Yup, I noticed that as well. I sent that to the API developers. However, rather then wait for then, is there any way I can circumvent this check, short of copying the whole code for WebClient into my pod with this variation?
brianWed 28 Feb 2018
s there any way I can circumvent this check, short of copying the whole code for WebClient into my pod with this variation?
No, that is a pretty fundamental low-level hook which determines how to map the raw socket input stream to the application level input stream
Jeremy Criquet Tue 27 Feb 2018
After deleting my comment, turns out there's still a bug.
brian Tue 27 Feb 2018
It only uses an absolute API if it detects you are using a proxy server for your given URI.
Double check your proxy settings in etc/web/config.props
Jeremy Criquet Tue 27 Feb 2018
After deleting my comment, turns out there's still a bug.
Jeremy Criquet Tue 27 Feb 2018
After deleting my comment, turns out there's still a bug.
Jeremy Criquet Tue 27 Feb 2018
After deleting my comment, turns out there's still a bug.
Jeremy Criquet Tue 27 Feb 2018
(Really shouldn't have jumped the gun on deleting my comments above...)
So, the issue that remains is I'm getting:
When I do
WebClient.readRes
.The resCode and resHeaders seem to be there.
So, it seems like something else is being included that is bugging out
WebClient
.Running the same URI through curl does not produce any problems (and gives me the requested content).
Jeremy Criquet Tue 27 Feb 2018
Here's the response line/headers:
Passing the response body into curl gives this for the first 16 characters:
It seems like that's a byte order mark? I suspect that's what's causing WebClient to bug out?
EDIT: Ahh that's why:
So, it seems it's unhappy with the server giving the header
Content-Encoding: UTF-8
.Is there any way that I can override this?
brian Tue 27 Feb 2018
That isn't a valid content-encoding which defines a compression encoding (not a character encoding).
See MSDN
Jeremy Criquet Wed 28 Feb 2018
Yup, I noticed that as well. I sent that to the API developers. However, rather then wait for then, is there any way I can circumvent this check, short of copying the whole code for
WebClient
into my pod with this variation?brian Wed 28 Feb 2018
No, that is a pretty fundamental low-level hook which determines how to map the raw socket input stream to the application level input stream