Age | Commit message (Collapse) | Author |
|
If rSession.m_Proxy.aName is a simple host-name, the CurlUri constructor
will fail with CURLUE_BAD_SCHEME, so just ignore the error here,
we only care about parsing out the username/password
Change-Id: Iec2d6e7315a5899ddddf6120a43199b75bf62db2
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/155834
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit 9b30b4b1678e8be15ba51d236bd9a3e693d8d3d6)
|
|
Even if it used to be valid before; unfortunately newly entered
credentials weren't actually used because the flag was never reset.
Change-Id: Ib36689f40ff780596b9cfe6fe589a6f2e79cfcd2
|
|
- don't require it to be successful.
Trying to connect to a server with self-signed CA results in:
warn:ucb.ucp.webdav.curl:6796:6568:ucb/source/ucp/webdav-curl/CurlSession.cxx:946:
curl_easy_perform failed: (35) schannel: next InitializeSecurityContext failed:
Unknown error (0x80092012) - The revocation function was unable to check revocation for the certificate.
Apparently schannel wants to check by default (called with
SCH_CRED_REVOCATION_CHECK_CHAIN) that all the certificates aren't
revoked, but the self-signed CA doesn't specify how to check.
Set it to only check revocation when the way to do so actually works,
via CURLSSLOPT_REVOKE_BEST_EFFORT, which sets these flags:
SCH_CRED_IGNORE_NO_REVOCATION_CHECK | SCH_CRED_IGNORE_REVOCATION_OFFLINE | SCH_CRED_REVOCATION_CHECK_CHAIN
Change-Id: I6d77ca23fe2012d8a5d65000b14775070b5c9a0f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/138204
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit f6a0ca0e92e41ad8fea71acdacdc7ec5e775dc59)
|
|
With curl 7.86, the test JunitTest_ucb_complex starts to fail:
.ftp://noname:nopasswd@*nohost.invalid
now executing open
com.sun.star.ucb.InteractiveAugmentedIOException:
at com.sun.proxy.$Proxy15.execute(Unknown Source)
at complex.ucb.UCB.executeCommand(UCB.java:63)
at complex.ucb.UCB.checkWrongFtpConnection(UCB.java:119)
because curl_easy_perform() now returns CURLE_URL_MALFORMAT where
previously it was CURLE_COULDNT_RESOLVE_HOST.
Map this to an exception the test expects.
Change-Id: Ifdb672946726ddb4cb4d9426b7e70eefac63f040
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/141877
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 022e476af44c0dfc97403dc0f3a3b63e731903e6)
|
|
ever since a system update, this test receives an
IllegalIdentifierException thrown by
ContentProvider::queryContent
in
ucb/source/ucp/webdav-neon/webdavprovider.cxx
Given that
(1) On a system where the test passes, no IllegalArgumentException is
thrown
(2) Comment in the test indicates that the test should be run with proxy
enabled, which is not the case in our build
(3) I am running with a system proxy enabled
(4) the IllegalIdentifierException looks quite valid
I suspect that I am the victim of a dodgy test and some new proxy
behaviour in Fedora31.
Change-Id: Id3fb3499a44b5672e4eed0a2749dc287d6e7a3cc
Reviewed-on: https://gerrit.libreoffice.org/83853
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit 649a37f1f4048eb10303b6d1a46050234807f8ce)
|
|
Change-Id: I63c3521de965ae3aba3079def8e29a189f9b44d2
|
|
This reverts commit 275b318cd55720b1106965c31e8d60619ae50ec9.
|
|
Change-Id: Idbfacc858dd0180c40e86bbd6fc9b0084bd9ed2d
|
|
Not necessary but probably faster.
Documentation says that CURL_LOCK_DATA_CONNECT isn't safe in a multi
threaded program.
