Commit Graph

978 Commits

Author SHA1 Message Date
Daniel Stenberg
0866c1b52d Michael Curtis' updates 2002-04-08 21:59:06 +00:00
Daniel Stenberg
5d28a857a4 based on Jacky Lam's "HTTP 1.0 304-only" fix, this change makes a 304 reply
always stop reading after the headers no matter what 'close' is.
2002-04-08 07:27:22 +00:00
Daniel Stenberg
a03fd7b81c T. Bharath pointed out the flaw in ConnectionExists() for how we didn't
check proxy connections for "deadness" before they were re-used
2002-04-04 12:23:14 +00:00
Daniel Stenberg
1cfcbc50a6 Fixes the problem Jacky Jam pointed out, where libcurl will "hang" for an
extra second after having downloaded headers-only
2002-04-04 12:19:56 +00:00
Daniel Stenberg
d560207ea1 This corrects VERBOSE PASV ftp transfers on AIX (and OSF1/Tru64)
Gerhard Herre reported this in bug report #536238
2002-04-04 06:07:56 +00:00
Daniel Stenberg
90b51831fd Tor Arntsen's fix for "CGI_Lite" compliance! 2002-04-03 11:11:01 +00:00
Daniel Stenberg
547e91dbf0 Clarence Gardner pointed out the not-taken-care-of return codes from SSL_read 2002-04-02 09:33:38 +00:00
Daniel Stenberg
04da96e044 Jacky Lam added memdebug.h include to prevent crashes when that is used 2002-04-02 06:36:47 +00:00
Daniel Stenberg
541e5a3b82 Jacky Lam cookie parser fix for domains with preceeding dot 2002-03-25 09:08:33 +00:00
Daniel Stenberg
5af61716aa Miklos Nemeth's update 2002-03-23 15:41:17 +00:00
Daniel Stenberg
67b0f9aacd no longer include "multi.h", it comes with the regular curl/curl.h now 2002-03-20 10:54:17 +00:00
Daniel Stenberg
854277bae5 crlf replacement on uploads did not work. test case 128 was added just now
to make sure it remains functional.
2002-03-20 10:53:24 +00:00
Daniel Stenberg
ca9760756a verbose FTP PASV output could output a "random" name from the stack based
array
Also, uploading data with --crlf cannot check that the proper size was
uploaded and thus should not warn if the sizes differ. This can be changed
in the future by having the expected size in the connectdata struct and then
increase the expected size for each byte that is added in the --crlf
replacement process.
2002-03-20 10:52:24 +00:00
Daniel Stenberg
ab6b0f9843 removed multi.h 2002-03-19 14:32:43 +00:00
Daniel Stenberg
6b4532b592 moved to ../include/curl/ 2002-03-19 10:34:06 +00:00
Daniel Stenberg
974f314f57 copyright string (year) update 2002-03-19 07:54:55 +00:00
Daniel Stenberg
eaff1a344e made it pass stricter compiler flags with less warnings 2002-03-19 07:32:35 +00:00
Daniel Stenberg
5c691ed835 verbose PASV transfers passed a bad buffer size to the name resolver functions
and it cause cause a crash.

Albert Choy found and fixed it.
2002-03-16 16:59:47 +00:00
Daniel Stenberg
f73864a045 corrected SSL builds 2002-03-15 14:46:59 +00:00
Daniel Stenberg
5f758fbd11 make sure we return CURLE_WRITE_ERROR if the write callback returned
an error, even if we were decoding a chunked-encoded transfer
2002-03-15 12:42:41 +00:00
Daniel Stenberg
fb29529a52 Jun-ichiro itojun Hagino <itojun@itojun.org>:
Now first check if IPv6 is supported, then use PF_UNSPEC. If not, use PF_INET.
It'll solve both the "slow name lookup" problem on IPv4 and still work fine on
IPv6 hosts.

Bug report #530204 has more details:
http://sourceforge.net/tracker/?func=detail&atid=100976&aid=530204&group_id=976
2002-03-15 09:54:30 +00:00
Daniel Stenberg
3cd2673077 bug report #530204 correctly identified that revision 1.52 broke ipv6
functionality and this change reverts this.

