Similar to commit 1f24d35033 (cppunittester:
use a dedicated desktop on Windows, 2023-07-09), this change allows to run
UI tests on Windows without them interferring with what you are doing.
The code testing LIBO_TEST_DEFAULT_DESKTOP environment variable to avoid
the separate desktop, and creating it, is extracted to a new function in
systools.
Change-Id: I6df3cb4ae581ae289e49b47d3135d01f632fbd95
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/170615
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
Since introduction of accessibility tests infrastructure in commit
0185ddd6d5 (Add infrastructure and
basic tests including slight UI interaction, 2022-08-01) and the
respective tests, running 'make check' on Windows produces pop-up
dialogs, while these tests run. These dialogs distract, steal focus,
swallow text that I type elsewhere, and may fail the tests if
accidental user input interferes with what the tests check.
This commit creates a dedicated system desktop for cppunittests,
which would be isolated from the active desktop, yet allow creation
of windows, thus not preventing accessibility testing. This way,
it workarounds unavailability of svp vcl plugin on Windows.
Change-Id: I83db54c82bfe98d14171355cc19cdd5767549fdb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/154194
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
The plugin in its current form is probably not intending to find these cases.
But it happened to do so on macOS with libc++, where the std::basic_string ctors
taking a final defaulted Allocator argument are instead implemented by a pair of
overloaded ctors, one taking no Allocator at all (see
671d1c6cd1 "Adapt loplugin:unnecessarygetstr to
libc++").
But these changes here are useful regardless, so lets leave it at that.
Change-Id: I2776671e2953182bdcad36432951a75f82412ebb
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/151410
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Often a developer debugging a failing CppUnit test wants a core dump with the
place where an uncaught exception is thrown. So if the newly introduced
CPPUNIT_PROPAGATE_EXCEPTIONS environment variable is set (to any value), disable
all the protectors that would otherwise catch such exceptions (and just report
some limited information about them).
Change-Id: I3052f71c0787583c496279a6f5b35a0299c357b3
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/143882
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
noting that this commit is completely blind. But we're not getting
good stacks on jenkins, and it doesn't look right to me that we're
using 64-bit functions on 32-bit stacks, so make the 32-bit paths
consistently use 32-bit functions and structures.
Change-Id: I413ab97001246c442df151700cc618df4391e03b
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/131964
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
> sal/cppunittester/cppunittester.cxx(553,115): error: format specifies type 'unsigned int' but the argument has type 'ULONG64' (aka 'unsigned long long') [-Werror,-Wformat]
> printf("\tat %s in %s: line: %lu: address: 0x%0X\n", pSymbol->Name, line->FileName, line->LineNumber, pSymbol->Address);
> ~~~ ^~~~~~~~~~~~~~~~
> %0llX
> sal/cppunittester/cppunittester.cxx(558,64): error: format specifies type 'unsigned int' but the argument has type 'ULONG64' (aka 'unsigned long long') [-Werror,-Wformat]
> printf("\tat %s, address 0x%0X.\n", pSymbol->Name, pSymbol->Address);
> ~~~ ^~~~~~~~~~~~~~~~
> %0llX
> sal/cppunittester/cppunittester.cxx(576,50): error: format specifies type 'unsigned int' but the argument has type 'DWORD' (aka 'unsigned long') [-Werror,-Wformat]
> printf("*** Exception 0x%x occurred ***\n\n",ex->ExceptionRecord->ExceptionCode);
> ~~ ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> %lx
...but where SYMBOL_INFO::Address is documented to be of type ULONG64, which in
turn is documented to be a typedef for unsigned __int64, for which there
is the MSVC-extension I64 format specifier
Change-Id: Ibed1d3fa49ac6fd988174c6041071292fc66dd3d
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/122872
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
so the multiple unit tests don't stomp on each other.
This fixes a couple of things in my earlier attempt
(*) actually set the env variable on Windows
(*) don't use a global variable to test the env var, because that
variable might be initialised BEFORE the env var is actually set
Change-Id: Id43a1dd2fbd324691e0b6578c9026b8a523012e8
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/121436
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
...for LIBO_INTERNAL_ONLY. These had been missed by
1b43cceaea "Make many OUString functions take
std::u16string_view parameters" because they did not match the multi-overload
pattern that was addressed there, but they nevertheless benefit from being
changed just as well (witness e.g. the various resulting changes from copy() to
subView()).
This showed a conversion from OStringChar to std::string_view to be missing
(while the corresponding conversion form OUStringChar to std::u16string_view was
already present).
