Data Objects

This topic applies to the Admin Console > Data > Objectssettings.

Data Objects are the tables, views, methods, stored procedures, functions and custom SQL made accessible from the Data Sources for reports.

Create, Edit, Delete Data Objects

• To add a new data object click Objects in the Main Menu and either:
  • click the Add icon at the top of the main menu
  • right-click and select Add from the context menu
  • or to quickly add data objects, use the Automatic Database Discoverytools
• To clone an object, right-click and select Clone from the context menu
• To edit an object either:
  • double click it
  • select it and click the Edit icon at the top of the main menu
  • right-click it and select Edit from the context menu
• To delete an object either:
  • select it and click the Delete icon at the top of the main menu
  • right-click it and select Delete from the context menu
• To save changes and new objects click the Apply or Okay buttons.

Data Object Properties

A data object has the following properties:

Name

Select the data source from the first dropdown. In the second drop-down select a data object. This field is required.

This will display all the of the source’s tables, views, methods, stored procedures, and functions.

To add custom SQL click the Add Custom SQL icon next to the Data Sources dropdown. See Custom SQL Objects.

The name of data objects may not contain the following characters:

{ } [ ] , . %

To create a data object for a database table (or other backing model) with one of these characters in its name, create a Custom SQL Objects with a simple SELECT clause. For example for a table named sales.figures.2020:

SELECT * FROM [sales.figures.2020]

Note that the identifier delimiters (the [ ]) around the table name will vary with the database type.

Alias

The user friendly name for the data object that gets displayed to end-users. This field is required.

An alias may not contain the following characters:

@ { } [ ] , . %

Unique Key Fields

The columns that uniquely identify a record. This field is required.

The folder used to group related data objects. A hierarchy can be created by using backslashes between names.

Northwind

Northwind\Employees

When configuring Multiple Data Models, separate the different folder names with a comma.

ID

An application-wide unique identifier for the data object. IDs are required when creating multiple data objects with that have the same name but come from distinct data sources. IDs can also be used to optimize Web Service and .NET Assembly calls. For more information see Data Object IDs.

By default all objects have an implicit object ID that is set behind-the-scenes. This field allows you to set the IDs explicitly. It is generally recommended to set explicit IDs for all data objects. As of v2019.1.12, this field will display the implicit ID by default.

Parameters

Parameters that are passed to stored procedures, table functions, Web Services or .NET Assembly methods. Clicking in the drop-down will bring up a menu. Click the Add button and select the parameter from the drop-down list. See Stored Procedures and Assembly Data Sources

Parameter values are passed in the order in which they are listed in the data object. It is critical to ensure that the order is correct.

Tenant Columns

Specify which columns contain tenant information and link the parameters accordingly.

This setting is used to filter data when multiple users’ information is held within the same table or view, and a column holds information identifying each user. Exago will only retrieve the rows where the column value matches the corresponding parameter.

Description

Provide an optional user-friendly description of this data object. When this object is hovered over in the Data Fields Pane, the description text will pop up in a tooltip.

Column Metadata

Specify any columns that should not be filterable, visible, or that should be read as a specific data type. See Column Metadata.

Schema Access Type

Specify how Exago should retrieve the schema for the data object. There are three possibilities:

• Default — Follow the global Schema Access Type setting in Other Settings.
• Datasource — Queries the data source for the schema.
• Metadata — Reads the schema from the stored metadata.

See Retrieving Data Object Schema.

Filter Dropdown Object

Specify an alternative data object or custom SQL to be queried when a user clicks a filter value dropdown (for example in the Report Designer’s Filters dialog, the Filters pane of the Dashboard or ExpressView Designer)

The Filter Dropdown Object must have a column with the same name as each column in the main data objects or an error message will be displayed to the user when clicking the dropdown.

To set a Filter Dropdown Object, choose a Data Source and then:

• select a table or view from the dropdown on the right, or
• click the Custom SQL icon to write custom SQL statements to retrieve filter values. Writing custom SQL for Filter Dropdown Objects is the same as writing Custom SQL Objects.

The filter dropdown object data source should return values faster than it’s corresponding “parent” data object. Therefore, .NET Assembly data sources are not recommended as Filter Dropdown Object data sources.

To remove a Filter Dropdown object, click the Deleteicon.

For more information, see Filter Dropdown Objects .

