Components perform all the data querying, processing, and storing in Data Flow. You can see the list of available components on the left of the data flow designer. Depending on what the components do, they are divided into three categories: sources, destinations, and transformations.
You can add components to your data flow by dragging them from the components list to the diagram. If the component has an input, connect the output of the necessary component to it. After this, you can click this component and configure its details in a sidebar on the right. Each kind of components has its own settings described below.
Source is a component that extracts data from a data source and starts the data flow. Source is the very first component you should add to the diagram.
After you have added Source, click it and configure its details — select a connection to get data from, select and configure the action to perform in order to obtain data and map its parameters if needed. You can also give Name to the component displayed on the diagram.
Source component has one regular output and no input. Its output returns the data obtained by the selected action.
Target is a component that stores input records in a data source. It can also write data to logs.
To configure Target, select a connection to load data to, select and configure the action to perform in order to store the input data and map its parameters if needed. You can also give Name to the component displayed on the diagram.
If batch data loading is supported for a connector used in the selected connection, the Target component loads data in batches into the data source by default. You may change this behavior by setting the Single Row parameter to True.
Target component has one input. It also has one regular output, and records, successfully loaded to the data source, are directed to this regular output. For some connectors, actions of the Target component also may obtain values of columns autogenerated in the data source, for example autogenerated IDs, and add the corresponding columns to the scope of the regular output.
Target component also has an error output, and records, failed to be loaded to the data source, are directed to this error output. Target component adds a column with the error message to the scope of this output.
Transformation components either transform the input data, i.e. add new columns to the input records, or modify variable values. They may also split input records into multiple outputs.
Split is a component that passes all the input records to each of its outputs. It has one input and can have any number of regular outputs, which are exact copies of the same input. It does not need any configuration, you just drag it from the component list to the diagram and connect its input and outputs to the corresponding components.
Conditional Split is a component that splits records into different outputs depending on the specified conditions.
To configure Conditional Split, you need to create one or more conditional outputs. You can also give Name to the component displayed on the diagram.
To create a conditional output, click next to the Conditional Outputs header and then specify the output name and condition.
For the condition, you need to enter an expression, using our expression syntax. The expression may refer input fields and data flow variables. You may also use the expression editor that provides helpful features to enter an expression, like code completion, validation, and previewing expression results for specified property values.
Note that the expression must evaluate to a boolean value. If it returns a value of another type, the record is sent to an error output.
The conditions are fulfilled in the same order as you specified in the list of Conditional Outputs. For example, the first record is checked whether it complies with the first condition — whether the condition evaluates to true. If it complies, the component sends the record to the first conditional output. This record is not checked anymore for compliance with other conditions in the next conditional outputs. The same way as first record all other records are checked and distributed between conditional outputs.
If the record does not match any of the conditions specified in the conditional outputs, it goes to the Default output, which is always available for a Conditional Split component.
In addition to regular outputs, the Conditional Split component has an error output. It sends rows to this output in two cases: when an expression evaluation fails with an error, and when expression returns a result of type other than bool.
When you connect a regular output of the Conditional Split component to the input of another component, you will be asked to select the output name in order to determine the output to connect to.
Row Count is a component that counts the input rows and writes the result to a data flow variable. It passes its input records to its output without any modification.
To configure Row Count, you need just to select a variable to put the number of the rows to. The variable must have a numeric type. You can also give Name to the component displayed on the diagram.
Lookup is a component that matches input records with records from another data source and adds columns of the matched records to the scope.
To configure Lookup, you need to specify Behavior and select a Connection to get data from. Afterwards, select and configure the action to perform in order to obtain data and map its parameters if needed. You can also give Name to the component displayed on the diagram.
The Behavior parameter determines how Lookup distributes input records that do not have a match or have multiple matches between Lookup outputs. Lookup always have a regular Default output and an Error output.
Here is the list and descriptions of possible Behavior values:
- FirstOrDefault - for multiple matches, first returned row is taken, and its corresponding fields are added to the input row. If no matches are found, the fields added by lookup are filled with NULL values.
- First - for multiple matches, first returned row is taken, and its corresponding fields are added to the input row. If no matches are found, the row goes to the Error output.
- Single - A row goes to the Error output both if no matches or if multiple matches are found. The row only goes to the Default output if a single match is found.
- SingleOrDefault - if no matches are found, the fields added by lookup are filled with NULL values. If multiple matches are found, the row goes to the Error output.
An Extend component adds calculated columns to input records. It allows you to add one or more columns, values of which are calculated based on the input column values and data flow variables.
To configure Extend, you need to set up its Output Schema using the Mapping Editor. Create parameters, which correspond to the result columns, and configure their mapping.
Extend component has one input and one regular Default output for successfully processed rows with calculated columns added. It also has the Error output, to which it passes records, for which one of the expressions failed with an error, and adds the Error column with an error message to them.
A Value component selects a value obtains the specified value calculated based on the values from input record and assigns it to the specified variable. It can select the min, max, first, or last value from all the processed records.
You can use the Value component, for example, to store the maximum value of record creation time in a parameter, and query only records with the created time greater than this value during the next data flow execution. Thus, every time the data flow starts, it will process only new records.
To configure the Value component, enter an expression using our expression syntax into the Value box. The expression may refer input fields and data flow variables. You may also use the expression editor that provides helpful features for entering an expression, like code completion, validation, and previewing expression results for specified property values. In a simple case the expression is usually just the name of an input column.
Then select the Value type — min, max, first, or last. Finally, specify a Variable to store the value in. You can also provide a Name for the component, which is displayed on the diagram.
The Value component has one input and one regular Default output for successfully processed rows. It sends successfully processed input rows to its output unchanged. It also has the Error output, to which it passes records, for which the expressions failed with an error, and adds the Error column with an error message to them.
Components that interact with data sources perform actions. Action is an operation performed by a component to obtain data from the data source or store them. These are the following components: Source, Lookup, Target.
Action parameters connect component input and variables with the action. They are mapped to input columns and data flow variables using expressions in the Mapping Editor. Their values are calculated for each input row in a batch, and then the action is executed for each row. Some actions are executed for each row separately, and some are executed in batches, for the batches of input rows.
For more information about actions, their configuration, and mapping their parameters see Actions. There you can also find the list of actions, common for most of supported connectors.