| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
See the [Meson website] for a full reference. To build, test and/or
install the librest library, you essentially need the following commands:
```
$ meson build
$ ninja -C build
$ meson test -C build
$ ninja -C build install
```
For the GNOME initiative, see
https://wiki.gnome.org/Initiatives/GnomeGoals/MesonPorting
[Meson website]: http://mesonbuild.com/
https://gitlab.gnome.org/GNOME/librest/merge_requests/1
|
|
|
|
| |
Most of them are taken from the Vala overrides: https://gitlab.gnome.org/GNOME/vala/-/blob/master/vapi/metadata/Rest-0.7.metadata
|
|
|
|
|
|
|
|
|
| |
Calling rest_xml_parser_parse_from_data() with an empty string ("")
currently causes a crash as xmlReaderForMemory() returns NULL in that
case, and we then try to dereference this pointer without checking it's
non-NULL.
https://bugzilla.gnome.org/show_bug.cgi?id=789053
|
|
|
|
|
| |
This is already done in finish_call and the g_strdup to
priv->status_message was leaking the previously set one.
|
|
|
|
|
| |
Before this, it was impossible to get the server payload from a proxy
call invoked via rest_proxy_call_upload.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The order in which GHashTable returns its key-value pairs is undefined.
Therefore the output of rest_xml_node_print can change based on the
GHashTable implementation. While not strictly necessary, it would be
nice to avoid that. Having a stable order, even if it is not
documented and depends on the current RestXmlNode code, is handy for
testing.
This was the main reason behind the tests/xml.c breakage.
https://bugzilla.gnome.org/show_bug.cgi?id=788960
|
|
|
|
|
|
|
|
|
|
| |
The rest_xml_node_add_child API explicitly documents NULL parents as
the way to create top-level nodes. This had broken the second half of
tests/xml.c.
Fallout from b11a1664cd4dfcc62b07b4a68adc220fd1eb8305
https://bugzilla.gnome.org/show_bug.cgi?id=788960
|
|
|
|
|
|
|
|
| |
This had broken tests/xml.c.
Fallout from 61a7b231bd8b9d1b8d02dca120389e79d38b428d
https://bugzilla.gnome.org/show_bug.cgi?id=788960
|
|
|
|
| |
https://bugzilla.gnome.org/show_bug.cgi?id=785818
|
|
|
|
| |
https://bugzilla.gnome.org/show_bug.cgi?id=785818
|
|
|
|
| |
https://bugzilla.gnome.org/show_bug.cgi?id=785811
|
|
|
|
| |
https://bugzilla.gnome.org/show_bug.cgi?id=785811
|
|
|
|
| |
https://bugzilla.gnome.org/show_bug.cgi?id=785811
|
|
|
|
| |
https://bugzilla.gnome.org/show_bug.cgi?id=785811
|
|
|
|
|
|
| |
Bump minimum GLib version to 2.44.0.
https://bugzilla.gnome.org/show_bug.cgi?id=785811
|
| |
|
| |
|
|
|
|
|
|
|
| |
Currently rest_proxy_call_sync (and everything else using
prepare_message) might return FALSE without setting @error at all. Add
two new error codes to indicate the possible problems and reliably
set the error if something went wrong.
|
| |
|
|
|
|
|
|
|
| |
It has been deprecated for a long while, and we were only using it for
proxy support. However, now that we switched to using SoupSession rather
than SoupSession{Sync,Async}, a default proxy resolver will be used, so
we don't need to set it explicitly.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This fixes:
==16342== 93 (64 direct, 29 indirect) bytes in 1 blocks are definitely lost in loss record 211 of 278
==16342== at 0x4C2BBAD: malloc (vg_replace_malloc.c:299)
==16342== by 0x5F5CE58: g_malloc (gmem.c:94)
==16342== by 0x5F73D42: g_slice_alloc (gslice.c:1025)
==16342== by 0x5F7436D: g_slice_alloc0 (gslice.c:1051)
==16342== by 0x56DA77A: soup_uri_new_with_base (soup-uri.c:283)
==16342== by 0x56DAA42: soup_uri_new (soup-uri.c:519)
==16342== by 0x4E48BB5: oauth2_proxy_extract_access_token (oauth2-proxy.c:384)
==16342== by 0x4009E6: test_url_fragment_no_access_token (oauth2.c:36)
==16342== by 0x5F7C983: test_case_run (gtestutils.c:2158)
==16342== by 0x5F7C983: g_test_run_suite_internal (gtestutils.c:2241)
