We’re working to improve the documentation on this point, but that paragraph basically says that you can have two cases: linear process and repeatable process.
In a linear process, the robot has only one transaction.
For example, suppose you want to get the name of the user and write it in Notepad. Since you want to do it only once, you check whether
1. If it is, then you ask the user’s name using an Input Dialog activity and stores the name in
out_TransactionItem. If it’s not
1 (meaning the first and only transaction was already processed), then you set
Nothing. This means that there’s no data left to be processed and the framework will conclude its execution.
Sample project: ReFrameworkLinearProcess.zip (526.2 KB)
On the other hand, in a repeatable process, the robot can have many transactions.
For example, suppose you want to read data from some people from a CSV file and write it in Notepad. In this case, you can set the
out_TransactionItem argument (in the GetTransactionData.xaml file) to something like
This means that you’ll get one row from the DataTable and that’ll be your transaction item.
One big advantage in this case is that you get a way to process each row independently and, if Notepad crashes during the process, the framework will automatically reopen it and continue the processing from where it stopped.
Sample project: ReFrameworkRepeatableWithDatatable.zip (780.0 KB)