platform/upstream/curl.git
22 years agocurl_formadd() now returns 'CURLFORMcode' instead of int, to better enable
Daniel Stenberg [Tue, 21 May 2002 07:44:27 +0000 (07:44 +0000)]
curl_formadd() now returns 'CURLFORMcode' instead of int, to better enable
checking for particular errors. curl/curl.h defines the errros

22 years agoRoland Zimmermann's hint, we use SSL_CTX_use_certificate_chain_file() instead
Daniel Stenberg [Mon, 20 May 2002 14:25:35 +0000 (14:25 +0000)]
Roland Zimmermann's hint, we use SSL_CTX_use_certificate_chain_file() instead
of the previous one that used SSL_CTX_use_certificate_file()

22 years agoremoved compiler warnings
Daniel Stenberg [Fri, 17 May 2002 08:15:33 +0000 (08:15 +0000)]
removed compiler warnings

22 years agoFreeBSD needs sys/types.h before we include sys/select.h that was included
Daniel Stenberg [Fri, 17 May 2002 07:57:13 +0000 (07:57 +0000)]
FreeBSD needs sys/types.h before we include sys/select.h that was included
mainly for AIX in the first place...! As reported in bug report #556869

22 years agoFixes bug report #556930 - we need to make sure that the data is all right
Daniel Stenberg [Fri, 17 May 2002 07:49:28 +0000 (07:49 +0000)]
Fixes bug report #556930 - we need to make sure that the data is all right
after we've realloc() the packed hostent struct.

22 years agoAdded item 4.10 after talks with Russ Freeman
Daniel Stenberg [Wed, 15 May 2002 21:40:29 +0000 (21:40 +0000)]
Added item 4.10 after talks with Russ Freeman

22 years ago7.9.7 commit
Daniel Stenberg [Mon, 13 May 2002 09:40:16 +0000 (09:40 +0000)]
7.9.7 commit

22 years agowe don't need win32sockets.c anymore, we support this internally
Daniel Stenberg [Mon, 13 May 2002 07:29:22 +0000 (07:29 +0000)]
we don't need win32sockets.c anymore, we support this internally

22 years agofopen.c added, a fopen() style emulation for URL reading
Daniel Stenberg [Mon, 13 May 2002 07:28:10 +0000 (07:28 +0000)]
fopen.c added, a fopen() style emulation for URL reading

22 years agoFriday's fixes
Daniel Stenberg [Sun, 12 May 2002 16:10:12 +0000 (16:10 +0000)]
Friday's fixes

22 years agoAIX 5.1
Daniel Stenberg [Fri, 10 May 2002 16:01:24 +0000 (16:01 +0000)]
AIX 5.1

22 years agoKein Roth made --trace-ascii look even better, and make OD 0A occurances
Daniel Stenberg [Fri, 10 May 2002 15:59:42 +0000 (15:59 +0000)]
Kein Roth made --trace-ascii look even better, and make OD 0A occurances
get output as plain newlines.

22 years agoAIX wants sys/select.h for the fd_set stuff in curl/multi.h, and even though
Daniel Stenberg [Fri, 10 May 2002 14:37:39 +0000 (14:37 +0000)]
AIX wants sys/select.h for the fd_set stuff in curl/multi.h, and even though
it is a bit ugly work-around to add this here, it is still a working work-
around! ;-)

22 years agoPatrick Smith's contributed docs improvements for when NLST is used by
Daniel Stenberg [Tue, 7 May 2002 23:36:53 +0000 (23:36 +0000)]
Patrick Smith's contributed docs improvements for when NLST is used by
curl...