Change-Id: Iae79c01330b48f443b94abf6101a50559cb46189
|
|
Change-Id: I8796aa7de6335c57818aa570a55fdc64d85bd9f9
|
|
Change-Id: Iafa8bdd183ef8a514b656ec41a9b7a6fa1e3acb9
|
|
Change-Id: I0018a9904857d7521895936dc27607a54523f300
|
|
Change-Id: I93e0c8f95beafc30b94296430352f2ae54e65b11
|
|
Sharepoint reports lack of authentication with 403 status and additional
X-MSDAVEXT_ERROR header value 0x000E0098.
Try to fallback to NTLM auth in this case, if a first attempt with
imported cookie didn't work.
Note that the build config of libcurl on Linux is such that adding
CURLAUTH_NEGOTIATE has no effect, but on WNT it causes the AuthMask
to be ambiguous and prevents curl from trying NTLM.
|
|
This reverts commit cc77bc0e5273c6cf404851624ce5b127cdd839f4.
|
|
Sharepoint reports lack of authentication with 403 status and additional
X-MSDAVEXT_ERROR header value 0x000E0098.
Try to fallback to NTLM auth in this case, if a first attempt with
imported cookie didn't work.
Change-Id: I0d6dca2989d276262547a61784a3d0ed8bff9abd
|
|
Improve error handling/logging, and do it only if the error code
0x000E0098 is received.
Change-Id: I47dada2ef08b21a43cdfa3db9eb2fcdb4043a04f
|
|
This reverts commit 9352ba337a84672a0ba60651023422de88105456.
|
|
Change-Id: I1d22ff039f10304588820388e62701ea1f515a80
|
|
This is just for testing, don't ship.
Change-Id: Ic2344c61cbe3cf419b698be526a20388c0ef4f37
|
|
Sharepoint may reject LOCK with HTTP/1.1 403 FORBIDDEN
and then a dialog pops up via UUIInteractionHelper that says
"Server error message: ."
Let's actually put some error message in the dialog, why not the
HTTP status line, plus a little prefix to see which method failed.
Change-Id: Ied895787f813c5cddcb18eb4f693d5bfc8c62076
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133835
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit cb64a52afc92891ab853b9bb1294610bb9ca98d0)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133856
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
(cherry picked from commit 061fbad1e33f68726ee5605c7903f4ae219d8539)
|
|
Sharepoint uses properties like these:
"http://schemas.microsoft.com/repl/resourcetag"
"urn:schemas-microsoft-com:Win32CreationTime"
They aren't standard and don't match our own ucbprops namespace, and it
looks like they should be handled by an encoding to a name like
"<prop:Win32CreationTime xmlns:prop=\"urn:schemas-microsoft-com:\">"
Unfortunately WebDAVResponseParser::endElement() didn't do that when
handling a PROPFIND reply to get the property names.
This causes a crash when all properties are copied in
UniversalContentBroker::globalTransfer(), which is called by
SfxMedium::DoBackup_Impl() when the setting
"/org.openoffice.Office.Common/Save/Document/CreateBackup"
is in effect.
Change-Id: I2d6480bfd2f828b6e7fc431ba4b333d95ec12718
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133769
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 65362d63b9019e45d1224ed6d78d4e1d443d6b00)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133727
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
(cherry picked from commit 8a8226630e7e2458835f21f57cb73769ef06c668)
|
|
With the ndr.de server, OPTIONS on an image URL results in a status 500
Internal Server Error and an unexpected closed connection:
16.591:info:ucb.ucp.webdav.curl:1634132:1634170:ucb/source/ucp/webdav-curl/CurlSession.cxx:285: debug log: 0x7f9c94004988: transfer closed with 222 bytes remaining to read
Apparently Neon always reported the closed connection problem with a
string:
"Could not read response body: connection was closed by server"
The HTTP status code is extracted in makeStatusCode() and the string
doesn't contain it, so the status is reported as 0 up the stack.
With curl, it can apparently happen that either CURLE_PARTIAL_FILE is
returned, in which case the status is ignored and reported as 0,
or CURLE_OK is returned and the status 500 is reported up the stack.
Adapt the handling in Content::open() to fall back to GET in case of a
500 status.
