On Wed, 17 Oct 2012, Paul Peng wrote:

> when it needed to handle chunk HTTP, the responses always lost data(checked
> through byte size and md5sum) compared with the responses I captured using
> wireshark. The most obviously losing data are the chunked sizes at the
> beginning of the HTTP responses' message body and the digital 0 to indicate
> ending at the end.
>
> I suspect my callback function got problem. Here are the relevant codes:

I'm not aware of any such problem in the chunked-encoding decoding code
(assuming you use a modern libcurl version)? We do have several test cases to
verify the chunked decoder.

It would be easy for you to just run the curl tool against such a URL to
verify if the problem is in your code or elsewhere. If you can repeat it with
curl, then please show us a byte stream that we can repeat this problem with!

-- 
  / daniel.haxx.se
-------------------------------------------------------------------
List admin: http://cool.haxx.se/list/listinfo/curl-library
Etiquette:  http://curl.haxx.se/mail/etiquette.html