22 years ago7.9.7-pre2
Daniel Stenberg [Tue, 7 May 2002 21:52:38 +0000 (21:52 +0000)]
7.9.7-pre2

22 years agoAdded --trace-ascii support
Daniel Stenberg [Tue, 7 May 2002 13:13:17 +0000 (13:13 +0000)]
Added --trace-ascii support

22 years ago--trace-ascii and --junk-session-cookies were added
Daniel Stenberg [Tue, 7 May 2002 13:12:12 +0000 (13:12 +0000)]
--trace-ascii and --junk-session-cookies were added

22 years agosupport for ingoring session cookies added
Daniel Stenberg [Tue, 7 May 2002 09:58:13 +0000 (09:58 +0000)]
support for ingoring session cookies added

22 years agothe code for case CURLOPT_DEBUGDATA code broke the CURLOPT_STDERR one!
Daniel Stenberg [Mon, 6 May 2002 18:30:17 +0000 (18:30 +0000)]
the code for case CURLOPT_DEBUGDATA code broke the CURLOPT_STDERR one!

22 years agorecent fiddling
Daniel Stenberg [Mon, 6 May 2002 13:44:00 +0000 (13:44 +0000)]
recent fiddling

22 years agoCURLOPT_FILE and CURLOPT_INFILE have better aliases now:
Daniel Stenberg [Mon, 6 May 2002 13:43:35 +0000 (13:43 +0000)]
CURLOPT_FILE and CURLOPT_INFILE have better aliases now:
CURLOPT_WRITEDATA and CURLOPT_READDATA

22 years agoAdded multi-post.c, based on the source file posted by Gustaf Hui
Daniel Stenberg [Mon, 6 May 2002 13:38:28 +0000 (13:38 +0000)]
Added multi-post.c, based on the source file posted by Gustaf Hui

22 years agoBetter support for being used with the multi interface without the *fd_set()
Daniel Stenberg [Sun, 5 May 2002 12:11:03 +0000 (12:11 +0000)]
Better support for being used with the multi interface without the *fd_set()
and proper select()ing have been made.

22 years ago--trace is a 7.9.7 function
Daniel Stenberg [Sun, 5 May 2002 09:09:17 +0000 (09:09 +0000)]
--trace is a 7.9.7 function

22 years agoadded -Z/--max-redirs and --trace
Daniel Stenberg [Sun, 5 May 2002 09:08:26 +0000 (09:08 +0000)]
added -Z/--max-redirs and --trace
changed order of a few others to keep the a-z order

22 years agomade VERBOSE output more like it used to be, HEADER_IN is thus also ignored
Daniel Stenberg [Sat, 4 May 2002 15:36:07 +0000 (15:36 +0000)]
made VERBOSE output more like it used to be, HEADER_IN is thus also ignored
by the internal debugfunction callback

22 years agoCURL_MAX_WRITE_SIZE is a new exported define that informs about the biggest
Daniel Stenberg [Fri, 3 May 2002 14:50:29 +0000 (14:50 +0000)]
CURL_MAX_WRITE_SIZE is a new exported define that informs about the biggest
sized buffer that may be passed to a write callback

22 years agofixed return code
Daniel Stenberg [Fri, 3 May 2002 12:40:37 +0000 (12:40 +0000)]
fixed return code

22 years agoanother week of changes
Daniel Stenberg [Fri, 3 May 2002 12:21:07 +0000 (12:21 +0000)]
another week of changes

22 years ago- Added "--trace [file]" to the command line tool. It makes a very detailed
Daniel Stenberg [Fri, 3 May 2002 12:14:09 +0000 (12:14 +0000)]
- Added "--trace [file]" to the command line tool. It makes a very detailed
  trace dump get stored, with a full protocol dump that includes all received
  and transmitted data. This could be a very effective tool for debugging what
  goes wrong. This dump includes every byte the way it is sent to/received
  from the server. The dump is the plain-text version, so SSL transfers will
  still be readable.

22 years agomade the DEBUGFUNCTION get called properly on a few more places, especially
Daniel Stenberg [Fri, 3 May 2002 12:07:32 +0000 (12:07 +0000)]
made the DEBUGFUNCTION get called properly on a few more places, especially
for DATA_IN and DATA_OUT.

22 years agomake sure our own printf() clones are used
Daniel Stenberg [Fri, 3 May 2002 12:06:04 +0000 (12:06 +0000)]
make sure our own printf() clones are used

22 years ago-D now stores all headers to the same file if multiple URLs are given on the
Daniel Stenberg [Fri, 3 May 2002 09:47:25 +0000 (09:47 +0000)]
-D now stores all headers to the same file if multiple URLs are given on the
command line!

