Loops overview

Loops are a basic element of control flow in computer programming that Postman Flows supports. As a visual programming environment, Flows has its own design patterns that you need to know to make looping structures work for you.

Loops with the Repeat block

In Flows, here’s the most basic design pattern for a loop:

1. Begin with a Repeat block. Use a Number block to set the Count input to the number of times you want Flows to iterate through the loop.

2. Connect a block that does work (for example, an Evaluate block) to the Repeat block.

3. Optionally, add more blocks that do work, connecting them together in a series.

4. Connect a Collect block to the last of the blocks from the previous step.

5. Connect a Display block to the Collect block.

This loop repeats the specified number of times. There’s no way to break out of the loop.

Here’s how the loop will look:

Loops with the For block

Since the For block iterates over a list, you need to provide a list to its input.

1. Begin with a List block. Alternatively, you can use an HTTP Request block to pass in a list from an API.

2. Connect a For block to the List block.

3. Connect a block that does work (for example, an Evaluate block) to the For block.

4. Optionally, add more blocks that do work, connecting them together in a series.

5. Connect a Collect block to the last of the blocks from the previous step.

6. Connect a Display block to the Collect block.

This loop goes all the way through the list - there’s no way to break out. Nor can you start in the middle.

Here’s how the loop will look:

Best practices for working with loops

When planning and creating loops in flows, consider the following facts and principles:

• Collect blocks are great for synchronizing loops, because the flow will wait until all data has been collected before continuing.

• Don’t try to create a variable within a loop. Flows treat variables as constants that can only be set once.

• In Flows, you can construct a double For loop that’s equivalent to a nested loop in programming languages like Python or JavaScript. For example, if a value X in the outer loop is needed to process other values in the inner loop, use an Evaluate block in the outer loop to map X to each item you’re sending through to the inner loop. (See the TypeScript map() function.)

Alternatives to loops

Flows can filter and map data. For computational efficiency, consider doing this with TypeScript in Evaluate blocks rather than with loops.

Here’s an example flow whose source data is a list of JSON objects that describe water and oil wells. The goal is to reveal correlations between a well’s elevation and its status.

• The first Evaluate block filters out all the wells whose elevation falls outside of a range of interest, thus selecting a manageable subset of wells to investigate. The range of interest is from 700 to 960 feet, inclusive, and the following expression performs the filtering:

  1
  tableOfWells.filter(
  2
  (row) => row.Elevation >= 700 && row.Elevation <= 960);

• The second Evaluate block maps values from the data’s Elevation and Status of Well fields to new JSON objects and outputs these as a list ordered by elevation from low to high. Displaying the transformed data as a table makes correlations easier to spot. The following expression performs the mapping and ordering:

  1
  wellList = wells.map((row) => ({["Elevation"]: row["Elevation"],["Status": row["Status of Well"]}));
  2
  wellList.sort((a, b) => a.Elevation - b.Elevation);

You could produce the same result using loops. But especially when operating on large data sets, you’ll get better performance using Evaluate blocks with the TypeScript methods filter(), map(), and sort().