Filter Dropdown Objects

The Filter Dropdown Object overrides the source of values that populate a filter value dropdown in the user interface (for example in a prompting filter dialog; in the Filters dialog in the Report, Dashboard or ExpressView Designer; in the Dashboard or Report Viewer).

By default, filter value dropdowns populate with data from their corresponding data object. Explicitly defining a Filter Dropdown Object in the configuration allows those default values to be replaced by another table, view or by custom SQL statements. This substitution happens transparently to the end user.

Specifying a Filter Dropdown Object may be most useful when the corresponding data object is a Stored Procedure, Web Service, or .NET Assembly that takes more than a few seconds to return data, as this can adversely affect user experience. Filter Dropdown Objects may also be useful when explicitly controlling filter values available to end users is desired.

Using a Filter Dropdown Object requires General > Filter Settings > Read Database for Filter Values to be True.

Filter Dropdown Objects may reference Data Sources other than their corresponding data object but there must be an identical column in the source data object and the Filter Dropdown Object.

For example, consider the Orders and OrderDetails tables of the Northwind sample data set. If the Orders table is used as the Filter Dropdown Object for the OrderDetails table, and a filter is created on the OrderId field, the Filter Dropdown Object will work since both tables contain an integer typeOrderId field.

Although not obvious in the figure, the order ID numbers (for example 10248, 10249, 10250) in the ExpressView filter dropdown below are read from the Orders table since it is the Filter Dropdown Object for OrderDetails.

On the contrary, there is no Quantity column in the Orders table. When attempting to filter on OrderDetails.Quantity, an error message will be displayed to the end user.

To prevent users from filtering on fields that are not available in the Filter Dropdown Object, set the Filterable property in the Filter Dropdown Object’s Column Metadata to False. Note that this will prevent all filtering on that field through the user interface. Filters on the column may still be added via one of the APIs.

Filter Dropdown Objects will completely replace the available filter dropdown values for their corresponding data object. If the Filter Dropdown Object has less or more values than its corresponding data object, users may not be able to filter on all possible values or be able to select values that do not correlate with records in the corresponding object.

The figure below demonstrates a less than ideal Filter Dropdown Object—the object only contains three Order IDs despite there being many more in the OrderDetails table, as well as an Order ID that does not exist in the OrderDetails table (for example 9999999). A user would only be able to filter by order ID 10248 or 10249.

Additional Resources

• Filters, segment on Filter Dropdown objects runs from 14:05–20:15.

Stored Procedures

Stored Procedures offer the ability to use high level code to modify the data set before it is sent to Exago.

For best performance, stored procedures should know what sorts and filters the user has set and whether to return the schema, a single column, or the entire data set. To accomplish this:

• Use the Call Type, Filter, Column and Sort Parameters in the Programmable Object Settings. These parameters will be passed from Exago to identically named parameters in the Stored Procedure. It is highly recommended to use these parameters as they can improve performance.
• Additional parameters may be passed by setting them in the Data > Object > Parameters tab for the Data Object.

As noted above in the Parameters section, parameter values are passed in the order in which they are listed. It is critical to ensure that the order listed is correct.

Important Note for Microsoft SQL Server

SQL Server has an attribute called FMTONLY that must be handled by all stored procedures.

FMTONLY has two possible values:

• ON: The stored procedure will only return the column schema. However all IF conditional statements are ignored and all of the code will be executed. This setting will fail if the stored procedure contains any temp tables.
• OFF: The stored procedure returns all of the data and the column schema. The stored procedure will correctly execute IF conditions.

The ON setting will cause problems if there are IF conditions in the procedure; however, only using the OFF setting will hurt performance if the Call Type Parameter in the Programmable Object Settings is not used.

The following example demonstrates how to use the Call Type, Column, Filter and SortStr Parameters to improve efficiency.

For SQL Servers, FMTONLY is set to OFF.

