Great question!
The “split” transform and the “extract” transform both create new columns based on patterns. However, the two transforms treat the original column and the matched pattern differently.
- Handling of the original column: When you apply a “split” transform to your dataset, Trifacta drops the original column. When you apply an “extract” transform to your dataset, Trifacta preserves the original column.
- Handling of the matched pattern: The matched pattern is the pattern that you identify in the wrangle script for the “split” transform and the “extract” transform. The “split” transform treats the matched pattern as a delimiter and splits the column around the matched pattern. The “extract” transform treats the matched pattern as the data element desired in the new column.