22 years agoclarified that you must keep the variables that you point to!
Daniel Stenberg [Thu, 2 May 2002 22:34:31 +0000 (22:34 +0000)]
clarified that you must keep the variables that you point to!

22 years agospell
Daniel Stenberg [Thu, 2 May 2002 22:15:21 +0000 (22:15 +0000)]
spell

22 years agouse and set the fd_set pointers instead of the actual values, as then we
Daniel Stenberg [Thu, 2 May 2002 22:14:31 +0000 (22:14 +0000)]
use and set the fd_set pointers instead of the actual values, as then we
work properly with the multi interface when the user has provided the
fd_sets!

22 years agoadded two pointers to the fd_set variables to read/write from, as sometimes
Daniel Stenberg [Thu, 2 May 2002 22:13:35 +0000 (22:13 +0000)]
added two pointers to the fd_set variables to read/write from, as sometimes
we need to point to user-provided fd_sets

22 years agoreturn CURLM_CALL_MULTI_PERFORM in one more case, and check return code
Daniel Stenberg [Thu, 2 May 2002 22:12:14 +0000 (22:12 +0000)]
return CURLM_CALL_MULTI_PERFORM in one more case, and check return code
from malloc()

22 years agomake sure the dns cache pointers in the easy handles are NULLed
Daniel Stenberg [Thu, 2 May 2002 18:07:38 +0000 (18:07 +0000)]
make sure the dns cache pointers in the easy handles are NULLed

22 years agoHanno Kranzhoff noticed we didn't properly reset the download/upload counters
Daniel Stenberg [Thu, 2 May 2002 08:52:09 +0000 (08:52 +0000)]
Hanno Kranzhoff noticed we didn't properly reset the download/upload counters
before transfers, when doing multiple ones on the same handle.

22 years agoJacky Lam's fix to make the realloc() of the hostent data work properly
Daniel Stenberg [Wed, 1 May 2002 11:36:13 +0000 (11:36 +0000)]
Jacky Lam's fix to make the realloc() of the hostent data work properly
even when the realloc() actually gets a new memory block

22 years agohow to set more than one cookie
Daniel Stenberg [Tue, 30 Apr 2002 07:37:02 +0000 (07:37 +0000)]
how to set more than one cookie

22 years agoimplemented curl_multi_info_read() which I had forgotten before!
Daniel Stenberg [Sat, 27 Apr 2002 22:21:51 +0000 (22:21 +0000)]
implemented curl_multi_info_read() which I had forgotten before!

22 years agoupdated some docs
Daniel Stenberg [Sat, 27 Apr 2002 20:19:55 +0000 (20:19 +0000)]
updated some docs

22 years agoclarified
Daniel Stenberg [Sat, 27 Apr 2002 18:31:49 +0000 (18:31 +0000)]
clarified

22 years agoone added, one edited
Daniel Stenberg [Sat, 27 Apr 2002 18:12:26 +0000 (18:12 +0000)]
one added, one edited

22 years agoshorter introduction
Daniel Stenberg [Sat, 27 Apr 2002 18:01:54 +0000 (18:01 +0000)]
shorter introduction

22 years agoupdated
Daniel Stenberg [Sat, 27 Apr 2002 18:00:10 +0000 (18:00 +0000)]
updated

22 years agoAdded BINDINGS, lists all available libcurl bindings as of this date.
Daniel Stenberg [Sat, 27 Apr 2002 17:02:38 +0000 (17:02 +0000)]
Added BINDINGS, lists all available libcurl bindings as of this date.

22 years agofixes on rainy saturday in the end of April, 2002
Daniel Stenberg [Sat, 27 Apr 2002 13:26:32 +0000 (13:26 +0000)]
fixes on rainy saturday in the end of April, 2002

22 years agoFixed the FTP response reader to deal with timeouts better. Previously it
Daniel Stenberg [Sat, 27 Apr 2002 13:24:06 +0000 (13:24 +0000)]
Fixed the FTP response reader to deal with timeouts better. Previously it
would reset the timeout for each incoming data, which would make veeery
slow responses be allowed to take even more time since the timeout would only
be reached if the time between two received data chunks was longer than the
set timeout value...

