Remove `status` from our internal Subprocess utility.
This facility is unused, so better to just remove it than figure out what its cross-OS semantics should be.
PiperOrigin-RevId: 345279290
Check for a high bit to see if a return value is a signal or an error code in googletest/test.
This is needed because for subprocess under python3 windows, a return value representing a C signal (such as 0x80000003) is represented as a large positive integer rather than a negative one.
PiperOrigin-RevId: 345270460
Include TargetConditionals.h before checking TARGET_OS_IPHONE
New versions of Clang require that you include TargetConditionals.h before checking the TARGET_OS_* macros. Include it where appropriate to keep the compiler happy.
PiperOrigin-RevId: 345106443
Use environment variable TEST_TMPDIR in Linux environments
for temporary directory if available otherwise use /tmp/.
Bazel sets the environment variable TEST_TMPDIR when launching tests.
PiperOrigin-RevId: 342058921
Use a tagged constructor for FlatTuple instead.
Some versions of MSVC are getting confused with that constructor and generating invalid code.
PiperOrigin-RevId: 342050957
Fix *_THROW tests under MSVC, which produce slightly different type
names from std::type_info (e.g. "class std::runtime_error" instead of
"std::runtime_error")
PiperOrigin-RevId: 341874178
Rollback change from
https://github.com/google/googletest/pull/1836. This change generates
a script on Windows to actually run each test, but the script itself
doesn't correctly report if the test passed.
This change will "break tests" that were already broken on Windows,
but weren't being reported as such.
PiperOrigin-RevId: 341850671
On a diskless system you cannot get the current directory. So if
death tests are disabled anyway, there is no point trying to
get current directory.
Without this fix, running tests on diskless systems will fail,
even when death tests are disabled.
Improve lookup of operator<< for user types
Without this fix, trying to use this class with googletest
struct Foo {};
template <typename OutputStream>
OutputStream& operator<<(OutputStream& os, const Foo&) {
os << "TemplatedStreamableInFoo";
return os;
}
results in an ambiguity error between the class' operator<< and the
operator<< in gtest-printers.h removed in this CL.
This fix also enables implicit conversions to happen, so that e.g.
we will find the base class operator<< if a subclass has no
operator<< of its own.
PiperOrigin-RevId: 336261221
Revision of recent DoubleNearPredFormat change to support more toolchains.
isnan() is a macro in C99, and std::isnan() is a function in C++11. The previous change used `isnan` directly, and broke some tests in open source.
This CL changes it to follow the practice in gmock-matchers.h, and spell uses of isnan as
(std::isnan)(f)
. The parens around `std::isnan` prevent it from being recognized as a macro in the preprocessor.
PiperOrigin-RevId: 333374377
Improve DoubleNearPredFormat output on bad epsilons
DoubleNearPredFormat will happily accept epsilon values (abs_error) that
are so small that they are meaningless. This turns EXPECT_NEAR into a
complicated and non-obvious version of EXPECT_EQ.
This change modifies DoubleNearPredFormat) so that when there is a
failure it calculates the smallest meaningful epsilon value, given the
input values, and then prints a message which explains what happened.
If a true equality test is wanted either pass a literal 0.0 as abs_error
or use EXPECT_EQ. If a check for being almost equal is wanted consider
using EXPECT_DOUBLE_EQ which, contrary to its name, verifies that the
two numbers are *almost* equal (within four ULPs).
With this change the flaky test mentioned in crbug.com/786046 gives this
output:
The difference between 4.2934311416234112e+18 and 4.2934311416234107e+18 is 512, where
4.2934311416234112e+18 evaluates to 4.2934311416234112e+18,
4.2934311416234107e+18 evaluates to 4.2934311416234107e+18.
The abs_error parameter 1.0 evaluates to 1 which is smaller than the minimum distance between doubles for numbers of this magnitude which is 512, thus making this EXPECT_NEAR check equivalent to EXPECT_EQUAL. Consider using EXPECT_DOUBLE_EQ instead.
