2021.2

OData API

The Thinkwise Indicium application tier offers an open API that can be used to perform CRUD-operations and execute procedures on the applications that have been made with the Thinkwise Platform. Through Indicium, a user will only have access to those tables, views and procedures for which the user has been authorized in IAM.

Indicium implements version 4.0 of the OData standard. The OData standard is well documented and as such this manual will not explain all of its features in detail. The OData documentation can be found using the link http://www.odata.org/documentation/.

Base URL

To access the API of a specific application, the URL of the request must always start with:

<web_app_root_url>/iam/<appl>/

In this template, <web_app_root_url> refers to the root URL of Indicium in IIS and <appl> refers to the ID or the alias of the application in question in IAM. Below, you will find two examples of valid base-URLs:

All requests that start with this base-URL will be subjected to authentication. The type of authentication used depends on the configuration of the user in question in IAM. All authentication follows the HTTP Basic authentication scheme, meaning that Indicium should only be exposed to the Internet over HTTPS.

Indicium running on a 2018.2 and earlier IAM must use <web_app_root_url>/api/<appl>/ instead.

Software Factory

The Indicium API provides access to project versions in the Software Factory development environment, as long as the Software Factory itself is available through the Indicium IAM.

<web_app_root_url>/sf/<sf_appl>/

The segment <sf_appl> refers to the application ID or application alias which can be found in the Software Factory at the desired runtime configuration. The runtime configuration must be activated for the user in the Software Factory to obtain access via Indicium.

For example:

Or, if you are using Universal GUI:

On the off chance that there are multiple Software Factory applications in IAM, the segment /sf/ can be substituted with the alias of another Software Factory application.

Indicium running on a 2018.2 and earlier IAM must use <web_app_root_url>/api/sf/_sf/<project_id>/<project_vrs_id>/<runtime_configuration_id> instead.

For example:

Standard operations

Indicium offers several standard operations which can be performed on every application.

The first standard operation is $metadata. By putting $metadata behind the base-URL of an application, the metadata document for that application can be retrieved. This metadata document contains an extensive description of all entities and operations available for that user in that application. It can be used to validate whether a user's permissions in IAM have been configured properly and as a guideline for writing your own Indicium client.

The second standard operation is /openapi. By putting /openapi behind the base-URL of an application, the API specification for that application can be retrieved in JSON format. This API specification can be used by a documentation generation tool like Swagger to display the API. Simply copy-paste the JSON content into the editor and a web-based UI will be provided with information about your API. It is also possible to import the specification into an API testing tool, like Postman. Postman will use the schemas defined in the API specification to create a new collection and automatically generate request and response bodies.

Performing CRUD-operations, functions and procedures

Aside from the standard operations, the Indicium API offers the ability to perform CRUD-operations on entities and to execute tasks, reports, functions and procedures on the database.

Note that every insert, update, task and report action is done by Indicium via a process called resource staging. For a more detailed explanation of this process and a description of the available endpoints and operations on each endpoint, see Resource Staging.

API actionHTTP VerbExample request
Selecting an entityGETRequest URL to retrieve all records from the project entity: https://server/indicium/iam/appl/project

Request URL to retrieve a specific record from the project entity given that it only has one numerical primary key column: https://server/indicium/iam/appl/project(123)

Request URL to retrieve a specific record from the sub_project entity which has a primary key of project_id and sub_project_id: https://server/indicium/iam/appl/sub_project(project_id=123,sub_project_id=321)

Request URL to retrieve the first 100 records from the project entity where the project_id is greater than 100, ordered by the project ID in ascending order: https://server/indicium/iam/appl/project?$top=100&$select=project_id, description&$filter=project_id gt 100&$orderby=project_id asc
Inserting an entityPOSTRequest URL to insert a new record into the project entity: https://server/indicium/iam/appl/project

Request body assuming that project_id is not an identity:
{"project_id": 123, "description": "a new project", "deadline": "2017-01-01T12:00:00.000"}

Only one record can be inserted per request. All query string parameters will be ignored.

See also Resource Staging.
Updating an entityPATCHRequest URL to update the description of an existing record of the project entity: https://server/indicium/iam/appl/project(123)

Request body:
{"description": "a changed project"}

Records can only be updated by specifying their entire primary key between the parentheses. All query string parameters will be ignored.

See also Resource Staging.
Deleting an entityDELETETo delete an existing record of the project entity: https://server/indicium/iam/appl/project(123)

Records can only be deleted by specifying their entire primary key between the parentheses. All query string parameters will be ignored.
Executing a taskPOSTRequest URL to call the task my_task with parameters project_id and description: https://server/indicium/iam/appl/my_task

Request body:
{"project_id": 123, "description": "my project"}

All query string parameters will be ignored.

See also Resource Staging.
Exporting a reportPOSTRequest URL to call the report my_report with parameters project_id and description: https://server/indicium/iam/appl/my_report

Request body:
{"project_id": 123, "description": "my project"}

All query string parameters will be ignored.

See also Resource Staging.
Calling a functionGETRequest URL to call the function my_function with parameters project_id and description: https://server/indicium/iam/appl/my_function(project_id=123,description='my project')

All query string parameters will be ignored.
Calling a procedurePOSTRequest URL to call the procedure my_procedure with parameters project_id and description: https://server/indicium/iam/appl/my_procedure

Request body:
{"project_id": 123, "description": "my project"}

All query string parameters will be ignored.