22 years ago(Curl_GetFTPResponse) moved some code to only get performed when actually
Daniel Stenberg [Sat, 27 Apr 2002 13:09:37 +0000 (13:09 +0000)]
(Curl_GetFTPResponse) moved some code to only get performed when actually
needed

22 years agoNow uses Curl_ as prefix for internal global symbols. curl_ should only be
Daniel Stenberg [Sat, 27 Apr 2002 13:07:51 +0000 (13:07 +0000)]
Now uses Curl_ as prefix for internal global symbols. curl_ should only be
used for "exported" globals.

22 years agoDNC cache prune crash removed, made the name spacing follow the general
Daniel Stenberg [Sat, 27 Apr 2002 13:06:40 +0000 (13:06 +0000)]
DNC cache prune crash removed, made the name spacing follow the general
rule: "Curl_" prefix for library-wide private symbols, "curl_" is for
exported symbols.

22 years agoopenbsd on Alpha was no problemos
Daniel Stenberg [Fri, 26 Apr 2002 12:58:07 +0000 (12:58 +0000)]
openbsd on Alpha was no problemos

22 years agotest suite portability fix
Daniel Stenberg [Fri, 26 Apr 2002 12:57:49 +0000 (12:57 +0000)]
test suite portability fix

22 years agohandles much larger POSTs, replaced snprintf() with sprintf() since this
Daniel Stenberg [Fri, 26 Apr 2002 07:48:05 +0000 (07:48 +0000)]
handles much larger POSTs, replaced snprintf() with sprintf() since this
needs to be more portable and in the test server we can skip the extra
safety

22 years agoanother week of changes
Daniel Stenberg [Fri, 26 Apr 2002 07:45:39 +0000 (07:45 +0000)]
another week of changes

22 years agoadjusted to the new Curl_resolv() proto
Daniel Stenberg [Thu, 25 Apr 2002 19:26:29 +0000 (19:26 +0000)]
adjusted to the new Curl_resolv() proto

22 years agoprevent persistant connections to do name resolves
Daniel Stenberg [Thu, 25 Apr 2002 19:18:19 +0000 (19:18 +0000)]
prevent persistant connections to do name resolves

22 years agoIn order to not get problems with DNS cache pruning, we no longer store
Daniel Stenberg [Thu, 25 Apr 2002 19:00:57 +0000 (19:00 +0000)]
In order to not get problems with DNS cache pruning, we no longer store
any name resolved data in any curl handle struct. That way, we won't mind
if the cache entries are pruned for the next time we need them. We'll just
resolve them again instead.

This changes the Curl_resolv() proto. It modifies the SessionHandle struct
but perhaps most importantly, it'll make the internals somewhat dependent
on the DNS cache not being disabled as that will cripple operations somewhat.
Especially for persistant connections.

22 years agothe TOTAL_TIME is not really total, as it excludes the connect time...
Daniel Stenberg [Thu, 25 Apr 2002 16:45:15 +0000 (16:45 +0000)]
the TOTAL_TIME is not really total, as it excludes the connect time...

22 years agouse the hostname pointer instead of using IPv4-only info from a struct
Daniel Stenberg [Thu, 25 Apr 2002 15:50:49 +0000 (15:50 +0000)]
use the hostname pointer instead of using IPv4-only info from a struct

22 years agoclarified a bit with the timecondition stuff
Daniel Stenberg [Wed, 24 Apr 2002 10:16:46 +0000 (10:16 +0000)]
clarified a bit with the timecondition stuff

22 years agono, this doesn't look like it bugs
Daniel Stenberg [Wed, 24 Apr 2002 10:16:00 +0000 (10:16 +0000)]
no, this doesn't look like it bugs

22 years agoNetBSD alpha works, also tried a newer FreeBSD on alpha
Daniel Stenberg [Wed, 24 Apr 2002 10:15:37 +0000 (10:15 +0000)]
NetBSD alpha works, also tried a newer FreeBSD on alpha