Tested:
I confirmed that this change detects the bad epsilon value that caused
crbug.com/786046 in Chromium and added a test for the desired output.
PiperOrigin-RevId: 332946880
gtest.cc: make ColoredPrintf static
the prototype was removed from gtest.h in cl/301446904; quiets a
-Wmissing-declarations warning
PiperOrigin-RevId: 329569020
Add millisecond precision to start timestamp in XML/JSON output
- Previous timestamp had format YYYY-MM-DDThh:mm:ss, now YYYY-MM-DDThh:mm:ss.sss
- This conforms to the ISO 8601 standard
PiperOrigin-RevId: 329503623
Mention matchers as an alternative to assertions in subroutines.
Matchers are often the better choice - they can provide more informative error messages and circumvent all of the complexity described in this section.
PiperOrigin-RevId: 326332149
Updated documentation for Value-Parameterized Tests: Per-default, a TEST_P without a corresponding INSTANTIATE_TEST_SUITE_P now causes a failing test.
PiperOrigin-RevId: 323622468
Detection of string_view type (whether it's std or Abseil)
is done in googletest/include/gtest/internal/gtest-port.h
with GTEST_INTERNAL_HAS_STRING_VIEW.
Make visible ParseInt32 in case users have separate gmock/gtest libraries and hidden-by-default symbols.
This function is still considered an internal implementation detail and is subject to change without notice. It is still unsafe/unsupported to link together libraries built at different commits.
PiperOrigin-RevId: 315405429
Fail TEST_Ps or TYPED_TEST_Ps that are defined but are not instantiated, as well as the opposite case, where INSTANTIATE_TEST_SUITE_P or INSTANTIATE_TYPED_TEST_SUITE_P is used but without any matching TEST_P or TYPED_TEST_P.
PiperOrigin-RevId: 315255779
Adds support for printing the types char8_t, char16_t, and char32_t
This changes prints these types as Unicode code points. It is possible
that there is a better way of printing these types, but that change is
more complex, and the format in which Googletest prints these types is
subject to change if someone implements a better way of printing them.
This fixes the C++20 build, which removed support for printing these types.
https://en.cppreference.com/w/cpp/io/basic_ostream/operator_ltlt2Fixes#2854
PiperOrigin-RevId: 314826912
Fixed#2823 - Make it so that a semicolon appearing after an invocation of GTEST_SUPPRESS_UNREACHABLE_CODE_WARNING_BELOW_ does not trigger a redundant semicolon warning.
This works by introducing an else block with a statement that intentionally does not end with a semicolon, forcing users to place the semicolon after the expansion. The approach here is preferred as opposed to removing semicolons that appear after each invocation because complete statements that do not have a visible semicolon or braces confuse users and code formatters, since the macro invocation looks superficially like an expression.
PiperOrigin-RevId: 311327491
The documentation for IsDebuggerPresent says that one just should
include windows.h, as that one is an umbrella header that includes
the header that declares IsDebuggerPresent. In older Windows SDKs,
debugapi.h didn't exist and IsDebuggerPresent was declared in
winbase.h (also included by windows.h).
This should fix issue #2822 properly.
This reverts commit a9f6c1ed14.
That commit cannot fix the issue it sets out to fix.
The original issue, #2822, was that building with a toolset
targeting XP compatibility is missing the debugapi.h header -
as debugapi.h didn't exist in older Windows SDKs.
Commit a9f6c1ed14 misinterpreted
the Microsoft documentation about IsDebuggerPresent. The information
about which header to use, "debugapi.h (include Windows.h)" means
that the function declaration currently lives in debugapi.h, but
for compatibility, just include the Windows.h umbrella header.
In older Windows SDKs (e.g. the v6.0a SDK), IsDebuggerPresent
is declared in winbase.h, and debugapi.h doesn't exist at all in those
versions.
Including Windows.h with a different capitalization than the existing
include won't help finding headers that don't exist.