Change-Id: I5f3dce7f67ce25c87bf3882c89e7ff2d386bcbad
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133707
Tested-by: Jenkins
Tested-by: Gabor Kelemen <kelemeng@ubuntu.com>
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 5801b887629dbd784116ec6878b61bb99e991647)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133723
Reviewed-by: Adolfo Jayme Barrientos <fitojb@ubuntu.com>
(cherry picked from commit 5459e6ef2554005ec045b76195ef1ba813a2dd34)
|
|
... so that the dialog in SfxMedium::LockOrigFileOnDemand() can show
something more useful than hard-coded URL, which has been used since
commit 99bdd887a6141883878978bad9beb35e7b326cd1 in 2009 when locking
was added.
This is half of the user info that is put into LO's own lock files
(see svt::LockFileCommon::GenerateOwnEntry()).
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133186
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 70bea73ffbe815c2d5b7067f20d3eebbb98df61f)
ucb: webdav-curl: check UseUserData before sending user name
Hope this should be acceptable.
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133188
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 6d7c3848560883034a4ff12ac1ef52d2d1d28606)
Change-Id: Iefac724644a536fc37c3c79ce862e25bd9be38af
|
|
commit b4576f3da4d90139fc5140962d13cb91dab98797 "tdf#82744: fix WebDAV
lock/unlock behaviour - part 3" added a call to get the
DAV:lockdiscovery property.
But WebDAVResponseParser puts lock related properties into a separate
return value maResult_Lock that is only returned for LOCK requests.
Just add it as a normal property too, then PROPFIND can get it, and
the dialog in SfxMedium::LockOrigFileOnDemand() no longer displays
"Unknown user".
Change-Id: Icee920588ea40b6e203b18287d75484528cfdebb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133185
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 68f41f4e2dcec65e8885cdfa658c0568a26182a9)
|
|
This is now the second bug filed because a server replies with 403 if
the UserAgent contains the string "curl".
Change-Id: I25ca2d255af76a7ff4e64dad900b1bf0b78de59f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133212
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 8d9c56e8f42428fd6695942c673bffb985d22ad5)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133146
Reviewed-by: Xisco Fauli <xiscofauli@libreoffice.org>
(cherry picked from commit 3fe6a03426a86ec6b18a9b712588619f19a72897)
|
|
... and in any case abort authentication after 10 failed attempts.
Apparently some PasswordContainer can turn this into an infinite loop.
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/132974
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 2bc4d1d22fdbd9d97c66bb53762b4b4bf7b61b47)
ucb: webdav-curl: oops, increment after checking
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/132982
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
Tested-by: Jenkins
(cherry picked from commit ab65a74998b498ff49c15db87fc14a9afa89d8bf)
Change-Id: Ib2333b371a770999e8407ce7e1af21512aadb70d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/132867
Tested-by: Jenkins
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
(cherry picked from commit 6b54e6a8e64233de63b826211b81a8ed6767483f)
|
|
Change-Id: Ifc98baec69e32d231e44b389a870e90debd5a389
|
|
Change-Id: I040e18ae4ddca11d1ed7303fce5d343d36c1c3b0
|
|
See tdf#74608 for motivation.
Change-Id: Ia86a4cc06819ce5820ab44d293790f7bd9a03711
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/99224
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit f3f0d05f72d8dbd9717910b4d6b42c54be3e106f)
|
|
to make the ongoing convert-to-UNO-constructor work easier
Change-Id: I99596561832b10538505cf4dfdabaf11d6e4d7cb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/98664
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
(cherry picked from commit 15f44cb6fbf2459dca522e4f5b7f525fd1d6c5e7)
|
|
Change-Id: Ic12c38be4426d7cb3c4134585df34a7107e8f4ce
|
|
Change-Id: Idf1d75817009286cd79a00c0ba154cea70e2ffda
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125908
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit b03e070420606d407df2ec5e9dfa7043ecc46177)
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/125895
Reviewed-by: Thorsten Behrens <thorsten.behrens@allotropia.de>
(cherry picked from commit db8edd163cf453bbb9b85c2a17c78da49df27cdf)
|
|
1 CppuniTest makefile should be sufficient, particularly since all 4
of them used gb_CppunitTest_use_library_objects so this should reduce
wasted space.
