Conditional Split

The ConditionalSplit transformation evaluates each incoming row against a predicate. If the condition is met, the row is routed to the main output; otherwise, it can be sent to a secondary path. This component is useful for branching logic in data flows, allowing clean separation of matching and non-matching records.

Overview#

The ConditionalSplit transformation allows you to route rows into different outputs based on a condition. If a row matches the provided predicate, it is sent to the primary output; otherwise, it can be routed to an alternate output. This is useful for separating data into different processing paths.

Buffering#

• Input Buffer: Buffers incoming rows before evaluation.
• Primary Output Buffer: Stores rows matching the predicate.
• Secondary Output Buffer: (optional) Stores rows that do not match.
• The transformation is non-blocking and processes each row as it arrives.

Example: Conditional Routing#

var split = new ConditionalSplit<MyRow>();
split.MatchPredicate = row => row.Col1 >= 3;
source.LinkTo(split);
split.LinkTo(dest);                   // Rows with Col1 >= 3
split.LinkUnmatchedTo(unmatched);    // Rows with Col1 < 3

You can track how many rows were routed to each output using the following properties:

• MatchCount: The number of rows that satisfied the predicate and were sent to the main output via LinkTo.
• UnmatchedCount: The number of rows that did not satisfy the predicate and were sent to the alternate output via LinkUnmatchedTo.

These counters are updated automatically during execution and can be used for monitoring or validation purposes after the data flow completes.

Working with Dynamic Objects#

Use the non-generic version with dynamic data:

var split = new ConditionalSplit();
split.MatchPredicate = row => ((dynamic)row).Col1 >= 3;

source.LinkTo(split);
split.LinkTo(dest);
split.LinkUnmatchedTo(unmatchedDest);