Package: libcurl3-dev
Version: 7.14.0-5
Severity: normal
I started rewritting a small web client using CURL, and for each
curl_* function, I had:
int errnum;
errnum=cur_*();
if (CURLE_OK!=errnum) {
fprintf(stderr, "%d: %s\n", __LINE__,
curl_easy_strerror(errnum));
exit(EXIT_FAILURE);
}
I read about CURLOPT_ERRORBUFFER, and it seemed to me that it would
simplify that repeated code a bit. So I added, immedately after
creating an easyhandle (involving an error check, as above):
char errbuf[CURL_ERROR_SIZE];
errnum=curl_easy_setopt(handle, CURLOPT_ERRORBUFFER, errbuf);
if (CURLE_OK!=errnum) {
fprintf(stderr, "%d: %s\n", __LINE__,
curl_easy_strerror(errnum));
exit(EXIT_FAILURE);
}
and then I deliberately didn't set a URL, and then called:
errnum=curl_easy_perform(handle);
if (CURLE_OK!=errnum) {
fprintf(stderr, "%d: %s\n", __LINE__, errbuf);
exit(EXIT_FAILURE);
}
(This time without the call to curl_easy_strerror()). But, I got
garbage string output, so apparently I have no clue how this works.
Does it just specify a different buffer, instead of a buffer to be
mallocated by curl? Could you improve the documentation? Maybe
include an example?
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
