Create common test static lib and move UTs to TAEF by bbowman · Pull Request #1849 · microsoft/WinObjC

bbowman · 2017-01-27T02:12:50Z

No description provided.

bbowman · 2017-01-27T02:15:38Z

#include <TestFramework.h>

I don't know what an .inl file is??? This looked like it needed to change though.

Refers to: tests/unittests/Starboard/StarboardAutoIdTests.inl:17 in 53f7de8. [](commit_id = 53f7de8, deletion_comment = False)

bbowman · 2017-01-27T02:16:31Z

@@ -0,0 +1,572 @@
+//******************************************************************************


This file needs the most review.

bbowman · 2017-01-27T02:17:25Z

Need to adjust the scripts still so that build definitions don't break. May be a bit non trivial since the output is a different format #Resolved

pradipd · 2017-01-27T02:25:28Z

Actually, we need 2 things:

We need to get wttlog.dll from somewhere to create output files.
We need a way to upload the results to VSO.

I need to do the same things for getting functional tests running on hte build machine.
So, please hold off merging before I figure that out. Hopefully I'll have it figured out tomorrow (we already do this in the lab).

In reply to: 275573399 [](ancestors = 275573399)

pradipd · 2017-01-27T02:26:12Z

Also, #1 is harder than #2.

In reply to: 275574352 [](ancestors = 275574352,275573399)

bbowman · 2017-01-27T02:32:57Z

EndProject

CoreGraphics, CoreImage, and COreGraphcis.Drawing had a different OutDir. This should be standardized. I need to figure out what I missed. #Resolved

Refers to: build/build.sln:90 in 53f7de8. [](commit_id = 53f7de8, deletion_comment = False)

bbowman · 2017-01-27T02:35:07Z

<OutDir>$(SolutionDir)$(Platform)\$(Configuration)\$(RootNamespace)</OutDir>

Remove. I missed a few. The new outDir is /build/Release or /build/Debug. Perhaps we should do $(SolutionDir)$(Platform)$(Configuration)\ so its close to the same? It just needs to be different from the frameworks out dir so that the package .zip stuff (soon to be removed) doesn't pick up uts. This means we will have 1 copy of the binaries around to support the UTs but that seems a lot better.

#Resolved #Resolved

Refers to: build/Tests/UnitTests/CoreImage/CoreImage.UnitTests.vcxproj:73 in 53f7de8. [](commit_id = 53f7de8, deletion_comment = False)

pradipd · 2017-01-27T02:53:38Z

Also, sorry, I didn't think of this sooner....

In reply to: 275574435 [](ancestors = 275574435,275574352,275573399)

pradipd · 2017-01-27T03:07:20Z

  </ImportGroup>
-
-  <PropertyGroup>
-    <StripAppContainerBit>true</StripAppContainerBit>


StripAppContainerBit [](start = 5, length = 20)

:) #Closed

DHowett-MSFT

This is a partial review until I get in tomorrow.

DHowett-MSFT · 2017-01-27T03:16:55Z

 #include "EbrStorageFile.h"

+#include <COMIncludes.h>
+#include <wrl\client.h>


nit: standardize on /; we do that elsewhere in WOC code #Resolved

DHowett-MSFT · 2017-01-27T03:18:26Z

      <GenerateDebugInformation>true</GenerateDebugInformation>
      <AdditionalDependencies>mincore.lib;libxml2.lib;icudt.lib;icuin.lib;icuuc.lib;libdispatch.lib;%(AdditionalDependencies)</AdditionalDependencies>
-      <AdditionalLibraryDirectories>$(StarboardBasePath)\Frameworks\limbo;$(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
+      <AdditionalLibraryDirectories>$(StarboardBasePath)\Frameworks\limbo;%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>


lol?!?! #Closed

How did this ever work? #Resolved

I mean I think it just ignores it if there is nothing there. WHich is true now. I forget what used to be there that CF needed. Scary indeed. #WontFix

DHowett-MSFT · 2017-01-27T03:20:35Z

  <ItemGroup>
-    <ClCompile Include="$(StarboardBasePath)\tests\unittests\Framework\Framework.cpp" />
+
    <ClangCompile Include="$(StarboardBasePath)\tests\unittests\CoreGraphics.Drawing\EntryPoint.cpp" />


This guy uses a custom entry point so that it can take command line arguments. Is there going to be a way to pass in run-specific state with TAEF/TE?