The improvement to loplugin:stringadd became necessary to fix
> [CPT] compilerplugins/clang/test/stringadd.cxx
> error: 'error' diagnostics expected but not seen:
> File ~/lo/core/compilerplugins/clang/test/stringadd.cxx Line 43 (directive at ~/lo/core/compilerplugins/clang/test/stringadd.cxx:42): simplify by merging with the preceding assignment [loplugin:stringadd]
> File ~/lo/core/compilerplugins/clang/test/stringadd.cxx Line 61 (directive at ~/lo/core/compilerplugins/clang/test/stringadd.cxx:60): simplify by merging with the preceding assignment [loplugin:stringadd]
> 2 errors generated.
Change-Id: Ie40de0616a66e60e289c1af0ca60aed6f9ecc279
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/107602
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
-visual: so that the human eye can easily find it
-at start: to identify the last test if it ends abruptly
-no spaces: easy scripting | grep RUN | awk '{print $2}'
A visually unique, eye-catching start banner
will be helpful when browsing through logs,
or when adding asserts and running a make sw.check.
[I frequently look for existing unit tests that match
a certain condition in the 15 ooxmlexport tests.]
Previously, the tests ran in the order that they existed
in the file. So if one asserted/crashed in the middle of
the run, you could easily guess which test it was
by finding the name of the test that just finished OK.
However, since
2f2246d22e
Author: Stephan Bergmann on Sat Aug 8 12:00:23 2020
external/cppunit: Run tests in deterministic order
the tests are sorted by name and run in
alphabetical order - which is not at all obvious or
intuitive. It certainly makes it harder to guess
what the next alphabetical name will be.
P.S. For consistency, I considered adding [____DONE_]
as well, but existing tooling might search for
"finished in:" and grab the first string (the test name)
so I don't want to break anything doing that.
Change-Id: I5521e681cd9f8304842290680f0389256fe3fe43
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/103268
Tested-by: Jenkins
Reviewed-by: Justin Luth <justin_luth@sil.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
At least on Windows, e6dfaf9f44 "Turn
OUStringLiteral into a consteval'ed, static-refcound rtl_uString" caused
CppunitTest_sc_tablesheetsobj to fail at exit, when the
static OUString aCacheName
in ScDocument::GetTable (sc/source/core/data/document.cxx) is destroyed but
references
constexpr OUStringLiteral gaSrcSheetName(u"SheetToCopy")
from test/source/sheet/xspreadsheets2.cxx in Library_subsequenttest, referenced
(only) by the CppunitTest_sc_tablesheetsobj library, and both those libraries
had already been unloaded from memory.
Change-Id: Icea85019611fbaec3603f37b234e9e6fe3502961
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/103103
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
... calling OleInitialize early in sal_main to initialize COM and concurrency.
Seeing intermittent hangs in main thread in CAPNDataObject::GetData calling
m_rIDataObjectOrg->GetData, and the inner stack waiting objects in the code
doing apartment switching (I forgot to save a stack to paste unfortunately),
I suspect that being related to incorrect concurrency model. OleInitialize
docs [1] mention:
Applications that use the following functionality must call OleInitialize
before calling any other function in the COM library:
* Clipboard
* Drag and Drop
* Object linking and embedding (OLE)
* In-place activation
...
Because OLE operations are not thread-safe, OleInitialize specifies the
concurrency model as single-thread apartment.
CoInitializeEx sets COINIT_MULTITHREADED by default, so possibly that might
get called somewhere before clipboard/OLE code is called. I hope that this
change would fix those hangs.
[1] https://docs.microsoft.com/en-us/windows/win32/api/ole2/nf-ole2-oleinitialize
Change-Id: I7213c9d6cb4bd0691a3ce355995157797d7db93f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94675
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
rtl/string.hxx and rtl/ustring.hxx both unnecessarily #include <sal/log.hxx> (and don't make use of it themselves), but many other files happen to depend on it.
This is a continuation of commit 6ff2d84ade to be able to remove those unneeded includes.
This commit adds missing headers to every file found by:
grep -FwL sal/log.hxx $(git grep -Elw 'SAL_INFO|SAL_INFO_IF|SAL_WARN|SAL_WARN_IF|SAL_DETAIL_LOG_STREAM|SAL_WHERE|SAL_STREAM|SAL_DEBUG')
in files formerly omitted for oversight or non-cxx extension
Change-Id: I327c573f44076c6ccfecf737eafccba2da72e1bd
Reviewed-on: https://gerrit.libreoffice.org/58600
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
auto-rewrite with <https://gerrit.libreoffice.org/#/c/47798/> "Enable
loplugin:cstylecast for some more cases" plus
solenv/clang-format/reformat-formatted-files
Change-Id: I7d89b011464ba5d2dd12e04d5fc9f65cb4daebde
Fail the test if CPPUNIT_TEST_NAME does not contain at least one
existing test, which is much better than silently reporting OK(0).
Change-Id: If2137764b24aa52dc35fe2368ee8fe38cb7dbece
Caolán McNamara