Change-Id: I485f0af0aee2b265b56c505ced44257834784e98
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/124608
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 895918b07da8fbe5603b3a96b3c58a8dc2d066e5)
|
|
There's no point in building the tests for neon any more, just build
them only if curl is used.
This finds a bug in CurlUri::Init() where query and fragment separators
need to be added manually.
Change-Id: I3dcd1512450522df2c2a7d223b2e29b6c0e794fb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/124607
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@allotropia.de>
(cherry picked from commit 5a44e3772adb281d16757cdac3120d56565a96e5)
|
|
TODO: ucb/Library_ucpdav1.mk changes
Change-Id: I528bd0b4ccbd3c8cfe98b74a94269ef99292f82e
|
|
we get a GError of G_IO_ERROR_FAILED_HANDLED which is documented
as "Operation failed and a helper program has already interacted with the user.
Do not display any error dialog." PENDING seems to describe this state
better than a generic error case and results in no error dialog.
Change-Id: Iee382e49edaa0a734526659971e0339991e2c03e
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/92171
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@cib.de>
|
|
INetURLObject to file UCP"
It turns out that there is lots of code in and around e.g.
SfxMedium::LockOrigFileOnDemand (sfx2/source/doc/docfile.cxx) that treats
document URLs non-generically and treats file URLs specially. That means that
it does not work to transparently rewrite a file URL as an smb URL in the
underlying UCB layer. So for libreoffice-6-4 revert the change again. (For
master towards LO 6.5, I plan to instead completely drop the feature of silently
treating certain forms of file URLs on Linux as smb URLs.)
This reverts commit 46c645bf4e9909f5296e75028f1f5434e83942d2.
Conflicts:
tools/qa/cppunit/test_urlobj.cxx
ucb/source/ucp/file/prov.cxx
Change-Id: I94ac0637115bea36924099951efd8542854a727d
Reviewed-on: https://gerrit.libreoffice.org/83396
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
...checking for unnecessary uses of more "fake bool" types.
In the past, some of the checks involving the types of variables or data
members, or the return types of functions, issued warnings that required
surrounding code to be changed too (e.g., when changing the signature of a
function whose address was taken). These checks have been tightened now to not
warn in such cases (which avoids warnings that require changes to additional
code, or changes that might even be impossible to make, at the cost of being
less aggressive about removing all unnecessary uses of those "fake bool" types).
Change-Id: I70eb75039817cda34ed611387ee27dc5f36a3e2e
Reviewed-on: https://gerrit.libreoffice.org/82554
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
tweak the plugin to be more permissive, then validate by hand
afterwards
Change-Id: I40c5c911fe6ff7e45baaca372abf7dac211d9654
Reviewed-on: https://gerrit.libreoffice.org/81942
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
Change-Id: I2eab990c15f845b44a3b598571aca361dadf9ff3
Reviewed-on: https://gerrit.libreoffice.org/81946
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
...to find StringLiteral on the RHS of +=. Which revealed that the
VisitCompoundStmt/checkForCompoundAssign logic needed to be fixed, too, so that
s += side_effect();
s += "literal";
s += side_effect();
only gets combined to
s += side_effect() + "literal";
s += side_effect();
and not all the way to
s += side_effect() + "literal" + side_effect();
Change-Id: I432e3458b933a7d0ad6141c747b675cc8b0f0ba4
Reviewed-on: https://gerrit.libreoffice.org/81804
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Also recheck hxx files after recent changes in f-u-i
Note: to build ucb/source/ucp/webdav one needs to use the
--with-webdav=serf
configure option
Found with bin/find-unneeded-includes
Only removal proposals are dealt with here.