It uses this state for:

Generation mode (writing images to disk without testing)

Comparison against a separate image root

Yes but I was a little lazy. Its called RuntimeParameters. YOu can see it for the test directory stuff. We'd want to hide away those things behind a macro / function probably.

DHowett-MSFT · 2017-01-27T03:26:25Z

-                                                                   test_info->test_case_name(),
-                                                                   test_info->name(),
+                                                                   CFSTR("TestImage.%s%s%@.png"),
+                                                                   GET_TEST_FULL_NAME().c_str(),


I cannot place a comment on the appropriate line since it's not in the diff.

DrawingTest uses RecordProperty to poke URL references out of the test output via XML.
A modified test runner can use these to display the images inline in its UI.

What's the TAEF equivalent? #Resolved

I haven't looked through how we're using TAEF logging, but Log::File(L"path to file") would result in the image being saved under TAEF's current output directory.

From the TAEF docs;

Log:: File(const wchar_t* pszFileName)
Log a test file to be saved. Files are saved to a folder named "WexLogFileOutput". This folder will be in the folder specified by the WTTRunWorkingDir environment variable if that variable exists. When using Wex.Logger outside of TAEF and the WTTRunWorkingDir environment variable does not exist, the WexLogFileOutput folder will be in the current directory. When using Wex.Logger along with TAEF and the WTTRunWorkingDir environment variable does not exist, the WexLogFileOutput folder will be in the TAEF output folder.

If you just want to log the path to the file, you could Log::Property(const wchar_t* pszName, const wchar_t* pszValue). Of course, we'll need wttlog.dll to generate test logs.

From the TAEF docs;

Log::Property(const wchar_t* pszName, const wchar_t* pszValue)
Log a name/value property pair. The value can be in xml format.

In reply to: 98139951 [](ancestors = 98139951)

Yeah sorry. I saw this but forgot to go make a macro for it. We'd want Log::Property I think. I'll make a macro for it. #Resolved

DHowett-MSFT · 2017-01-27T03:26:41Z

+#ifndef GET_TEST_FULL_NAME
+#define GET_TEST_FULL_NAME                                                                \
+    []() {                                                                                \
+        std::string alternateTestName = _ExchangeAlternateTestName(std::string(), false); \


This scares me, but I'll have to look a little more at why.

I agree. I was trying to avoid having to use a .cpp for this but I remembered that I already did that work for the logger so I could make this a bit nicer my having a class with more sane storage and helpers.

DHowett-MSFT · 2017-01-27T03:28:03Z

+#endif
    testing::InitGoogleTest(&argc, argv);

    _configImpl = std::move(std::make_shared<CommandLineDrawingTestConfigImpl>(argc, argv));


Wait whoa no way, TE can pass args in argc/v?!

no. it can't. this is me stamping out argc/v to interop with the below call. See my comment on runtime params.

DHowett-MSFT · 2017-01-27T03:31:04Z

@@ -27,7 +28,8 @@
 #include "TestUtils.h"

 TEST(NSFileManager, GetAttributes) {
-    NSString* testFileFullPath = getPathToFile(@"/data/NSFileManagerUT.txt");
+    NSString* testFileFullPath =
+        [[NSString stringWithUTF8String:GET_CURRENT_TEST_DIRECTORY().c_str()] stringByAppendingPathComponent:@"/data/NSFileManagerUT.txt"];


I still like the idea of there being one nice shared thing that does this.. but for "lowest common denominator" support it would have to be either CFString or std::string/std::wstring returning. The CG.Drawing tests run as C++ only.