==16342== by 0x5F7CB4E: g_test_run_suite_internal (gtestutils.c:2253)
==16342== by 0x5F7CD5D: g_test_run_suite (gtestutils.c:2328)
==16342== by 0x5F7CD80: g_test_run (gtestutils.c:1596)
==16342== by 0x400B64: main (oauth2.c:65)
==16342==
==16342== 133 (64 direct, 69 indirect) bytes in 1 blocks are definitely lost in loss record 262 of 278
==16342== at 0x4C2BBAD: malloc (vg_replace_malloc.c:299)
==16342== by 0x5F5CE58: g_malloc (gmem.c:94)
==16342== by 0x5F73D42: g_slice_alloc (gslice.c:1025)
==16342== by 0x5F7436D: g_slice_alloc0 (gslice.c:1051)
==16342== by 0x56DA77A: soup_uri_new_with_base (soup-uri.c:283)
==16342== by 0x56DAA42: soup_uri_new (soup-uri.c:519)
==16342== by 0x4E48BB5: oauth2_proxy_extract_access_token (oauth2-proxy.c:384)
==16342== by 0x400A25: test_access_token_simple (oauth2.c:44)
==16342== by 0x5F7C983: test_case_run (gtestutils.c:2158)
==16342== by 0x5F7C983: g_test_run_suite_internal (gtestutils.c:2241)
==16342== by 0x5F7CB4E: g_test_run_suite_internal (gtestutils.c:2253)
==16342== by 0x5F7CD5D: g_test_run_suite (gtestutils.c:2328)
==16342== by 0x5F7CD80: g_test_run (gtestutils.c:1596)
==16342== by 0x400B64: main (oauth2.c:65)
==16342==
==16342== 137 (64 direct, 73 indirect) bytes in 1 blocks are definitely lost in loss record 263 of 278
==16342== at 0x4C2BBAD: malloc (vg_replace_malloc.c:299)
==16342== by 0x5F5CE58: g_malloc (gmem.c:94)
==16342== by 0x5F73D42: g_slice_alloc (gslice.c:1025)
==16342== by 0x5F7436D: g_slice_alloc0 (gslice.c:1051)
==16342== by 0x56DA77A: soup_uri_new_with_base (soup-uri.c:283)
==16342== by 0x56DAA42: soup_uri_new (soup-uri.c:519)
==16342== by 0x4E48BB5: oauth2_proxy_extract_access_token (oauth2-proxy.c:384)
==16342== by 0x400A97: test_url_encoding_access_token (oauth2.c:51)
==16342== by 0x5F7C983: test_case_run (gtestutils.c:2158)
==16342== by 0x5F7C983: g_test_run_suite_internal (gtestutils.c:2241)
==16342== by 0x5F7CB4E: g_test_run_suite_internal (gtestutils.c:2253)
==16342== by 0x5F7CD5D: g_test_run_suite (gtestutils.c:2328)
==16342== by 0x5F7CD80: g_test_run (gtestutils.c:1596)
==16342== by 0x400B64: main (oauth2.c:65)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The current code is using xml = g_strconcat (xml, ...) which is causing
some leaks as g_strconcat returns a newly allocated string. Using
GString avoids this issue without constantly freeing the intermediate
strings.
This fixes multiple leaks like:
==16611== 18 bytes in 1 blocks are definitely lost in loss record 124 of 301
==16611== at 0x4C2BBAD: malloc (vg_replace_malloc.c:299)
==16611== by 0x5F5CE58: g_malloc (gmem.c:94)
==16611== by 0x5F75B8E: g_strconcat (gstrfuncs.c:585)
==16611== by 0x4E450CF: rest_xml_node_print (rest-xml-node.c:287)
==16611== by 0x4E451DA: rest_xml_node_print (rest-xml-node.c:305)
==16611== by 0x4E450F8: rest_xml_node_print (rest-xml-node.c:292)
==16611== by 0x4009A0: main (xml.c:40)
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
soup_form_request_new_from_hash can return NULL, e.g. if the given url
is malformed.
|
|
|
|
|
| |
priv->session was a SoupSession since the last commit, which also
removed every practical usage or priv->session_sync.
|
| |
|
|
|
|
| |
Based on initial work from Timm Bäder <mail@baedert.org>
|
| |
|
| |
|
|
|
|
| |
Where the checked version is < 2.36, which we now require
|
| |
|
| |
|
|
|
|
| |
An equivalent _parse_token_response exists that does not contain a typo.
|
|
|
|
| |
https://bugzilla.gnome.org/show_bug.cgi?id=658725
|
|
|
|
|
|
|
|
| |
steal_oauth_params() creates a list of strdup'ed strings containing the
name of the ouath parameters which should be stolen from the call
arguments. However, the code makes sure to free the list, but never
frees its content, causing a leak. This commit makes sure the dup'ed
strings are freed.
|
|
|
|
|
|
|
|
| |
This prevents a crash due to cur_node being NULL.
Found by scan-build.
https://bugzilla.gnome.org/show_bug.cgi?id=719550
|