CREATE PROCEDURE [dbo].[sp_webrpt_person]
@callType INT,
@columnStr varchar(1000),
@filterStr varchar(1000),
@fullFilterStr varchar(1000),
@sortStr varchar(1000)
AS
SET NOCOUNT ON /*for performance reasons */
SET FMTONLY OFF /*force procedure to return data and process IF conditions */
declare @sql varchar(2000)
declare @columnInfo varchar(1000)
declare @orderbyClause varchar(1000)
declare @operator varchar(max)
if @callType = 0 /*return schema; no need to return any rows */
begin
 set @sql = 'select * from vw_webrpt_person where 0 = 1'
end
else
if @callType = 1 /*return all data for execution */
begin
 set @orderbyClause = ''
 if @sortStr is not NULL AND @sortStr <> 'null' set @orderbyClause = ' ORDER BY ' + @sortStr
 set @sql = 'select ' + @columnStr + ' from vw_webrpt_person where ' + @filterStr + @orderbyClause
end
else
if @callType = 2 /*return filter dropdown values; limit # rows to some value; sortStr will indicate which direction user is scrolling the dropdown*/
begin
set @orderByClause = ''
 set @operator = ' >= '
 if @sortStr is not NULL AND @sortStr <> 'null' AND @sortStr <> '' Begin SET @orderByClause = ' ORDER BY ' + @sortStr END
 if (RIGHT(@sortStr, 4) = 'DESC')Begin SET @operator = ' <= 'END
 set @columnInfo = '[' + @columnStr + ']'
 SET @sql = 'select top 100 ' + @columnInfo + ' from Northwind.dbo.OrderDetails where ' + @columnInfo + @operator + @filterStr + ' AND ' + @fullFilterStr + @orderByClause
end
exec(@sql)

Table Value Functions

Table Value Functions can be used as data objects. Any available table value functions of a data source will be displayed in the Extensions tab under Functions. Exago handles table value functions similar to views and tables except it will pass any parameters set in the Data > Object tab or in the Programmable Object Settings.

See Table-Valued Functions.

Custom SQL Objects

Exago can use custom SQL as a data object.

1. Provide a unique Object Name for the new data object. It cannot match any existing data object in the system configuration nor contain any of the following characters, including spaces: [ ] { } . , @
2. Choose the Data Source that this Custom SQL Object will retrieve data from. Not all data sources in the system may support Custom SQL objects.
3. In the large code window enter the full SQL statement. This statement will be inserted as a subquery when it is sent to the Data Source for processing.
   

   

   ORDER BY clauses and Common Table Expressions are not supported in subqueries.

4. System variables called parameters, that change based on certain environmental situations (for example who is running the report) may optionally be added to the SQL statement by selecting the name from the dropdown and clicking the Add button.
5. Click the Test icon to check the SQL statement for validity.
6. Click Okay to close the Custom SQL Object dialog and continue editing the data object.

Data Object Macros

Macros can be embedded in Custom SQL Objects to make them even more dynamic. Each macro allows for different SQL to be used according to the circumstances in which the data object is being called. Below are the details and examples of available macros.

IfExecuteMode(string trueCondition, string falseCondition)

IfExistReportDataObject(string dataObjectName, string trueCondition, string falseCondition)

Data Object IDs

There are three ways in which you can utilize data object IDs.

Adding Multiple Data Objects with the Same Name

IDs distinguish Data Objects that have the same Name but come from different Data Sources. When adding multiple Data Objects with the same name, make sure each data object has a unique ID.

Avoiding Issues from Changes to Object Names

Providing IDs for all the Data Objects prevents issues if the name of the underlying table, view or stored procedure is changed.

Calling a Single Web Service/.NET Assembly/Stored Procedure

Web Services, .NET Assemblies and Stored Procedures comprise a group called Programmable Objects. These objects can retrieve parameters from Exago and the host application in order to control what data is exposed to the user.

Generally for Web Services and .NET Assemblies each data object calls a distinct method. Similarly each Stored Procedure is its own data object. By using data object IDs a single method/stored procedure can be called. This method can then return data or schema based on the data object ID.

To call a single Web Service/.NET Assembly/Stored Procedure:

1. Provide a name for Data Object ID Parameter Name in Programmable Object Settings.
2. Create a method/procedure in the service, assembly or procedure that utilizes the object ID parameter to return the appropriate data/schema.
3. For each data object:
   1. Select Object in the Main Menu and click the Add button
   2. Select the single Service/Assembly/Procedure
   3. Provide an Alias and an IDfor the object
   4. Select the Unique Key Fields.
   5. Click Okay or Apply to save the object.

Example: This stored procedure uses the object ID parameter (@objectID@) to return different data/schema information for different object IDs.