I'm not sure what you mean... this is a std::string returning thing. The thing is just a bit weird. Are you wanting a more formalized class or something? I agree that the macro "function" isn't super clean ...

aballway

🔑

aballway · 2017-01-27T04:06:57Z

      <GenerateDebugInformation>true</GenerateDebugInformation>
      <AdditionalDependencies>mincore.lib;libxml2.lib;icudt.lib;icuin.lib;icuuc.lib;libdispatch.lib;%(AdditionalDependencies)</AdditionalDependencies>
-      <AdditionalLibraryDirectories>$(StarboardBasePath)\Frameworks\limbo;$(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
+      <AdditionalLibraryDirectories>$(StarboardBasePath)\Frameworks\limbo;%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>


limbo? #Closed

yeah don't worry about it. There has been a lot of cleanup of the old code. You don't know half of the horrors of this project ;-) #WontFix

aballway · 2017-01-27T04:15:52Z


+  std::vector<std::pair<std::string, TestMetaFactoryBase<typename TestCase::ParamType>*>> GetTestMetaFactories() {
+    std::vector<std::pair<std::string, TestMetaFactoryBase<typename TestCase::ParamType>*>> toReturn;
+    for (typename TestInfoContainer::iterator test_it = tests_.begin(); test_it != tests_.end(); ++test_it) {


nit: can make this range-based for? #Resolved

+1, also auto would be nice

In reply to: 98142566 [](ancestors = 98142566)

I'll try but I think I did that and it was complaining about needing the typename stuff. Maybe auto will work though.

aballway · 2017-01-27T04:17:21Z

+    };
+
+    // Get StorageFile reference for file in local folder
+    [WSStorageFile


I feel like I'm looking at node :(

what do you mean? Is there something better I should be doing?

No I don't think so, just oddly reminiscent of callbackapalooza

aballway · 2017-01-27T04:18:42Z

+
+    TEST_CLASS_SETUP(AssetsLibraryClassSetup) {
+        bool success = AssetsLibraryTestVideoSetup("AssetsLibraryTestVideo.mp4");
+        success &= FunctionalTestSetupUIApplication();


Assuming we don't actually run the test if TEST_CLASS_SETUP fails, should we use the non-short circuiting &=? #WontFix

I dunno this is copied from the current behavior in develop. I tried to not actually change the functional tests just consolidate them. But I may have messed up. You make a good point about if setup fails though to not run the test. I don't do that right in my test_p shim ... #WontFix

aballway · 2017-01-27T04:19:26Z

+//
+// Cortana Activation Tests
+//
+class CortanaVoiceCommandForegroundActivation {


takes me back 😿 #Closed

jaredhms · 2017-01-27T15:59:28Z

+  std::vector<std::pair<std::string, TestMetaFactoryBase<typename TestCase::ParamType>*>> GetTestMetaFactories() {
+    std::vector<std::pair<std::string, TestMetaFactoryBase<typename TestCase::ParamType>*>> toReturn;
+    for (typename TestInfoContainer::iterator test_it = tests_.begin(); test_it != tests_.end(); ++test_it) {
+      linked_ptr<TestInfo> test_info = *test_it;


linked_ptr [](start = 6, length = 20)

nit: linked_ptr& #Resolved

jaredhms · 2017-01-27T16:08:27Z

+    // Pass null to indicate default app and app delegate classes
+    UIApplicationInitialize(nullptr, nullptr);
+}
+


this shouldn't be resurrected; we now call UIApplicationInitialize in FunctionalTestHelpers::FunctionalTestSetupUIApplication() #Resolved

yeah sorry bad merge on this file. I just wanted to delete everything in it basically but y'all made a bunch of changes to it. #Resolved

jaredhms · 2017-01-27T16:09:40Z

+
+class ActivatedAppReceivesToastNotificationTests {
+    BEGIN_TEST_CLASS(ActivatedAppReceivesToastNotificationTests)
+    END_TEST_CLASS()


nit: anywhere we're not adding class-level metadata, we can just use TEST_CLASS(ActivatedAppReceivesToastNotificationTests) (no need for begin/end) #WontFix

I know but I'm not changing how it was for people already. This is just transplanted code. #lazy. #WontFix

jaredhms · 2017-01-27T16:15:48Z

+    //     2. Use the same mechanism as in TEST_METHOD below to export a method from the WinObjC test file and call it here.
+    //     3. If you do not need this functionality feel free to remove this for your test class.
+    TEST_CLASS_CLEANUP(SampleTestClassCleanup) {
+        return true;


