summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Merge tag '2.46.0' into baserock/morphbaserock/morphMark Doffman2014-04-01269-12717/+29849
|\ | | | | | | 2.46.0
| * 2.46.02.46.0Dan Winship2014-03-242-2/+6
| |
| * 2.45.922.45.92Dan Winship2014-03-172-1/+23
| |
| * soup-socket: fixed leak in new_socket_client()Slava Monich2014-03-161-2/+7
| |
| * soup-session: force port for proxy URLSlava Monich2014-03-161-1/+1
| | | | | | | | | | | | Otherwise libsoup is unable to connect through proxies that use default port (say, 80 for http). Proxy resolver resolves port-less proxy URLs into network address with zero port, which is then passed to connect()
| * tests: use g_test_bug() to annotate test casesDan Winship2014-03-1525-293/+485
| |
| * tests: remove debug_printf()s that are redundant with test namesDan Winship2014-03-1522-246/+13
| |
| * tests: split up some test programs into more testsDan Winship2014-03-1510-624/+831
| |
| * auth-test: remove some erroneous SKIP_IF_NO_APACHEsDan Winship2014-03-151-6/+0
| |
| * redirect-test: remove a flaky testDan Winship2014-03-101-103/+0
| | | | | | | | | | | | | | | | Bug 651446 involved a race condition when queuing a message while another message was being redirected on the same host. The fix included a regression test, but later rewrites of SoupSession made the race condition no longer possible, and made the test itself slightly flaky. So just remove it.
| * xmlrpc-test: use g_test_skip()Dan Winship2014-03-091-4/+1
| | | | | | | | | | | | There was a comment about not using g_test_skip() here because of an old glib bug, but we're using g_test_skip() in lots of other places now, so...
| * test-utils: change http_debug command-line flagDan Winship2014-03-091-1/+1
| | | | | | | | g_test_init() intercepts "-h" for "help", so make http-debug be "-H"
| * configure.ac: belatedly bump glib requirement to 2.38Dan Winship2014-03-091-2/+2
| | | | | | | | | | The tests use the TAP driver now, which only exists in glib 2.38 and later.
| * Move to the platform categoryAlexandre Franke2014-02-231-0/+1
| |
| * 2.45.902.45.90Dan Winship2014-02-172-1/+32
| |
| * tests: skip individual tests rather than whole test programsDan Winship2014-02-1715-166/+144
| | | | | | | | | | | | | | | | | | | | | | Use g_skip_test() to skip individual tests rather than just returning status 77 from the test program as a whole. In several cases, we still end up skipping more than necessary, due to test cases that need to be split up more. Remove the "MISSING_REGRESSION_TESTS_PACKAGES" functionality, since the skipped tests are now pointed out explicitly.
| * Merge branch 'content-sniffing-update'Dan Winship2014-02-1715-159/+605
| |\ | | | | | | | | | | | | | | | | | | | | | Update SoupContentSniffer (and sniffing-test) to match the current version of the MIME sniffing spec. https://bugzilla.gnome.org/show_bug.cgi?id=648849 https://bugzilla.gnome.org/show_bug.cgi?id=715126
| | * sniffing: Add audio/video sniffingGustavo Noronha Silva2014-02-178-0/+135
| | |
| | * sniffing: Bring image sniffing up-to-date with the MIMESNIFF specGustavo Noronha Silva2014-02-176-99/+127
| | |
| | * sniffing: Bring feed vs HTML up-to-date with the MIMESNIFF specGustavo Noronha Silva2014-02-174-43/+120
| | | | | | | | | | | | | | | | | | | | | * decide on that before doing the image sniffing to match the spec * use const char* and g_str_has_prefix for comparisons to make it more legible * deal with rdf:RDF tags
| | * sniffing: Add comment to the XML types bitGustavo Noronha Silva2014-02-171-0/+1
| | |
| | * sniffing: test that it allows leading whitespace in doctype headersAlberto Garcia2014-02-173-0/+16
| | | | | | | | | | | | https://bugzilla.gnome.org/show_bug.cgi?id=715126
| | * sniffing: Adjust the general unknown MIME type algorithmGustavo Noronha Silva2014-02-172-55/+187
| | | | | | | | | | | | | | | | | | This change adjusts the pattern matching table to the current form of the MIMESNIFF spec, adding a check for a tag-terminating byte and using the formally defined sniff-scriptable flag.
| | * sniffing: Implement handling of the X-Content-Type-Options headerGustavo Noronha Silva2014-02-172-1/+24
| | |
| | * sniffing: Implement the check-apache-bug flagGustavo Noronha Silva2014-02-174-10/+40
| | | | | | | | | | | | | | | | | | Run the text or binary algorithm when some specific text/plain Content-Types are provided, since older versions of apache would send that type for binary files. http://mimesniff.spec.whatwg.org/#dfnReturnLink-0
| | * sniffing: Document the specification we are targetingGustavo Noronha Silva2014-02-171-0/+4
| |/
| * SoupConnection: Explicitly set SoupAddress protocol.Matthew Barnes2014-02-171-3/+16
| | | | | | | | | | | | | | | | | | So GProxyResolver can distinguish between "http" and "https" requests, and choose the appropriate proxy settings. Previously the SoupAddress protocol was left unset, so soup_address_connectable_proxy_enumerate() assumed "http". https://bugzilla.gnome.org/show_bug.cgi?id=724316
| * tests: use TAP, and install the testsDan Winship2014-02-0818-133/+1014
| | | | | | | | | | | | | | | | | | Import glib-tap.mk to run the tests in TAP mode, and to provide rules to install the tests. This required some restructuring of httpd.conf-related stuff in order to come up with a configuration that would work for all three cases (srcdir==buildir, srcdir!=builddir, and installed).