 ALTER PROCEDURE "dbo" . "Exago_example" @callType INT,
                                         @objectID NVARCHAR(max)
AS
    SET nocount ON
    SET fmtonly OFF

    IF @objectID = 'Produce'
      BEGIN
          IF @callType = 0
            BEGIN
                SELECT productid,
                       productname,
                       supplierid,
                       unitprice,
                       unitsinstock
                FROM   products
                WHERE  categoryid = 1001
            END
          ELSE IF @callType = 1
            BEGIN
                SELECT productid,
                       productname,
                       supplierid,
                       unitprice,
                       unitsinstock
                FROM   products
                ORDER  BY productid
            END
          ELSE IF @callType = 2
            BEGIN
                SELECT productid,
                       productname,
                       supplierid,
                       unitprice,
                       unitsinstock
                FROM   products
                ORDER  BY productid
            END
      END

    IF @objectID = 'Orders0'
      BEGIN
          IF @callType = 0
            BEGIN
                SELECT orderid,
                       orderdate,
                       requireddate,
                       shippeddate,
                       customerid
                FROM   orders
                WHERE  customerid = 0
            END
          ELSE IF @callType = 1
            BEGIN
                SELECT orderid,
                       orderdate,
                       requireddate,
                       shippeddate,
                       customerid
                FROM   orders
                ORDER  BY orderid
            END
          ELSE IF @callType = 2
            BEGIN
                SELECT orderid,
                       orderdate,
                       requireddate,
                       shippeddate,
                       customerid
                FROM   orders
                ORDER  BY orderid
            END
      END  

Cloning Data Objects v2019.1+

A of v2019.1+, data objects may be cloned via the Admin Console. A cloned data object inherits all its base attributes from its parent except its Alias and ID, which must be entered manually in order to create the clone.

To clone a data object, right-click it and select Clone.

Enter a new Alias and ID for the cloned object.

In v2019.2.18+, select how to handle the folder and description that will be associated with this cloned object:

 

• Inherit Folder — check this checkbox to copy the folder location of the object that this clone was created from. Uncheck to provide a different folder location in the text field below. When checked, the value copied is shown in the field below. The default value is checked.
• Inherit Description — check this checkbox to copy the description text of the object that this clone was created from. Uncheck to provide a unique description in the text field below. When checked, the value copied is shown in the field below. The default value is checked.

The cloned object will then appear in the Objects menu:

Functionality of Cloned Data Objects

• A cloned data object is linked to its parent object. Any changes made to a parent data object will be reflected in its clone.
• As of v2020.1.17+ and v2021.1.5+, cloned data objects may point to a different (but identical) object in the corresponding data source (for example a view, a table). This requires editing the XML configuration file. See the Data Objects in Config File XML Reference (All Nodes but General) for more information.
• Deleting the parent data object will also delete its linked clone.
• Clones do not inherit joins from their parent objects. Joins need to be manually created between clones and other data objects.
• Clones cannot be distinguished from other data objects within the report designers. There is no visual or functional difference between a cloned data object and a normal object outside of the Admin Console.
• A cloned data object cannot itself be cloned or duplicated.

Column Metadata

Column metadata refers to the properties of each column in the data object.

Column metadata overrides culture settings.

There are several ways to add column metadata to a data object:

• for all data objects in a data source in bulk utilizing the Building Metadata in Bulk
• with the Add All Complete Objects, Joins and Metadata v2021.2+ on the corresponding Data Source (v2021.2+)
• manually for each data object, as described below

To modify the metadata of a column, either:

• double-click it
• select it then click the Add button.

Click the Read Schema button to quickly create column metadata for each column in the data object.

To remove metadata for a column, select it in the right panel and click the Delete icon.

To save changes to Column Metadata, click the Okay button. To discard the changes, click the Cancel button.

The following properties of each column can be modified:

Column Alias

The name of the data field that the end-users see.

Column aliases may not contain the following characters:

{ } "

Column Description v2016.3+

Data fields can have description text added. If the data field is hovered over in a selection screen in the Report Designer, the description text will pop up in a tooltip.

Plain Text

Tool tip text can be added verbatim in the Column Description field. In-line HTML tags like <b> can also be used if desired. Press Okay when done, then Apply the change.

Language File

You can also add description elements to the language file, and reference them in the Column Description field.

In the language file, add new elements to the <AdminObjects> section using the following format:

<element id="uniqueIdentifier" tooltip="Description Text"></element>

HTML tags must be encoded like so:

• Encode < as <
• Encode > as >
• Encode ” as "

For example, the following tooltip string encodes “<b>Description</b> Text”, which displays as “Description Text.”