This should return FunctionalTestCleanupUIApplication #Resolved

kk #Resolved

msft-Jeyaram · 2017-01-27T19:08:04Z

+        _EbrGetBaseWriteablePath

        ; REMOVE / CLEANUP C++ EXPORTS
        ?format@string@woc@@YA?AV?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@PBDZZ


format@string@woc@@ya?AV?$basic_string@DU?$char_traits@D@std@@v?$allocator@D@2@@std@@Pbdzz [](start = 9, length = 90)

I assume this mangled export is still needed? #WontFix

probably? Despite outward appearances, my goal is not to fix everything potentially wrong with this project. #WontFix

Although somewhat disappointing.

msft-Jeyaram · 2017-01-27T19:16:06Z

// Copyright (c) 2016 Microsoft Corporation. All rights reserved.

nit: update this #Resolved

Refers to: Frameworks/include/ErrorHandling.h:3 in 53f7de8. [](commit_id = 53f7de8, deletion_comment = False)

bbowman · 2017-01-27T23:01:38Z

+                    } \
+                    catch (const ::WEX::TestExecution::VerifyFailureException&) \
+                    { \
+                    } \


@DHowett-MSFT @jaredms this means that I should bail out and log a failure here right? Test Cleanup should still run though? Basically put the TestBody in the first try block?

Yeah, I think so. gtest tries them all (?) in their own exception handlers i think.

bbowman · 2017-01-27T23:03:14Z

+                    auto factory = metaFactory.second->CreateTestFactory(*param); \
+                    auto test = factory->CreateTest(); \
+                    \
+                    WEX::Logging::Log::StartGroup(currentTest.c_str()); \


WEX::Logging::Log::StartGroup(currentTest.c_str()); \ [](start = 20, length = 53)

would be useful to also log what the parameters used here are. I think gtest has a method for that somewhere. I dunno if I can get to it off of the ParamGenerator though?

bbowman · 2017-01-27T23:04:02Z

+                 size_t i = 0; \
+                 static std::wstring_convert<std::codecvt_utf8_utf16<wchar_t>> converter; \
+                 static const std::wstring c_stringPrefix = L #prefix L"_" L#className L"." + converter.from_bytes(metaFactory.first.c_str()); \
+                 for (typename ::testing::internal::ParamGenerator<className::ParamType>::iterator param = paramGenerator.begin(); param != paramGenerator.end(); ++param, ++i) { \


ypename ::testing::internal::ParamGeneratorclassName::ParamType::iterator [](start = 23, length = 76)

probably use auto& here?

pradipd · 2017-01-30T18:43:29Z

 <#
 .SYNOPSIS
-    Runs Tests for Islandwood project.
+    Runs Functional Tests for Islandwood project.


Can we just merge this with run-functionaltests? Would be nice to only "maintain" 1 script.
Granted, once these are written, they usually don't have to be modified.

pradipd · 2017-02-07T07:03:59Z

This is all sorted out.
You should be able to use the Run TAEF and Publish Task in VSTS.

In reply to: 275577758 [](ancestors = 275577758,275574435,275574352,275573399)

bbowman · 2017-02-08T23:04:25Z

//******************************************************************************

Looks like the test_p benchmark tests fail... Is it possible that you aren't using standard GTEST macros for that?

Refers to: tests/Benchmark/EntryPoint.cpp:1 in a3246d1. [](commit_id = a3246d1, deletion_comment = False)

DHowett-MSFT · 2017-02-08T23:10:14Z

@@ -54,7 +48,22 @@ class CommandLineDrawingTestConfigImpl : public DrawingTestConfigImpl {

 public:
    CommandLineDrawingTestConfigImpl(int argc, char** argv)


the TestConfig base class is pure virtual so that we don't need to clowncar all of the TAEF behaviours into the "command line" one ;P

like: there are so many ifdefs in this file it might simply be easier to have a win32 file that has its own config impl and the os x file that keeps the commandline one.

I agree in principle. I'll let you handle that. There were already a number of ifdefs in there for WIN32 so it seems like we need to think more about how to divide up the entrypoint code for the two.

