Don't auto-send STOMP content-length header if one was explicitly set

[scop]

I'm not sure if a code path where this (= sending more than one content-length header) happens exists, but nothing prevents the issue from happening ATM.

[johnament]

Does this introduce any form of a security issue? What happens on the receiver side if the content length doesn't match the header?

[scop]

I don't think there's any way to know. But then again there probably isn't any way to know what might happen on client side if there are duplicate or multiple content-length headers.

[clebertsuconic]

I think this is ok.. besides.. I need to solve a little conflict with my other change.. so I will merge this now.

@johnament if you still think there's something wrong we can still review this (commit-then-review).. The ideas on these PRs is just to avoid mistakes not to censor commits, we can still rollback later if there's an issue with it.

[clebertsuconic]

@johnament I'm not trying to prevent any discussions here though.. please don't take me wrong.. I'm just saying I will go ahead and merge this.. don't read anything on it please :)

[johnament]

@clebertsuconic none taken at all.
I just know that issues like this tend to lead to buffer overflow type errors. It's technically an unrelated issue, I'll take a look at the server side.