Compare Datasets
Compare two lists by a key, row by row.
Do things in Compare Datasets
Add a Compare Datasets step. Choose the row to see all its fields.
Later steps read a step’s output by its name: see Use data from earlier steps.
What changed in version 2?
New steps use version 2.0.0. A step made before keeps 1.0.0 and works as it always has. The step’s panel shows its version at the top right. On an older step it reads v1.0.0 · 2.0.0 available: press it to choose the version.
In 2.0.0: Ignore Type Differences is on by default, so 30 and "30" count as the same value. In version 1.0.0 it is off, and types must match exactly.
Not in your list? See When an app isn’t here.
For developers
In the API, which calls apps pieces, Compare Datasets is pieceId: "compare" with actionId: "compare". Get an app returns every field.
It computes from its inputs only (pure). See How versions and side effects work.
Compare Datasets › Compare Datasets
Compare input A (old) with input B (new) by a key field.
- Input A (old) · Required
The baseline array, such as {{steps.previous.rows}}.
- Input B (new) · Required
The new array to compare against A.
- Key Field(s) · Required
Field(s) that identify a row, present in both arrays. For a composite key, list several comma-separated (such as orderId,lineId), and all must match. Dot-notation nested paths are supported.
- Fields to Skip Comparing · Optional
Comma-separated fields to ignore when deciding whether two matched rows differ (such as updatedAt, lastSeen). They still appear in the output rows.
- Ignore Type Differences · Optional
Treat a number and the same number written as text (30 and "30"), or true and "true", as equal. Rows read from a CSV or a Sheet hold text.
On by default.