Overview
Etleap’s wrangler simplifies the process of transforming your data through an intuitive user interface.
This section guides you through common wrangling steps such as:
- Parsing JSON data by flattening JSON objects, turning JSON arrays into tables, and fetching values from JSON structures using JMESPath queries.
- Extracting or manipulating substrings by splitting, editing, cutting, or extracting.
- Handling dates, date-times, and timestamps by turning them into columns with the appropriate types in the destination table.
- Adding metadata such as the file path or modification time to the destination table.
There are also guides for common use cases such as:
- Removing invalid data. This is useful in cases where the source data needs cleaning in order to be useful or correct downstream.
- Deleting or editing existing wrangler steps in order to change an existing wrangler script.
Limitations
- The Wrangler does not support outputting schemas with more than 2000 columns.
- A Wrangler script can only contain one Partition Output transform.
- For database sources, there must be a step (Parse CSV or Split Column) that generates the same number of columns as the source table’s schema.
- This step must be followed by a schema-interpreting step (Rename Columns) that assigns the same names as the source table schema.