JSON rules apply

A list filter on three elements (i.e., values) might look like this:

{"type":"list","filterId":<id>,"attributeId":<id>,"values":[id1, id2, id3]}

An interval filter on two dates might look like this:

{"type":"interval","filterId":<id>,"values":["2019-04-01","2019-04-30"]}

A floating filter on the last 7 days would look like this:

{"type":"floating","filterId":<id>,"values":[-7,0]}

For other date filter formats, consult the GoodData help pages on filtering.