Including Windows.h with a capital W breaks cross compilation with mingw
toolchains, where the header always has been spelled with a lower case
W. When building on native windows, the file system is case insensitive
and the capitalization doesn't matter.
This fixes issue #2840.
Explicitly define copy constructors used in googletest tests
As of C++11, providing a user-declared copy assignment operator should
suppress the availability of an implicit default copy constructor.
Classes that provide (or delete) a copy assignment operator must provide
their own copy constructor if one is desired. This may be an explicit
default copy constructor if appropriate.
As googletest is a C++11 codebase, this change should be made without
qualification.
This addresses the -Wdeprecated-copy warnings issued by trunk clang:
While compiling googletest/test/googletest-death-test-test.cc:
In file included from .../googletest/test/googletest-death-test-test.cc:33:
.../googletest/include/gtest/gtest-death-test.h:196:8: error: definition of implicit copy constructor for 'ExitedWithCode' is deprecated because it has a user-declared copy assignment operator [-Werror,-Wdeprecated-copy]
void operator=(const ExitedWithCode& other);
^
.../googletest/test/googletest-death-test-test.cc:279:16: note: in implicit copy constructor for 'testing::ExitedWithCode' first required here
EXPECT_PRED1(pred0, status0);
^
While compiling googletest/test/googletest-param-test-test.cc:
.../googletest/test/googletest-param-test-test.cc:502:8: error: definition of implicit copy constructor for 'NonDefaultConstructAssignString' is deprecated because it has a user-declared copy assignment operator [-Werror,-Wdeprecated-copy]
void operator=(const NonDefaultConstructAssignString&);
^
.../googletest/test/googletest-param-test-test.cc:507:36: note: in implicit copy constructor for 'NonDefaultConstructAssignString' first required here
Combine(Values(0, 1), Values(NonDefaultConstructAssignString("A"),
This matches other changes made elsewhere in the googletest codebase,
such as 306f3754a7. Perhaps those previous changes did not consider
test code.
PiperOrigin-RevId: 307495126
googletest-param-test-test.cc:502:8: error:
definition of implicit copy constructor for
'NonDefaultConstructAssignString' is deprecated because it has a
user-declared copy assignment operator [-Werror,-Wdeprecated]
void operator=(const NonDefaultConstructAssignString&);
^
googletest-port-test.cc:97:11: error:
definition of implicit copy constructor for 'Base' is deprecated because
it has a user-declared destructor [-Werror,-Wdeprecated]
virtual ~Base() {}
^
None of these are strictly needed for correctness.
A large number of them (maybe all of them?) trigger `-Wdeprecated`
warnings on Clang trunk as soon as you try to use the implicitly
defaulted (but deprecated) copy constructor of a class that has
deleted its copy assignment operator.
By declaring a deleted copy assignment operator, the old code
also caused the move constructor and move assignment operator
to be non-declared. This means that the old code never got move
semantics -- "move-construction" would simply call the defaulted
(but deprecated) copy constructor instead. With the new code,
"move-construction" calls the defaulted move constructor, which
I believe is what we want to happen. So this is a runtime
performance optimization.
Unfortunately we can't yet physically remove the definitions
of these macros from gtest-port.h, because they are being used
by other code internally at Google (according to zhangxy988).
But no new uses should be added going forward.
This change updates testing::internal::IsAProtocolMessage to return true not
just for full proto messages but also for lite ones (i.e. those inheriting
directly from MessageLite).
PiperOrigin-RevId: 304286535
The extra tests check that the xml output is correct when a failure occurs before skipping, and that the right skip message is added to the file.
The json file had to be fixed because it's the same executable.
Add --gtest_fail_fast support to googletest.
- Analogous functionality to to golang -test.failfast and python --failfast
- Stops test execution upon first test failure.
- Also add support Bazel equivalent env var (TESTBRIDGE_TEST_RUNNER_FAIL_FAST)
PiperOrigin-RevId: 302488880
Google Test and Google Mock require matching versions to work,
so this requirement should be described in the pkgconfig files.