However, with this revert we bring back problems on (some/all?) non-IPv6
enabled Linux machines that have getaddrinfo().
2002-03-15 08:45:09 +00:00
Daniel Stenberg
468b787272 if CURLOPT_POSTFIELDS isn't set, but CURLOPT_POST is, we will assume that
we should read the POST-data from the read callback
2002-03-14 14:39:23 +00:00
Daniel Stenberg
cfdb6f851c 2002 2002-03-14 14:37:41 +00:00
Daniel Stenberg
7886f120f3 CURLOPT_POST deserved a new comment with the new POST-by-callback support 2002-03-14 14:37:16 +00:00
Daniel Stenberg
76fe69b133 no longer attempts to SIZE a NULL pointer, as that wasn't very clever
(but didn't crash or anything)
2002-03-13 13:13:19 +00:00
Daniel Stenberg
e1bae4fc7e Setting CURLOPT_PASSWDFUNCTION to NULL now restores the internal function. 2002-03-13 13:10:52 +00:00
Daniel Stenberg
bc9705f758 sendf() now deals with Curl_write() returning -1 properly, which it might
do if the write would've blocked
2002-03-13 13:09:37 +00:00
Daniel Stenberg
c819e234b8 now supports all options in arrays, except the CURLFORM_ARRAY itself 2002-03-13 12:10:20 +00:00
Daniel Stenberg
9f374c2050 Added support for CURLFORM_FILENAME to set the filename field of a file
part.
2002-03-11 15:18:59 +00:00
Daniel Stenberg
fe3c874001 detect fclose(NULL) 2002-03-08 15:31:44 +00:00
Daniel Stenberg
017be8a882 Jean-Philippe Barrette-LaPierre fixed the CURLOPT_PASSWDFUNCTION to make
NULL set back the internal default function
2002-03-08 15:06:42 +00:00
Daniel Stenberg
f51f2417c5 Brad corrected the include path (again) 2002-03-06 22:19:16 +00:00
Daniel Stenberg
80b004a57d Wesley Laxton's CURLOPT_PREQUOTE work 2002-02-28 23:31:23 +00:00
Daniel Stenberg
ea8476a2dc Ralph Mitchell's SSL problems made me notice that we didn't increase the
header byte counter properly
2002-02-28 15:13:35 +00:00
Daniel Stenberg
cb85ca18ab more fancy alloc, we store the size in each allocated block so that we
can destroy the full allocated area just before we free it
2002-02-28 12:37:05 +00:00
Daniel Stenberg
0cacbc892c always allocates at least 64 bytes for real, and damages them before free 2002-02-28 12:18:15 +00:00
Daniel Stenberg
6753c3c715 made building outside the source tree work again, Kevin Roth reported 2002-02-27 15:09:23 +00:00
Daniel Stenberg
3612c3774e made Max-Age work as defined in the RFC.
my brain damaged fix to not parse spaces as part of the value is now fixed
to instead strip off trailing spaces from values.
2002-02-27 07:38:04 +00:00
Daniel Stenberg
66b8f48a88 When saving a cookie jar, set field 1 (counted from 0) properly to TRUE if the
domain starts with a dot.
2002-02-26 13:18:08 +00:00
Daniel Stenberg
a23a897ad2 removed crash on weird input, this also better discards silly input 2002-02-26 13:07:53 +00:00
Daniel Stenberg
e6ed3478ea automake usage and options cleanup 2002-02-25 14:08:18 +00:00
Daniel Stenberg
b32a39f44f oops, #if not #ifdef 2002-02-25 10:12:04 +00:00
Daniel Stenberg
e4fefd088d cygnus can't include winsock.h even though it has it, why we need to
make a different and more complicated check for when to include it
2002-02-25 08:20:29 +00:00
Daniel Stenberg
b1ffe7b74a better time selection for the connect timeout 2002-02-22 15:44:37 +00:00
Daniel Stenberg
cbca19d6c2 lib/config.h.in added to dist 2002-02-22 07:51:23 +00:00
Daniel Stenberg
721b05e343 Nico Baggus' VMS tweaks 2002-02-20 13:46:53 +00:00
Daniel Stenberg
a333bddeeb Andrés García solved bug report #515228 by making sure the progress meter
is updated even if everything is read in one single pass, as the windows
functions apparantly does more often than other systems.
2002-02-20 13:38:34 +00:00
Daniel Stenberg
4c6a52fe90 corrected reference to multi-using examples 2002-02-19 11:02:01 +00:00
Daniel Stenberg
792d73a9cf include winsock.h on window boxen to work smoother there 2002-02-19 11:00:34 +00:00