22 years agoremoved warning about signed/unsigned comparison
Daniel Stenberg [Tue, 23 Apr 2002 14:57:37 +0000 (14:57 +0000)]
removed warning about signed/unsigned comparison

22 years agothe 80 column police narrowed this source code! B-]
Daniel Stenberg [Tue, 23 Apr 2002 14:56:21 +0000 (14:56 +0000)]
the 80 column police narrowed this source code! B-]

22 years agofixes bug report #547484, no_proxy doesn't properly strip off port numbers
Daniel Stenberg [Tue, 23 Apr 2002 13:34:28 +0000 (13:34 +0000)]
fixes bug report #547484, no_proxy doesn't properly strip off port numbers
from the host names before comparing

22 years agouse new timecond defines
Daniel Stenberg [Tue, 23 Apr 2002 00:05:21 +0000 (00:05 +0000)]
use new timecond defines

22 years agoallow binding the local end of a connection even when using IPv6, thus we
Daniel Stenberg [Mon, 22 Apr 2002 23:56:13 +0000 (23:56 +0000)]
allow binding the local end of a connection even when using IPv6, thus we
now have --interface working properly

22 years agouse double where it is supposed to
Daniel Stenberg [Mon, 22 Apr 2002 23:54:43 +0000 (23:54 +0000)]
use double where it is supposed to

22 years agouse sclose() to close sockets
Daniel Stenberg [Mon, 22 Apr 2002 23:53:49 +0000 (23:53 +0000)]
use sclose() to close sockets

22 years agorenamed the TIMECOND defines to be CURL_ prefixed
Daniel Stenberg [Mon, 22 Apr 2002 23:53:15 +0000 (23:53 +0000)]
renamed the TIMECOND defines to be CURL_ prefixed

22 years agoThe timecond stuff now have CURL_ prefixes
Daniel Stenberg [Mon, 22 Apr 2002 13:33:56 +0000 (13:33 +0000)]
The timecond stuff now have CURL_ prefixes

22 years agoJacky Lam's adjust resolve-buffer size patch applied. Slightly edited
Daniel Stenberg [Mon, 22 Apr 2002 13:31:16 +0000 (13:31 +0000)]
Jacky Lam's adjust resolve-buffer size patch applied. Slightly edited
by Daniel.

22 years agopack_hostent does not handle 64 bit pointers correctly.
Daniel Stenberg [Fri, 19 Apr 2002 11:00:26 +0000 (11:00 +0000)]
pack_hostent does not handle 64 bit pointers correctly.
A Bjørn Reese patch.

22 years agoPrune old hostcache entries with each call...
Sterling Hughes [Wed, 17 Apr 2002 20:13:55 +0000 (20:13 +0000)]
Prune old hostcache entries with each call...
This can be optimized a tidbit, but this is a start.

22 years agoDirk Manske's two new infos
Daniel Stenberg [Wed, 17 Apr 2002 07:21:17 +0000 (07:21 +0000)]
Dirk Manske's two new infos

22 years agoDirk Manske's fix for HTTP response code 301, that now behaves more like
Daniel Stenberg [Wed, 17 Apr 2002 07:16:49 +0000 (07:16 +0000)]
Dirk Manske's fix for HTTP response code 301, that now behaves more like
browsers do. Which thus *breaks* the RFC 2616...

22 years agoDirk Manske brought the patch that introduces two new CURLINFO_* values:
Daniel Stenberg [Tue, 16 Apr 2002 07:59:20 +0000 (07:59 +0000)]
Dirk Manske brought the patch that introduces two new CURLINFO_* values:
CURLINFO_REDIRECT_TIME and CURLINFO_REDIRECT_COUNT.

22 years agoJonatan Lander fixed the "Disables POST, goes with GET" output to be more
Daniel Stenberg [Mon, 15 Apr 2002 13:47:06 +0000 (13:47 +0000)]
Jonatan Lander fixed the "Disables POST, goes with GET" output to be more
connected with reality! ;-)

22 years agoThis makes formposting with a specified file missing fail. curl_easy_perform
Daniel Stenberg [Mon, 15 Apr 2002 11:19:03 +0000 (11:19 +0000)]
This makes formposting with a specified file missing fail. curl_easy_perform
will then return CURLE_READ_ERROR.