This change is derived from the one used for the Fedora gtest package.
Signed-off-by: Neal Gompa <ngompa13@gmail.com>
When building packaged shared libraries for use, having the version
set makes it so that the soname is set correctly for parallel installation.
This change is derived from the one used for the Fedora gtest package.
Signed-off-by: Neal Gompa <ngompa13@gmail.com>
Exposition of the problem:
> cat main.cpp
TEST(errnoTest, errnoTest)
{
ASSERT_EQ(errno, 0);
}
int main(int argc, char** argv)
{
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
Compiled with gcc 10 like this:
> g++ -pthread -o runtest main.cpp -Wl,-Bstatic -lgtest -Wl,-Bdynamic
Before patch:
> ./runtest
[==========] Running 1 test from 1 test suite.
[----------] Global test environment set-up.
[----------] 1 test from errnoTest
[ RUN ] errnoTest.errnoTest
[ OK ] errnoTest.errnoTest (0 ms)
[----------] 1 test from errnoTest (0 ms total)
[----------] Global test environment tear-down
[==========] 1 test from 1 test suite ran. (0 ms total)
[ PASSED ] 1 test.
(output is colored, I run this inside an interactive terminal).
> ./runtest | cat
[==========] Running 1 test from 1 test suite.
[----------] Global test environment set-up.
[----------] 1 test from errnoTest
[ RUN ] errnoTest.errnoTest
main.cpp:5: Failure
Expected equality of these values:
(*__errno_location ())
Which is: 25
0
[ FAILED ] errnoTest.errnoTest (0 ms)
[----------] 1 test from errnoTest (0 ms total)
[----------] Global test environment tear-down
[==========] 1 test from 1 test suite ran. (0 ms total)
[ PASSED ] 0 tests.
[ FAILED ] 1 test, listed below:
[ FAILED ] errnoTest.errnoTest
1 FAILED TEST
(output is not colored, since IsTTY return false, because of the pipe,
however it also clobbered errno for the tests).
After the patch, both cases are working fine:
> ./runtest
[==========] Running 1 test from 1 test suite.
[----------] Global test environment set-up.
[----------] 1 test from errnoTest
[ RUN ] errnoTest.errnoTest
[ OK ] errnoTest.errnoTest (0 ms)
[----------] 1 test from errnoTest (0 ms total)
[----------] Global test environment tear-down
[==========] 1 test from 1 test suite ran. (0 ms total)
[ PASSED ] 1 test.
> ./runtest | cat
[==========] Running 1 test from 1 test suite.
[----------] Global test environment set-up.
[----------] 1 test from errnoTest
[ RUN ] errnoTest.errnoTest
[ OK ] errnoTest.errnoTest (0 ms)
[----------] 1 test from errnoTest (0 ms total)
[----------] Global test environment tear-down
[==========] 1 test from 1 test suite ran. (0 ms total)
[ PASSED ] 1 test.
Refactor function GetNextPrime so that the loop precondition is checked before
loop instead of during every loop run. Also by removing the loop condition,
it shows that the only exit from the loop is the return statement.
PiperOrigin-RevId: 293932783
Disable warning C4800 for Visual Studio 2019.
The compiler warning C4800 is disabled by default in Visual Studio 2019,
but it can be enabled on the command line. The only version of
Visual Studio that does not support warning C4800 is Visual Studio 2017.
PiperOrigin-RevId: 292624510
Deleted an orphaned duplicate file and exclude another that shouldn't be part of :gtest_all_test.
This showed up while trying to debug the presubmit failure for: https://github.com/google/googletest/pull/2683
PiperOrigin-RevId: 291398123
Add extra filtering so that the reported message differentiates between the case where INSTANTIATE_TEST_SUITE_P is missing vs. the case where TEST_P is missing.
PiperOrigin-RevId: 290114508
Add GTEST_ALLOW_UNINSTANTIATED_PARAMTERIZED_TEST to mark a paramaterized test as allowed to be un-instantiated.
