On Tue, Oct 21, 2008 at 01:40:49AM +0200, Ragnar Lonn wrote:
> I'm experimenting with using libcurl to simulate multiple web clients
> fetching things from a webserver, but I must be doing something wrong
> somewhere. I've scanned the list archives and read the docs but can't
> understand why it won't work! What happens is my program will only read
> part of the data from the webserver. I.e. if I run the program and tell
> it to perform 1 fetch of an URL that I know is 8 kB, it will execute and
> report getting around 2 kB when it finishes.
>
> What seems to happen is libcurl sets running_handles to 0, which makes me
> stop execution but not all data will have been read at that point. Maybe I
> have misunderstood something about how it is supposed to work?
>
> Any help would be very appreciated. I'll include the code below. It uses
> epoll and has been tested on a Linux (Ubuntu) machine.
[...]
> size_t client_write_data(void *buffer, size_t size, size_t nmemb, void
> *userp) {
> total_bytes += (unsigned int)(nmemb * size);
> }

You're not returning anything from the write callback. Which means that
libcurl treats the garbage that is returned as a code to abort the download.
You should really turn on warnings in your compiler to catch this kind
of thing.

>>> Dan

-- 
http://www.MoveAnnouncer.com              The web change of address service
          Let webmasters know that your web site has moved