<element id="uniqueIdentifier" tooltip="&lt;b&gt;Description&lt;/b&gt; Text"></element>

After adding the element to the language file, add the ID string to the Column Description field. Press Okay when done, then Apply the change.

Data Type

The type of data Exago should treat the data field as any of the following valid types:

• String
• Date
• Datetime
• Time
• Int
• Decimal
• Image
• Float
• Boolean
• Guid
• Currency

Filterable

Whether this field can be used to filter reports. The available options are:

To prevent users from attempting to build a filter in the Report Designer on an unsupported data type, Filterable should be set to None if the column’s data type is not one of the supported types in the Data Type section above.

This setting affects the availability of the column in the user interface. Filters may still be programmatically added to reports with the API regardless of the setting here.

Visible

If set to False, the data field will not be listed for users.

<entity>  ...
    <column_metadata>
    	<col_source>ExagoFormula</col_source>
    	<col_name>FullName</col_name>
    	<col_type>string</col_type>
    	<col_alias>FullName</col_alias>
    	<col_description>First and Last Name</col_description> 
    	<filterable>false</filterable>
    	<col_value>{Employees_0.FirstName} &amp; ' ' &amp; {Employees_0.LastName}</col_value>
    	<col_sortandgroupbyvalue>{Employees_0.LastName}</col_sortandgroupbyvalue>
 	</column_metadata>
</entity>

Right({Employees.SocialSecurityNumber},4)

Month({Orders.OrderDate})

{Employees.FirstName} & ' ' & {Employees.LastName}

Random(0,65536)

Retrieving Data Object Schema

Many of the dialogs throughout Exago require schema information such as column name, data type, filterable, sortable and visible status, sort-and-group-by value etc…. To enhance performance, schema information can be stored in the application as column metadata. Exago can then read this metadata instead of querying the data source.

While storing the schema as column metadata improves performance, updates to the column metadata will be required whenever columns are added, removed, or re-titled.

Building Metadata in Bulk

The use of column metadata reduces the frequency in which the databases are queried for schema information. To make the process of adding metadata simple, this information can now be built in bulk through the Admin Console.

Metadata can be built in bulk:

• with the Add All Complete Objects, Joins and Metadata v2021.2+ on the corresponding Data Source (v2021.2+)
• with the Bulk Metadata feature described below

Right-click a data object, then click Bulk Metadata.

Exago will then process the information for each data object under that data source and add metadata information for each field within these objects. The Schema Access Type of each object will automatically be set to Metadata for each object.

A new Bulk Metadata tab will open in the Admin Console displaying the processing information.

The metadata information that has been processed and added to each data object may then be edited or removed in the Column Metadata menu.

Please note, however, that if metadata has not been added to or is not enabled for a data object, the following warning will appear next to the data object in the Admin Console. This warning serves to further encourage the use of metadata and notify the system administrator that this data object will query the database each time it requires schema information.

Building Metadata Manually

In versions prior to v2019.1, data sources are queried for schema information by default. This process, however, may cause performance issues if the data sources take a considerable amount of time to return the schema. To avoid these performance issues, metadata may be added manually via the Column Metadata menu.

For Exago to retrieve schema information from the metadata:

1. In Admin Console > General> Other Settings, set Schema Access Type to Metadata. This will force Exago to get all schema information from the metadata for all data objects.
   

   

   Alternatively this setting can be overwritten for individual data objects by setting the Schema Access Type property.

2. For each data object open the Column Metadata menu.
   1. Click the Read Schema button. A message will appear asking you to confirm you want to continue. Click Okay.
   2. Click Okay to close the Column Metadata menu.
   3. Press Okay or Apply to save the data objects.

Other metadata options such as aliasing can still be utilized.

Reading Images from a Database

Exago can read images from a database and load them directly into a cell of a report. When images are stored in a database as a binary string there are two ways that Exago can load them into a report.

1. Change Column Metadata
   1. Open the Column Metadata menu and for the image column set Data Type to Image.
   2. Place the data field containing the images into the desired cell of a report. Upon execution the images will be loaded into the cell.
2. Without changing Column Metadata
   1. Place the data field that contains the images into the LoadImage function. Upon execution Exago will interpret the binary and load the images into the cell.