This allows test suites, that are defined in libraries and, for other reasons, get linked in (which should probably be avoided, but isn't always possible) to be marked as allowed to go uninstantiated.
This can also be used to grandfather existing issues and expedite adoption of the checks with regards to new cases before they can be fixed.
PiperOrigin-RevId: 289581573
Removing GTEST_API from TrueWithString.
This type is only used in test code on one side of a DLL boundary so it is not
necessary.
PiperOrigin-RevId: 288927929
Add option (default to disabled) to make C++ type parameterized tests (TYPED_TEST_P) fail when they're not instantiated.
When an un-instantiated TYPED_TEST_P is found, a new test will be inserted that emits a suitable message. For now, that is just a notice, but the hope it to flip the bit to make it fail by default.
PiperOrigin-RevId: 286408038
This reverts commit 20b5b8ecc7.
Reason for revert: Breaks existing applications, such as ANGLE
(angleproject.org), requires adding an extra prefix that needs to be
typed for no reason (when testing from command line) and increases the
logs' sizes by a non-trivial amount due to the very large number of
tests ANGLE runs.
Original commit message:
Add a compile time check to ensure that the parameters to TEST_P and
INSTANTIATE_TEST_SUITE_P are not empty. Some compilers already fail
in that case and, even where it works, it's likely to result in
technically invalid code by virtue of creating reserved identifiers:
https://en.cppreference.com/w/cpp/language/identifiers
First, every project is perfectly capable of adding a prefix if they
want to support such a compiler. This change penalizes every
project.
Second, using a prefix such as `_p` also results in reserved
identifiers, so this change is not really solving the problem. For that
matter, instead of generating `gtest_##prefix##...`, you can generate
`gtest_x##prefix##...` to correctly fix the issue, including when empty
prefixes are used.
This reverts commit 4c25e2b87f.
Reason for revert: Reverting the commit that disallows empty prefixes.
Original commit message:
Remove a test case rendered obsolete by disallowing empty argument for INSTANTIATE_TEST_SUITE_P.
Remove the code that it was testing.
Empty arguments can be passed to macros per C99 and C++11 specs, which
can then be forwarded to other macros.
Visual Studio's compiler has a bug in the following situation:
#define A(x) #x
#define B(x, y) A(x)
B(, b)
In the above case, Visual Studio first expands x to nothing, then
complains that A is not invoked with the right amount of arguments.
However, x in A(x) is still one argument, even if it expands to no
preprocessing tokens. See also https://stackoverflow.com/a/7674214.
Detect when C++ parametric tests (TEST_P) are not instantiated.
When an un-instantiated TEST_P is found, a new test will be inserted that will emit a warning message.
This can be made to error with minor code edits.
In the future, that is intended to be the default.
PiperOrigin-RevId: 284901666
Change googletest to notice failures during SetUpTestSuite() and TearDownTestSuite().
Previously, errors that occurred during those functions were logged but otherwise ignored. After this change, such failures will cause the test to fail and a brief summary will be printed at the bottom of the test log.
See https://github.com/google/googletest/issues/2330.
PiperOrigin-RevId: 284033342
Make generating goldens work with Python 3
golden_file.write(output) fails with Python 3 with:
TypeError: a bytes-like object is required, not 'str'
PiperOrigin-RevId: 282316101
Use standard C++11 integer types in gtest-port.h.
Remove testing::internal::{Int,Uint}{32,64} in favor of types
guaranteed to be in <cstdint> since C++11.
Tests for built-in integer type coverage are switched from
{Int,Uint}64 to [unsigned] long long, which is guaranteed by
C++11 to exist and be at least 64-bit wide.