| * tests: Add resources/* to soup-test.gresource, add utility functionsDan Winship2014-02-0812-217/+174
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Add the files in resources/ to soup-test.gresource, and add soup_test_load_resource() to get the contents of one of them, and then use this in places that were previously reading them by hand. Except forms-test, which needs to pass a filename to curl, so have it use index.txt instead of one of the resource files. Now none of the tests access the resources/ directory at runtime. Also add soup_test_get_index() to read index.txt rather than reimplementing it in multiple test programs.
| * tests: initial port to the gtestutils frameworkDan Winship2014-02-0836-3849/+2392
| | | | | | | | | | Some programs need to be split up into more tests, and the debug output is mostly not updated for the new format.
| * updated kn.poShankar Prasad2014-02-051-14/+13
| |
| * tests: headers: allow checks on non present headers fieldsLionel Landwerlin2014-01-191-2/+23
| | | | | | | | https://bugzilla.gnome.org/show_bug.cgi?id=722341
| * tests: headers-parsing: allow checks with case insensitive headers fieldsLionel Landwerlin2014-01-191-1/+1
| | | | | | | | https://bugzilla.gnome.org/show_bug.cgi?id=722341
| * soup_uri_to_string: document that passwords are never outputDan Winship2014-01-181-0/+3
| |
| * tests: ignore proxy settingsDan Winship2014-01-073-0/+7
| | | | | | | | | | | | | | | | | | | | Some of the tests depend in various ways on the fact that they are not going through a proxy, so make sure they really aren't (by forcing the use of the dummy GProxyResolver). Likewise, pass "--noproxy *" to curl in the tests that use it, to make sure it doesn't pick up $http_proxy, etc. Pointed out by Dirkjan Ochtman on IRC.
| * examples: fix simple-proxyDan Winship2014-01-071-1/+1
| | | | | | | | startup error checking was busted, making this not even work
| * soup-tld: accept ASCII-encoded hostnames tooDan Winship2013-12-182-11/+82
| | | | | | | | | | | | | | | | It's inconvenient for callers to have to ensure that the hostname they pass in is UTF-8 (since they themselves may not need to care). So accept ASCII-encoded hostnames too, and add the corresponding punycode tests from the publicsuffix.org test list that we were ignoring before.
| * 2.45.32.45.3Dan Winship2013-12-162-1/+58
| |
| * Update the public suffix list (and its test)Dan Winship2013-12-112-152/+770
| |
| * Misc docs updatesDan Winship2013-12-1114-65/+84
| |
| * docs: Update client-side tutorial for 2.42, document request APIsDan Winship2013-12-115-202/+391
| |
| * SoupSocket: set TCP_NODELAY on socketsDan Winship2013-12-101-0/+1
| | | | | | | | | | | | | | The old SoupSocket did this but it accidentally got removed in the GSocket port because I thought it didn't matter. It does. Noticed by Andy Wingo.
| * fix up non-literal format string issuesRyan Lortie2013-12-092-6/+11
| | | | | | | | | | | | | | Fix up a couple of errors picked up by -Werror=format-nonliteral on clang. https://bugzilla.gnome.org/show_bug.cgi?id=720082
| * no-ssl-test: fix indentationDan Winship2013-12-081-151/+151
| |
| * soup-version: add 2.46Dan Winship2013-12-081-0/+15
| |
| * Update Chinese simplified translation甘露(Gan Lu)2013-12-071-4/+4
| |
| * Update Chinese simplified translation甘露(Gan Lu)2013-12-071-21/+46
| |
| * no-ssl-test: fix for GDummyTlsBackend changesDan Winship2013-11-201-5/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | This was assuming that setting "ssl-use-system-ca-file" would fail when using the dummy TLS backend, but the fact that it failed was just a side effect of the fact that GDummyTlsBackend returned a different GTlsDatabase every time you called g_tls_backend_get_default_database(). Now that that's fixed, setting "ssl-use-system-ca-file" does appear to work. Of course, SSL won't work at all with the dummy backend, so it doesn't really matter whether setting this property works or not, so just remove that part of the test, so this will pass with either old or new glib.
| * SoupMessage: add :request-body-data and :response-body-data propertiesDan Winship2013-11-172-15/+74
| | | | | | | | | | | | | | Make it easier to use the request and response bodies from introspection by providing accessors to get them as GBytes. https://bugzilla.gnome.org/show_bug.cgi?id=704105
| * session-test: fix to work with the dummy TLS backendDan Winship2013-11-171-8/+13
| | | | | | | | | | | | do_property_tests() apparently didn't work with the dummy TLS backend, half because of a glib bug, and half because of a session-test bug. This fixes the session-test bug; the glib bug is fixed in glib master.
View Lorry Controller Status