It is important to apply URL encoding (also known as Percent-encoding) to all value literals in the request URL. Value literals can only occur between the parentheses ( ) after the name of an entity and in the value of the $filter query string parameter.

Subroutines

For security reasons, subroutines are not exposed by the Indicium OData API by default. To expose a subroutine using Indicium, enable the API option of the subroutine in the Software Factory. The service name of a subroutine and its parameters can be changed using the Alias settings.

Supported OData operations

Indicium supports the following OData operations and query string options in the URL:

  • $metadata
  • $top
  • $skip
  • $select
  • $filter
  • $orderby
  • $expand
  • $apply

Indicium supports filtering on and ordering by the display value of a lookup as well. This can be achieved by specifying reference_id/display_column_id in the request URL, rather than simply the ID of the column. For example: http://server/indicium/iam/123/table?$filter=contains(reference_id/display_column_id,'value')

In addition to the standard OData operations, Indicium also supports a custom $prefilter operation which can be used to add prefilters to requests in order to select records from entities. For example: http://server/indicium/iam/123/table?$prefilter=prefilter_1,prefilter_2

Expand

The $expand feature allows you to select any properties from the entity on the other side of the navigation property. The navigation properties for lookups will always lead to the entity which contains the display property of the lookup, which is not necessarily the same entity as the source_tab of the corresponding lookup reference. This is also why the navigation properties for lookups have been renamed. Using the reference ID's is misleading, because they imply that the source_tab is on the other end of the navigation property.

Examples:

When using $expand, you can use all of the supported operations such as $select, $filter, $orderby, $prefilter on the target entity, but for now only $select is supported on the source entity in the $expand clause.

The response of an $expand request will look like this:

{
   "value":
   [
      {
         "project_id":15,
         "sub_project_id":25,
         "name":"Board of supervisors",
         "hourly_rate":135.00,
         "planned_start_date":"2006-08-18T00:00:00Z",
         "planned_end_date":"2007-08-19T00:00:00Z",
         "hours_budgeted":337.00,
         "actual_start_date":"2009-11-18T00:00:00Z",
         "finished":1,
         "finished_on_date":"2009-09-24T00:00:00Z",
         "finished_text":1,
         "hours_booked":806.25,
         "sub_project_name":"Board of supervisors",
         "start_date_and_end_date":"2006-08-18 | 2007-08-19",
         "transl_project_id":
         {
            "description":"Hydraulics"
         }
      }
   ]
}

Apply with Groupby

You can also use the $apply=groupby clause to select lookup values, but this method should typically only be used when selecting properties that are part of the lookup reference or the display property, otherwise you're better off using $expand. All columns specified in the $apply=groupby clause will in fact be added to a GROUP BY clause in the resulting SQL statement.

Example:

The response of an $apply=groupby request will look the same as the response of an $expand request, depending on the properties you select or group by:

{
   "value":
   [
      {
         "project_id":15,
         "transl_project_id":
         {
            "description":"Hydraulics"
         }
      }
   ]
}

It's also possible to apply an aggregate function to the display property, when a key can result in multiple display values or if you want to provide an alias to the display value.

Example:

<http://server/indicium/api/123/sub_project?$apply=groupby((project_id),aggregate(transl_project_id/description with min as display_field)>

Apply with standard computations

Standard computations are a great way to extract data from a column without the need to store them in a different column with a default expression. For example, if the computation Year is called, the computation will return the year of the supplied dates. Or, if you want to get a part of a column value, you can use Substring to retrieve it.

Indicium supports the following standard computations:

ComputationInputOutputExample
DateDate, DateTimeDate$apply=Compute(Date(columnname) as Alias))
TimeTime, DateTimeTime$apply=Compute(Time(columnname) as Alias))
IsoweekDate, DateTimeInt$apply=Compute(Isoweek(columnname) as Alias))
WeekdayDate, DateTimeInt$apply=Compute(Weekday(columnname) as Alias))
YearDate, DateTimeInt$apply=Compute(Year(columnname) as Alias))
MonthDate, DateTimeInt$apply=Compute(Month(columnname) as Alias))
DayDate, DateTimeInt$apply=Compute(Day(columnname) as Alias))
HourTime, DateTimeInt$apply=Compute(Hour(columnname) as Alias))
MinuteTime, DateTimeInt$apply=Compute(Minute(columnname) as Alias))
SecondTime, DateTimeInt$apply=Compute(Second(columnname) as Alias))
CeilingInt, NumericDecimalInt$apply=Compute(Ceiling(columnname) as Alias))
SubstringVarchar, NvarcharCharNcharVarchar$apply=Compute(Substring(columnname, Startindex, Length) as Alias))
ConcatVarchar, NvarcharCharNcharVarchar$apply=Compute(Substring(FirstColumn, SecondColumn) as Alias))

If $compute is used in the URL, the alias must be included in $select in order to be selected. The alias is also available in $filter, $orderby and $groupby.

Example:

compute(year(column) as alias)

Becomes:

select datepart(year, column) as alias

Extended Api

In addition to the standard OData Api Indicium offers several extended operations.

Query ParameterDescription
$deselectA comma separated list of column id’s which are not requested, this can help reduce the response size when for example binary data is stored in columns. i.e. $deselect=column_one, column_two
$prefilterCan be used to add prefilters to requests in order to select records from entities
← InstallationProcess flows API →