PiperOrigin-RevId: 281565263
We use "MemoryIsNotDeallocated" to aovid internal expected leak reported
in Windows _Crt report, like:
{
#ifdef _MSC_VER
MemoryIsNotDeallocated memory_is_not_deeallocated;
#endif
static ThreadIdToThreadLocals* map = new
ThreadIdToThreadLocals();
return map;
}
But int the above code, only "new ThreadIdToThreadLocals()" is
protected, if we invoke "insert()" function of the return value,
the memory allocated in "insert()" will be reported to _Crt report
also. This change try to fix this issue.
googletest 1.10.0 fails to compile on msvc version 19.00.23917
with one compilation error:
src\googletest\include\gtest\internal\gtest-internal.h(1188) : error C2039:
'FlatTupleBase<testing::internal::FlatTuple<bool,bool>,testing::internal::IndexSequence<0,1> >':
is not a member of 'testing::internal::FlatTuple<bool,bool>'
This PR fixes the compilation error by explicitly specifying the full type that Indices is
located in the base type.
Correctly deal with stringification, and forbid empty arguments where they could slip thought, in the type parameterized test API.
Note: even where empty args work, it's likely to result in technically invalid code by virtue of creating reserved identifiers:
https://en.cppreference.com/w/cpp/language/identifiers
PiperOrigin-RevId: 279330971
Clone+exec death test allocates a single page of stack to run chdir + exec on.
This is not enough when gtest is built with ASan and run on particular
hardware.
With ASan on x86_64, ExecDeathTestChildMain has frame size of 1728 bytes.
Call to chdir() in ExecDeathTestChildMain ends up in
_dl_runtime_resolve_xsavec, which attempts to save register state on the stack;
according to cpuid(0xd) XSAVE register save area size is 2568 on my machine.
This results in something like this in all death tests:
Result: died but not with expected error.
...
[ DEATH ] AddressSanitizer:DEADLYSIGNAL
[ DEATH ] =================================================================
[ DEATH ] ==178637==ERROR: AddressSanitizer: stack-overflow on address ...
PiperOrigin-RevId: 278709790
Remove the threads link library variable references from the pkg-config
Cflags: field, removing -lpthread(s) from the compile flags.
"-l*" linker flags should only be part of the Libs: section and should
not be part of the Cflags: section in pkg-config files.
This was first suggested in
https://github.com/google/googletest/pull/2006 and further discussed in
https://github.com/google/googletest/pull/2483 .
Add a matcher `testing::ReturnRoundRobin` which, on each call, returns the next element in the sequence, restarting at the beginning once it has reached the end.
PiperOrigin-RevId: 276312136
Split the scoped trace examples into two snippets.
This doesn't have an effect in the github markdown renderer, but in some other renderers/templates, this leads them to be independently copy-pastable. In particular, the markdown rendering that Google uses internally has a copy button for each code snippet, which, before this change, would copy both lines. Bad copy button, no cookie! After this change, there will be two such buttons, one per line, and clicking a copy button will copy only the one snippet it is next to, and not the other. This is desirable because nobody will ever want to copy both lines, only one or the other.
PiperOrigin-RevId: 276079009
Remove bool_constant in favor of std::integral_constant<bool, ...>;
The one non-trivial use of bool_constant has been changed to have significantly
fewer template specializations.
PiperOrigin-RevId: 275842490
Closes#2208
Previously, skip messages were invisible, so debugging skips was hard.
Now we have this:
```
$ ./googletest/gtest_skip_test
Running main() from /home/lesha/github/snarkmaster/googletest/googletest/src/gtest_main.cc
[==========] Running 3 tests from 2 test suites.
[----------] Global test environment set-up.
[----------] 1 test from SkipTest
[ RUN ] SkipTest.DoesSkip
/home/lesha/github/snarkmaster/googletest/googletest/test/gtest_skip_test.cc:38: Skipped
skipping single test
[ SKIPPED ] SkipTest.DoesSkip (0 ms)
[----------] 1 test from SkipTest (1 ms total)
...
```
Remove a test case rendered obsolete by disallowing empty argument for INSTANTIATE_TEST_SUITE_P.
Remove the code that it was testing.
PiperOrigin-RevId: 275040108
