How to format FileIDs in Text Aggregator as clean JSON array for HTTP request?

[ReiFel]

Hi everyone,

I’m trying to build a flow where I upload multiple files and then send their FileIDs to an OpenAI HTTP request.

I already have all the FileIDs collected in a variable called AllFileIDs. Now I want to use a Text Aggregator to format them as a clean JSON array, like this:

{
  "file_ids": ["file-abc123", "file-def456", "file-ghi789"]
}

But I’m struggling to get the syntax right in the Text Aggregator.

I need each FileID in quotation marks, separated by commas, so the final result can go directly into the HTTP request body.

How can I best achieve this?

Any tips on the correct Text Aggregator settings or alternative approaches would be much appreciated!

Thanks a lot in advance!

[Stoyan_Vatov]

Hey there,

can you share the input bundle as well so we can see what it originally looks like?

[ReiFel]

Hey Stoyan,
thanks for your replie.
The second screenshot is the input bundle.
so basically I have an unclear amount of “all-file-ids” as an array. (also potentially 0)

[ReiFel]

The best would even be to get the a JSON in the format which I showed above from this Array aggregator, which has a nested structutre, where the file IDS are separated in pdf-file-ids and other-file-ids. (seen in the screenshot attached)
What i get from my current text aggregator (in the other attaced screenshot), is just the file names, without exclamationmar, and the 2 “other files” are not even separated with a space.
How could correct json come out, so i could feed the file IDs to a HTTP request to an OpenAI thread?

Thank you for your help.