aboutsummaryrefslogtreecommitdiff
path: root/tests/server
AgeCommit message (Collapse)Author
2005-02-22Curl_base64_decode() now returns an allocated bufferDaniel Stenberg
2004-12-14make sure the ipv6 http server gets its pid stored in a separate fileDaniel Stenberg
2004-12-13Fixed missing braces warning.Gisle Vanem
2004-12-12Missing 'in6addr_any' in MingW's lib. ld bug?Gisle Vanem
2004-12-11HTTP IPv6 support added to the test suiteDaniel Stenberg
2004-12-09ignore the getpart toolDaniel Stenberg
2004-12-09close the connection when a bad test number was requestedDaniel Stenberg
2004-12-02added comment about port number in CONNECT string being used as test numberDaniel Stenberg
2004-11-29removed commentDaniel Stenberg
2004-11-29include setup.h firstDaniel Stenberg
2004-11-29zero terminate the buffer spitout() returns, as the sws.c code depends onDaniel Stenberg
that!
2004-11-29Enable test cases to provide sections base64-encoded to be able to testDaniel Stenberg
with binary data.
2004-11-29utility to base encode data passed on stdinDaniel Stenberg
2004-10-07use curlx_strnequal() from the private lib sources instead of strncasecmp()Daniel Stenberg
for maximum portability
2004-06-22modified some logging outputDaniel Stenberg
2004-06-21when the client disconnects prematurely, dump the request as received thusDaniel Stenberg
far
2004-06-15skip the pid from the loggingDaniel Stenberg
2004-05-28delete trailing whitespaceDaniel Stenberg
2004-05-06%ld for longDaniel Stenberg
2004-05-06int/long fixDaniel Stenberg
2004-04-30include the full size of the sent response in the logDaniel Stenberg
2004-04-21log the WAIT commandDaniel Stenberg
2004-04-17Gisle Vanem: patches to make sws.c compile under MingW/MSVC isDaniel Stenberg
attached. And some cosmetic fixes.
2004-04-16remade the logging function to better deal with removed logfiles duringDaniel Stenberg
the execution of the tests
2004-03-31Added "swsbounce" magic: if this keyword is present in a <data> section itDaniel Stenberg
sets the "swsbounce" magic mode. If there follows a request for the SAME test number and the SAME part number, this mode will make the server bump the part number internally and thus return a different <dataNUM> section than it otherwise would. Test case 153 uses this in case you need an example. It is pretty involved and hard-to-use, but then the situation is pretty special over all. Enjoy.
2004-03-23minor variable type cleanupsDaniel Stenberg
2004-03-10use size_t for string lengthsDaniel Stenberg
2004-03-09Added -I$(top_srcdir)/include, since lib/setup.h might include files fromDaniel Stenberg
the external curl include dir.
2004-03-09Andrés García-fix to make it build with mingwDaniel Stenberg
2004-03-05Major rewrite of the test HTTP server to allow more fancy features to makeDaniel Stenberg
better tests with the issue12-patch applied. This change also includes Andrés García's win32-fixes. Made the logging look better/more readable in sws.log
2004-02-20fix protos to prevent warningsDaniel Stenberg
2004-02-20added getpart.h for the spitout() protoDaniel Stenberg
2004-02-13make the path constDaniel Stenberg
2004-02-12provide a source path to the servers to make them find the tests when runDaniel Stenberg
outside the source dir, not needing any symlinks
2004-01-30when we receive a request overflow, we still dump the incoming request toDaniel Stenberg
the dump file to make it easier to understand and debug the situation
2004-01-07updated year in the copyright stringDaniel Stenberg
2003-11-24new way to trick sws to return special data on CONNECTDaniel Stenberg
2003-10-30only use siginterrupt() if it really exists on the platformDaniel Stenberg
2003-10-09portability fix by using setup.h from the lib directoryDaniel Stenberg
2003-10-09not used anymoreDaniel Stenberg
2003-09-15make the NTLM part numbers get increased instead of plainly assigned, asDaniel Stenberg
this then makes redirection tests work (as the new test 89)
2003-09-03Peter Pentchev found two problems. One realloc problem that could allocateDaniel Stenberg
too little data, and one case of not zero-terminating the returned string. I chose a slightly different patch than the one Peter provided.
2003-08-08Gisle Vanem fixed a single-byte overflowDaniel Stenberg
2003-07-19fixed the CONNECT thing againDaniel Stenberg
2003-07-19If the data contents contains the word 'swsclose', then this server willDaniel Stenberg
disconnect the client after the response have been sent. This also happens if the respons is zero byte long. In all other cases (unless an error happens), it will now maintain the connection to allow proper persistant connection testing. This was required for the NTLM testing to work so I finally had to fix this. Of course most of the existing HTTP tests will be adjusted to work with this new rule of test file syntax for HTTP tests. Also fixed the log function to deal with varargs for better logging.
2003-07-01David Byron's patch that allows a client to make the server quit with aDaniel Stenberg
magic url.
2003-06-11basic NTLM supportDaniel Stenberg
2003-05-22Digest support addedDaniel Stenberg
2003-04-03Added log output for when the writing of the input HTTP request is successfulDaniel Stenberg
or unsuccessful. Used to track down the recent cygwin test suite problems.
2003-03-16typecase getpid() to int to prevent compiler warningDaniel Stenberg