summaryrefslogtreecommitdiff
path: root/lib
Commit message (Collapse)AuthorAgeFilesLines
* when a chunked error is noticed, store the error number in the error stringDaniel Stenberg2003-01-211-1/+1
| | | | to enable better error-tracking
* Markus F.X.J. Oberhumer's patch that reduces memory usage quite a bit byDaniel Stenberg2003-01-203-1/+10
| | | | | only allocating the scratch memory buffer once it is needed and not always in the handle.
* given passwords in netrc must be respected accordinglyDaniel Stenberg2003-01-201-5/+8
|
* reverted bad header replacementDaniel Stenberg2003-01-161-18/+274
|
* copyright year update in the source headerDaniel Stenberg2003-01-1671-344/+88
|
* Allow CURLINFO_PRIVATE to be NULL, patch by Markus OberhumerDaniel Stenberg2003-01-161-1/+1
|
* no TABs in source codeDaniel Stenberg2003-01-151-9/+9
|
* removed a TABDaniel Stenberg2003-01-151-1/+1
|
* Steve Oliphant pointed out that test case 105 did not work anymore and thisDaniel Stenberg2003-01-101-0/+3
| | | | was due to a missing fix for the password prompting
* if userpwd is "username:", this now implies a blank password while onlyDaniel Stenberg2003-01-092-5/+13
| | | | | | "username" will cause libcurl to prompt for password. Bryan Kemp noticed. test case 136 is added for this
* This fixed yet another connect problem with the multi interface and ipv4Daniel Stenberg2003-01-091-2/+10
| | | | stack. Kjetil Jacobsen reported and verified the fix.
* removed unused codeDaniel Stenberg2003-01-091-74/+2
|
* proper indentDaniel Stenberg2003-01-091-44/+46
|
* share.h is now a used header fileDaniel Stenberg2003-01-091-1/+1
|
* Updated more and now looks and and the API possibly works almost like theDaniel Stenberg2003-01-092-107/+81
| | | | design document specifies. There is still no code inside that uses this.
* updated to use the modified share-typesDaniel Stenberg2003-01-084-6/+18
|
* fixed a very, very rare and very, very little memory leakJean-Philippe Barette-LaPierre2003-01-081-3/+12
|
* Philippe Raoult's fix to handle wildcard certificate name checksDaniel Stenberg2003-01-071-1/+39
|
* Simon Liu's HTTP200ALIASES-patch!Daniel Stenberg2003-01-073-2/+45
|
* Only output valid filetime.Daniel Stenberg2003-01-071-16/+24
| | | | Return file-error if 550 is returned when trying MDTM
* when sending an error message to the debugfunction, we append a newline soDaniel Stenberg2003-01-071-3/+13
| | | | that the output looks better
* Matthew Blain's improvements for debug buildsDaniel Stenberg2003-01-071-2/+3
|
* indent fixDaniel Stenberg2003-01-061-2/+2
|
* fix bug (?) :-)Sterling Hughes2003-01-061-6/+8
| | | | | | previously, if you called curl_easy_perform and then set the global dns cache, the global cache wouldn't be used. I don't see this really happening in practice, but this code allows you to do it.
* return -1 even if SSL_pending() doesn't return non-zero, as we don't reallyDaniel Stenberg2002-12-291-4/+2
| | | | | care how many bytes that is readable NOW. Philippe Raoult reported the bug in 7.10.3-pre3.
* Curl_base64_decode() fixed by Matthew BDaniel Stenberg2002-12-191-2/+6
|
* Fixed the usage of SSL_read() to properly return -1 if the EWOULDBLOCKDaniel Stenberg2002-12-191-13/+11
| | | | | | | | | situation occurs, which it previously didn't! This was reptoed by Evan Jordan in bug report #653022. Also, if ERROR_SYSCALL is returned from SSL_write(), include the errno number in the error string for easier error detection.
* CURLE_HTTP_NOT_FOUND => CURLE_HTTP_RETURNED_ERRORDaniel Stenberg2002-12-181-1/+1
|
* Removed weird special multi interface condition that caused bug reportDaniel Stenberg2002-12-171-6/+0
| | | | #651464.
* EAGAIN on older (correct) glibc versions indicate a problem and not the needDaniel Stenberg2002-12-161-4/+16
| | | | | | | | | for a bigger buffer and this is indeed badness for us. Making this work on both old and new glibc versions require an ugly loop that in its worse form cause 45 bad loops when using the correct glibc and a non-resolving host name... :-/ We want a better fix. Badly.
* conn->bits.tcpconnect now keeps track of if this connection is connectedDaniel Stenberg2002-12-133-3/+20
| | | | or not
* Evan Jordan's fix for a memory leak. Bug report 650989.Daniel Stenberg2002-12-131-4/+9
|
* make a little work-around for file:// in _is_connected() and voila, now theDaniel Stenberg2002-12-131-0/+5
| | | | | multi interface works with file:// URLs fine (previously it crashed). This won't make it work on Windows though...
* The initial HTTP request can now be sent in multiple parts, as part of theDaniel Stenberg2002-12-103-76/+171
| | | | | | regular transfer process. This required some new tweaks, like for example we need to be able to tell the tranfer loop to not chunky-encode uploads while we're transferring the rest of the request...
* send_buffer is no more hereDaniel Stenberg2002-12-101-7/+0
|
* A normal POST now provides data to the main transfer loop via the usualDaniel Stenberg2002-12-092-15/+45
| | | | | read callback, and thus won't put a lot of stress on the request sending code (which currently does an ugly loop).
* The fread() callback pointer and associated pointer is now stored in theDaniel Stenberg2002-12-096-237/+241
| | | | | connectdata struct instead, and is no longer modified within the 'set' struct as previously (which was a really BAAAD thing).
* read and write as much as possible until end of data or EWOULDBLOCK beforeDaniel Stenberg2002-12-051-481/+504
| | | | returning back to the select() loop. Consider this a test so far.
* changed proto for Curl_krb_kauth()Daniel Stenberg2002-12-051-1/+1
|
* make WIN32 defined for Borland properly, as told by Alexander J. OssDaniel Stenberg2002-12-041-3/+2
|
* The waiting for the 226 or 250 line expected to come after a transfer isDaniel Stenberg2002-12-041-10/+20
| | | | | | | complete is now only made for 60 seconds and if no data was received during those 60 seconds, we store a special error message (preparing to make this a special error code) as this most likely means that the control connection has died while we were transferring data.
* Curl_GetFTPResponse() takes a different set of parameters and now return aDaniel Stenberg2002-12-035-123/+126
| | | | | proper CURLcode. The default timeout for reading one response is now also possible to change while running.
* updated to realityDaniel Stenberg2002-12-032-2/+10
|
* if the PWD reply parser failed, we leaked memoryDaniel Stenberg2002-12-021-1/+3
|
* more to ignoreDaniel Stenberg2002-12-011-1/+2
|
* only use Content-Length: header if not transfering data chunkedDaniel Stenberg2002-12-011-15/+28
|
* let the Curl_FormReader() return 0 when it reaches end of data to that theDaniel Stenberg2002-11-291-1/+1
| | | | chunked transfer work
* fix the hash init to call the correct dns cleanup functionDaniel Stenberg2002-11-281-3/+2
|
* added compareheader protoDaniel Stenberg2002-11-281-0/+4
|
* compareheader() was moved over to http.c and got a Curl_ prefixDaniel Stenberg2002-11-281-61/+7
| | | | | The chunked transfer upload never stopped due to a silly add before we checked for >0!
View Lorry Controller Status