Change-Id: I69c3a869394b0667b4d51e18e5293000872227a5
Reviewed-on: https://gerrit.libreoffice.org/81496
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
|
|
if one side of the expression is a compile-time-constant, we don't need
to worry about side-effects on the other side
Change-Id: Iee71ea51b327ef244bf39f128f921ac325d74e2b
Reviewed-on: https://gerrit.libreoffice.org/81589
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
It is unclear whether this will really be errors in a final C++20 standard, see
<https://github.com/llvm/llvm-project/commit/
974c8b7e2fde550fd87850d50695341101c38c2d> "[c++20] Add rewriting from comparison
operators to <=> / ==" for details.
Change-Id: I3a7c2574a65952c8d255107823a1d1e18b2cbe4c
Reviewed-on: https://gerrit.libreoffice.org/81268
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
found by the simple expidient of putting asserts in
the resize routine. Where an explicit const size is used,
I started with 32 and kept doubling until that site
did not need resizing anymore.
Change-Id: I998787edc940d0a3ba23b5ac37131ab9ecd300f4
Reviewed-on: https://gerrit.libreoffice.org/81138
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
See <https://gitlab.gnome.org/GNOME/glib/merge_requests/1170> "Avoid C++20
deprecated assignment to volatile" for a potential fix in GLib (but also see my
question at <https://github.com/llvm/llvm-project/commit/
4a6861a7e5b59be24a09b8b9782255d028e7aade#commitcomment-35540755> "[c++20]
P1152R4: warn on any simple-assignment to a volatile lvalue" whether that Clang
warning is actually wanted here).
With glib2-devel-2.62.1-1.fc31.x86_64, the suppressed warnings looked like
> libreofficekit/source/gtk/lokdocview.cxx:327:1: error: use of result of assignment to object of volatile-qualified type 'volatile gsize' (aka 'volatile unsigned long') is deprecated [-Werror,-Wdeprecated-volatile]
> G_DEFINE_TYPE_WITH_CODE (LOKDocView, lok_doc_view, GTK_TYPE_DRAWING_AREA,
> ^
> /usr/include/glib-2.0/gobject/gtype.h:1617:56: note: expanded from macro 'G_DEFINE_TYPE_WITH_CODE'
> #define G_DEFINE_TYPE_WITH_CODE(TN, t_n, T_P, _C_) _G_DEFINE_TYPE_EXTENDED_BEGIN (TN, t_n, T_P, 0) {_C_;} _G_DEFINE_TYPE_EXTENDED_END()
> ^
> /usr/include/glib-2.0/gobject/gtype.h:2034:3: note: expanded from macro '_G_DEFINE_TYPE_EXTENDED_BEGIN'
> _G_DEFINE_TYPE_EXTENDED_BEGIN_REGISTER(TypeName, type_name, TYPE_PARENT, flags) \
> ^
> /usr/include/glib-2.0/gobject/gtype.h:2005:7: note: expanded from macro '_G_DEFINE_TYPE_EXTENDED_BEGIN_REGISTER'
> g_once_init_leave (&g_define_type_id__volatile, g_define_type_id); \
> ^
> /usr/include/glib-2.0/glib/gthread.h:257:17: note: expanded from macro 'g_once_init_leave'
> (void) (0 ? *(location) = (result) : 0); \
> ^
Change-Id: If67ad04f8fb242f50b43a1d98ad2b28c4bed55a4
Reviewed-on: https://gerrit.libreoffice.org/80937
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
Replace them with default initialization or calloc
Change-Id: I747f53c2ced2d0473fd5a5ede4f8520a0633dcc1
Reviewed-on: https://gerrit.libreoffice.org/80805
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
|
|
look for OUStringBuffer append sequences that can be turned
into creating an OUString with + operations
Change-Id: Ica840dc096000307b4a105fb4d9ec7588a15ade6
Reviewed-on: https://gerrit.libreoffice.org/80809
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|
|
which defeat the *StringConcat optimisation.
Also make StringConcat conversions treat a nullptr as an empty string,
to match the O*String(char*) constructors.
Change-Id: If45f5b4b6a535c97bfeeacd9ec472a7603a52e5b
Reviewed-on: https://gerrit.libreoffice.org/80724
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
|