Jocelyn A.

you want to create a csv with 7 columns on the header and 10 columns on the data? that's fairly non standard. most consumer would error trying to read this. you could perhaps write the header line first with a row of 7 values which are your 7 headers at the start of the pipeline, then append the data with the csv formatter, without headers. again though, are you sure this is what is required?

deduplicate works with multiple documents, which is what isn't ultra pipeline compatible. If all you want is to remove exact match within a collection (list of objects) it might be possible with an expression, but there is no snap for this.

it can have any value yes, but the google api I use in the first http client doesn't include the filename in the response, hence I take it from the original object which has it from a previous step. I also change the extension to pdf as I am converting the file to a pdf in the second HTTP client.

I actually had to solve the same issue recently, in my case it was in between 2 google drive APIs. The main issue was caused by the binary input view from the HTTP Client which expects lower case $['content-location'] and $['content-type']. You can use a mapper in between the two to solve this problem, this should solve this for now, while we work to enhance this behaviour so that the HTTP Client Input view support all case variations of the Content* headers.