22 years ago7.9.6 commit
Daniel Stenberg [Mon, 15 Apr 2002 06:58:04 +0000 (06:58 +0000)]
7.9.6 commit

22 years agoDirk Manske made libcurl strip off white spaces from the beginning of cookie
Daniel Stenberg [Sun, 14 Apr 2002 18:21:17 +0000 (18:21 +0000)]
Dirk Manske made libcurl strip off white spaces from the beginning of cookie
contents.

22 years agocheck for _MSC_VER as well as __STDC__ when deciding to use the ## operator
Daniel Stenberg [Sun, 14 Apr 2002 17:29:35 +0000 (17:29 +0000)]
check for _MSC_VER as well as __STDC__ when deciding to use the ## operator
or not

22 years agoremoved unused variable
Daniel Stenberg [Sun, 14 Apr 2002 07:33:44 +0000 (07:33 +0000)]
removed unused variable

22 years agoAdd protos and change return value of curl_hash_count....
Sterling Hughes [Sat, 13 Apr 2002 01:56:22 +0000 (01:56 +0000)]
Add protos and change return value of curl_hash_count....

22 years agoSpeed up the hash code considerably, removing a bunch of legacy crud
Sterling Hughes [Fri, 12 Apr 2002 23:40:19 +0000 (23:40 +0000)]
Speed up the hash code considerably, removing a bunch of legacy crud

22 years agocut off an old section
Daniel Stenberg [Fri, 12 Apr 2002 14:41:24 +0000 (14:41 +0000)]
cut off an old section

22 years agohttp_proxy must be lower case
Daniel Stenberg [Fri, 12 Apr 2002 14:33:57 +0000 (14:33 +0000)]
http_proxy must be lower case

22 years ago7.9.6-pre5
Daniel Stenberg [Fri, 12 Apr 2002 11:39:27 +0000 (11:39 +0000)]
7.9.6-pre5

22 years agoexplain all mailing lists better
Daniel Stenberg [Fri, 12 Apr 2002 10:33:40 +0000 (10:33 +0000)]
explain all mailing lists better

22 years agoadded libcurl-errors
Daniel Stenberg [Fri, 12 Apr 2002 10:10:36 +0000 (10:10 +0000)]
added libcurl-errors

22 years agostore TIMER_CONNECT even if the connect failed
Daniel Stenberg [Fri, 12 Apr 2002 10:03:59 +0000 (10:03 +0000)]
store TIMER_CONNECT even if the connect failed

22 years agocorrected the verbose output for connects and fixed the connect time stamp
Daniel Stenberg [Fri, 12 Apr 2002 08:18:38 +0000 (08:18 +0000)]
corrected the verbose output for connects and fixed the connect time stamp
better for FTP (any protocol with protocol-specific connect actions)

22 years agothe HTTP request is a CURLINFO_HEADER_OUT
Daniel Stenberg [Fri, 12 Apr 2002 07:53:12 +0000 (07:53 +0000)]
the HTTP request is a CURLINFO_HEADER_OUT

22 years agoJean-Philippe Barrette-LaPierre's patch applied
Daniel Stenberg [Fri, 12 Apr 2002 07:23:57 +0000 (07:23 +0000)]
Jean-Philippe Barrette-LaPierre's patch applied

22 years agoJean-Philippe Barrette-LaPierre provided his patch that introduces
Daniel Stenberg [Fri, 12 Apr 2002 07:21:11 +0000 (07:21 +0000)]
Jean-Philippe Barrette-LaPierre provided his patch that introduces
CURLOPT_DEBUGFUNCTION and CURLOPT_DEBUGDATA.

22 years agoerror codes *ARE* documented now
Daniel Stenberg [Fri, 12 Apr 2002 07:19:43 +0000 (07:19 +0000)]
error codes *ARE* documented now