Solved
How to change Content-Length?
ReadyAPI seems to calculate the Content-Length but doesn´t account for special symbols like Umlaute (ä,ö,ü) which adds additional bytes to the content-length. I added a "content-length" manually to ...
Forum Discussion
It definitely appears to be a bug and in the process of debugging this issue we discovered more problems. Other Header Values appear to be ignored, too, even though they appear correct in ReadyAPI. Even though we use utf-8 umlaute appear as ?. The same call made from Postman with the exact same content and header works correctly. When we tried to check the traffic via Fiddler, to see what exactly is being passed from one to the other, we were not able, nothing appears in Fiddler, even though when making the same call from Postman it appears correctly in Fiddler.
Taz,
I encountered similar issues with ReadyAPI and SoapUI when I had special characters or encoding mismatches in my request data. Especially when the request data are exchanged via MS Teams or similar tools, unexpected characters may be inserted.
Postman cleaned the data and sent it modified, so no issue appeared, whereas ReadyAPI tried to use exactly what I inserted and ended up in error.
I would recommend double-checking your request data (encoding, invisible special chars, etc.) Once I removed the special chars, ReadyAPI worked fine.
Thanks. Unfortunately it seems the problem is maybe lying somewhere else. I figured out why it isn't working. We are putting text into the body using utf-8. This works fine when there's no special character in it like Umlaute. I tried now to convert the text first to ANSI and pasted it then into the body. Now it works correctly. Yet this is bad, because in our tests we're picking up content from another request and putting the response straight into this request. It means now we first have to convert it to ANSI Format.
Taz,
if there's a correct UTF in the beginning, you should be able to paste it into ReadyAPI with no issues. ReadyAPI is fine with UTF8 and national characters, we also use them. Check also the encoding headers in your HTTP request.
I recommend verifying whether the source data are correct. Converting data to ASCII is an unwanted modification of the test data, they need to be the same as in reality.
After contacting support we found the solution. See my latest post.