In order for Data Splitter to produce any results actions must be defined. Actions can be run :
- before / after a run,
- before / after processing each stream (file, email, URL),
- when a node's pattern is recognized in the input.
They are defined via the Definitions menu :
- pre- and post- run actions
- pre- and post- stream actions
- action groups
Actions can also be defined for any node via the node's drop-down menu.
To define or modify an action :
- start the appropriate action definition dialog,
- press the F1 key in the Action column,
- select the desired action from the list,
- enter the appropriate parameters in the action's dialog.
The following list of actions outputs the current file name to two output streams -
- a text file,
- an HTML file :
The actions ...
Append data to a target. Data includes :
Targets include :
Similar to the Send to action (below), which overwrites the target, if applicable.
Clear the indicated variable
Execute action group
Execute a group of actions. See the action groups topic, below.
Execute an SQL statement contained in a variable.
Execute a user-defined function.
Stop scanning the current input stream or stop the run. See the Halt topic.
Add a positive number to a counter.
Output table row
Add a new row to the specified database table.
Return from processing the current node group.
This is a powerful feature that allows scanning of "streams within streams". Also known as "recursion".
Send data to a target, overwriting if applicable. Similar to the Append to action (above). Overwrites variables, but equivalent to "Append to" for file destinations.
Set a counter to a value.
Start node =
Set the start node to a designated node.Special considerations regarding the start node assignment action
1. If this action is used to reassign the start node, the new assignment remains in effect until :
- the next time a start node assignment action is performed, or
- the user manually resets the start node.
So, if the start node is initially node A, and the start node assignment action reassigns it to node B, node B will be the start node :
- at the start of the next run,
- if the solution file is saved in that state (with node B as the start node).
In most cases, if the start node assignment action is used on a node, it should also be performed in either the pre-run or (more likely) the pre-stream actions to set the desired start node appropriately.
2. Start nodes should be given unique tags. When the start node assignment action is performed Data Splitter scans its internal list of nodes until it finds the first matching tag. If two nodes share that tag the first one in the list will be used. Note that the internal list is not guaranteed to be in node number order. Therefore, using the start node assignment action to set the start node to a node with a non-unique tag may produce unpredictable results.
Actions can be grouped into a "procedure" called an action group, which can be called with the Execute action group action. This action has one required parameter, the action group tag.
"Execute action group" has a checkbox option, "Execute once per input stream". Executing an action group oncer per input stream can be useful, for example, when performing a text search. An action group called "OutputFileName" can be created to output the file name and formatting information. This action group is called when the search text is recognized, but only "once per input stream". This way, the file name is output just once for each file containing the search text.