Alternative to keyValues Function for Pipeline Expressions in ADF

Question

I'm working with Azure Data Factory (ADF) and I need to create a key-value map within a pipeline expression. While the keyValues function works perfectly in Data Flows, it's not available in pipeline expressions.

My scenario involves two separate arrays, one containing keys and another containing corresponding values. I want to combine them into a key-value map format like this:

["key1" -> "value1", "key2" -> "value2", "key3" -> "value3"]

Is there a way to achieve this key-value map creation using pipeline expressions in ADF or Synapse Studio?

Answer 1

Hi @Glasier

Thanks for the question and using MS Q&A platform.

ASAIK there is no direct way to create key-value pairs using the pipeline expression builder in ADF, like you can in dataflows. Here is a method to build the required key-value pairs in ADF pipelines:

I have used 2 parameters with these values:

Let's say req is the variable where we want to store our final key-value pairs. I have initialized it with an empty object {}.

In the 'ForEach' activity, I set the items value to @range(0, length(pipeline().parameters.keys)) to generate the index.

Now, I have taken a variable called temp to apply union on the current value of req and the dynamically built key-value pair for current iteration.

@string(union(json(variables('req')),json(concat('{"',pipeline().parameters.keys[item()],'":"',pipeline().parameters.values[item()],'"}'))))

Finally, I update the req variable in each iteration with the current temp value, i.e., @variables('temp').

After running the pipeline, it will generate the following output in the req variable:

ADF pipelines do not support variables of the object type. If you need to use an object stored as a string, you can convert it to an object using the @json() function.

Hope this helps. Do let us know if you any further queries.

---

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.