You’re offline. This is a read only version of the page.
13
Completely new to Power Automate:
My flow responds to the arrival of an email with a pdf attachment, and whose body contains certain key values, by POSTing that attachment to an api in a remote system. More precisely, it makes an api call to check if the remote system already has a row matching these keys, and if it does not, then it makes two more calls:
A. One api call to create a new row to contain these keys (this works fine)
B. Another api call, to a different endpoint, to attach the pdf to this new row <-- resulting pdf is corrupt
So (B) is a POST whose body is the content of the PDF.
All this runs fine end to end, but the pdf is unreadable (that is, as a pdf). If I compare it to the original (that is, in a text editor) they don't match. The file produced by (b) is also much larger, almost twice the size of the original.
I've already validated (not shown here) that there only one attachment, ending in pdf. Include Attachments is Yes.
Here I've created the new row (A) and I'm putting together the post to B.
When I download the attachment content the input does not match the output, in that the output has a leading 'body' key.
Input
Share
Report{"name":"attachmentContent","value":"%PDF-1.7\n%����\n4 0 obj\n<</Subtype/Image/Width 1728/Output
{"body":{"name":"attachmentContent","value":"%PDF-1.7\n%����\n4 0 obj\n<</Subtype/Image/Width 1728/I also tried multipart, following several examples posted here, but the result was no different.
What's going on here? I appreciate your help!
Hi,
Thanks for your follow up post! I am glad that you were able to get this figured out. Hopefully this will prove useful for future communities users who are having similar issues. Remember to set the solution for this thread.
Regards,
Alex
-------
Community Support Team _ Alex Rezac
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
Here is the fix. It is similar to the fix described here.
The problem is this block where I populate attachmentContent. Somehow it gets all messed up, and I get back content (not shown here) that looks like my pdf but is corrupted somehow, and is almost twice the correct length.
So remove that variable and do this instead:
Note that my flow is for cases where there is only a single attachment. I'm not sure what would happen if there were multiples.
I'm not sure why the Body presentation does not preserve the wrapping call to base64ToBinary() -- this looks like a bug - but it works anyway.
#1
WarrenBelz
146,645
Most Valuable Professional
#2
RandyHayes
76,287
Super User 2024 Season 1
#3
Pstork1
65,997
Most Valuable Professional
All responses (
Answers (
