Qlik REST In Sense (otto) - a RESTful Engine API wrapper

Inspired by project chartcacher by Alexander Karlsson.

The intention was to provide an easy access to apps, objects and its data on a Qlik Sense server thru a REST API to integrate with other systems.

We at TIQ Solutions do provide a support and maintenance subscription for this project and also customization. There is also a commercial version on it's way having more endpoints and authentication.

See also: routes.js

GET /v1/docs - all docs on server, returns getDocList().qDocList

GET /v1/doc/{docId} - app layout, returns getAppLayout().qLayout

GET /v1/doc/{docId}/objects - all objects of app, returns getAllInfos().qInfos

GET /v1/doc/{docId}/serialize - serialize an app into a JSON object, more details here: mindspank/serializeapp

GET /v1/doc/{docId}/object/{objId} - object layout, returns getLayout().qLayout

GET /v1/doc/{docId}/object/{objId}/layout - layout data, returns getLayout().qLayout depending on object type it contains qHyperCube and qDataPages

GET /v1/doc/{docId}/object/{objId}/data - object data, returns first data page of either getLayout().qHyperCube or qListObject depending on object type chart/listbox, not data for pivot tables

GET /v1/doc/{docId}/object/{objId}/pivotdata - object data, returns getLayout().qPivotDataPages for pivot tables

GET /v1/doc/{docId}/object/{objId}/layers - object data, returns getLayout().layers for maps

POST /v1/doc/{docId}/hypercube - give a column list array or HyperCubeDef JSON as payload (request body) and get back the evaluated getLayout().qHyperCube, without data page

Examples for payload to define a hypercube:

1. a list of columns as string, measures start with equal sign, all others are treated as dimensions (hint: use brackets for dimensions):

[
    "[Date.autoCalendar.Date]",
    "[Case Owner Group]",
    "=Avg([Case Duration Time])",
    "=Count({$<Status -={'Closed'} >} Distinct %CaseId )"
]

2. a list of NxDimension and NxMeasure structure objects, can be mixed with column strings like in 1.:

 [
    "[Date.autoCalendar.Date]",
    {"qDef": {"qFieldDefs": ["Case Owner Group"], "qFieldLabels": ["Group"]}},
    {"qDef": {"qDef": "=Avg([Case Duration Time])", "qLabel": "Avg Case Duration Time"}},
    {"qDef": {"qDef": "=Count({$<Status -={'Closed'} >} Distinct %CaseId )", "qLabel": "Open Cases"}}
]

3. a HyperCubeDef (see link below)

See Qlik help also:

• Columns
• NxDimension
• NxMeasure
• HyperCubeDef

POST /v1/doc/{docId}/hypercube/size - give a give a column list array or HyperCubeDef JSON as payload (request body) and get back the evaluated size of getLayout().qHyperCube as the following JSON object:

{
    "columns": 3,
    "rows": 89,
    "pages": 1
}

POST /v1/doc/{docId}/hypercube/json/{pageNo*} - give agive a column list array or HyperCubeDef JSON as payload (request body) and get back the evaluated getLayout().qHyperCube.qDataPages[0].qMatrix (the first data page or page number given as last URL param) transformed into a JSON collection of data rows (eg. easy to use with Qlik REST Connector), date and timestamps are delivered in UTC-time:

[
    {
        "Date": "2014-10-11T22:00:00.000Z",
        "Cumulative New Cases": 4,
        "Cumulative Closed Cases": 0
    },
    {
        "Date": "2014-10-18T22:00:00.000Z",
        "Cumulative New Cases": 5,
        "Cumulative Closed Cases": 3
    },
...
]

GET /wdc - Tableau WDC to connect q-risotto endpoint /v1/doc/{docId}/hypercube/json

cd sources

npm install

• Copy the files manually with admin priviliges into
  C:\Program Files\Qlik\Sense\ServiceDispatcher\Node\q-risotto\

• Then append the following configuration options to
  C:\Program Files\Qlik\Sense\ServiceDispatcher\services.conf
  This will let the Service Dispatcher know how to run the module, this step has to be re-applied in an upgrade of Qlik Sense Server.

[q-risotto]
Identity=Qlik.q-risotto
Enabled=true
DisplayName=q-risotto
ExecType=nodejs
ExePath=Node\node.exe
Script=Node\q-risotto\server.js

[q-risotto.parameters]

Adjust ./src/config/config.json to work with a Qlik Sense Server like this:

{
    "enigmaSchema": "enigma.js/schemas/12.34.11.json",
    "engineHost": "<your Qlik Sense Server hostname or IP>",
    "enginePort": 4747,
    "globalAppId": "engineData",
    "userDirectory": "Internal",
    "userId": "sa_repository",
    "certificatesPath": "C:/ProgramData/Qlik/Sense/Repository/Exported Certificates/.Local Certificates",
    "port": 3000
}

Adjust ./src/config/config.json to work with the dockerized QIX engine like this:

{
    "enigmaSchema": "enigma.js/schemas/12.34.11.json",
    "engineHost": "qix-engine",
    "enginePort": 9076,
    "globalAppId": "engineData",
    "userDirectory": null,
    "userId": null,
    "certificatesPath": null,
    "port": 3000
}

Use the docker-compose.yml file provided in the repo and start it:

docker-compose up

See config.json for configurations.

Start on Qlik Sense server with npm start or integrate into Qlik Sense ServiceDispatcher.

Navigate with browser or other tools to https://<qlik sense server name>:1338/<endpoint>

Test q-risotto API calls with Postman:

Use QlikView with the Qlik REST Connector to retrieve data from Qlik Sense:

Use q-risotto WDC with Tableau to retrieve data from Qlik Sense:

Use q-risotto in your R environment and retrieve data from Qlik Sense. Here you will find the R convenience package 'senser' to wrap q-risotto and complex Qlik structs like qHyperCubeDef:

https://github.com/ralfbecher/senser

Ralf Becher

• irregular.bi
• twitter/irregularbi
• github.com/ralfbecher

Copyright © 2016 Ralf Becher

Released under the MIT license.