DHowett-MSFT · 2017-02-08T23:12:45Z

+        std::string tempBuffer;
+        uint32_t maxPath = _MAX_PATH;
+        tempBuffer.resize(maxPath);
+        _NSGetExecutablePath(&tempBuffer[0], &maxPath);


This introduces a dependency on Foundation, which not all of our test modules can take. We also cannot take that dependency on OS X.

This is only for OSX and it is what was happening before anyway somewhere ... I didn't make up that code at least.

aballway · 2017-02-08T23:13:32Z

The only macros it uses are TEST_P and INSTANTIATE_TEST_CASE_P, but it still has the GTEST current_test_info()->name(), which is probably causing the crash. I'm building it now so I'll see if there are any other things that would cause an issue. Thanks for fixing everything else up!

DHowett-MSFT · 2017-02-08T23:13:47Z

+__if_not_exists(GetTestFullName) {
+    static std::string GetTestFullName() {
+        std::string toReturn;
+        return toReturn + ::testing::UnitTest::GetInstance()->current_test_info()->test_case_name() + "." +


Why add toReturn+ here? It looks like it's being done for std::string coercion? That's not great.

Yeah that is what it was for. I agree not great. I'll change.

bbowman · 2017-02-08T23:15:34Z

I changed the test info name dealio and it seems to all pass for whatever that means...

In reply to: 278493789 [](ancestors = 278493789)

DHowett-MSFT · 2017-02-08T23:14:58Z

 *.xcf filter=lfs diff=lfs merge=lfs -text
 *.gif filter=lfs diff=lfs merge=lfs -text
 *.ttf filter=lfs diff=lfs merge=lfs -text
+*.nupkg filter=lfs diff=lfs merge=lfs -text


Is this intended to be part of this change or part of the packaging change?

This change does bring in the TAEF nupkg.

DHowett-MSFT · 2017-02-08T23:15:52Z

-const std::wstring& _IwGetWritableFolder() {
-    return g_WritableFolder;
+    unsigned int rawLength;
+    return WindowsGetStringRawBuffer(basePath.Get(), &rawLength);


length is marked _Out_opt_ in the header; don't make it populate something you're discarding immediately.

DHowett-MSFT · 2017-02-08T23:16:09Z


-    // Set our writable and temp folders
-    IwSetWritableFolder(Windows::Storage::ApplicationData::Current->LocalFolder->Path->Data());
+    // Set our temp folders


nit: not plural anymore

with the single line code, the comment is not necessary anymore (in the past, this used to be several lines :))

DHowett-MSFT · 2017-02-08T23:17:10Z


+#include <tuple>
+
 // Copying these macros from WexTestClass.h, but not including


Based on this comment, the following macros could simply be destroyed? Or does this exist for cross-platform support?

I didn't look at that. I assume those run cross-plat so i think we should keep as is.

DHowett-MSFT · 2017-02-08T23:18:24Z

 SB_EXPORT int EbrGetWantedOrientation();

 SB_EXPORT const wchar_t* IwGetWritableFolder();
-SB_EXPORT void IwSetWritableFolder(const wchar_t* folder);


I'd personally prefer to see all the Iw Writable Folder changes in a separate pull request with a separate commit (instead of squashed all together at the end). It will help future bisecting if we regress something.

In fact, I think all framework changes should be separate from this testing-related PR.

DHowett-MSFT · 2017-02-08T23:22:09Z

+
+// TODO::
+// todo-nithishm-04182016 - Investigate a better way to support ASSERT_*_MSG and EXPECT_*_MSG macros without re-defining them
+// here. Bug 7244185 is tracking this work.


Should we port this VSO item to a GitHub item?

Just remove the TODO. If we didn't need it after so long, we don't really need it :)

DHowett-MSFT · 2017-02-08T23:23:00Z

+#define FRIEND_TEST(test_case_name, test_name) friend class test_case_name##_##test_name
+
+// clang-format off
+template <unsigned NumChar>


A comment explaining this and why we need it would be excellent.

DHowett-MSFT · 2017-02-08T23:23:11Z

+            str1[7] == 'D') ? L"TRUE" : L"FALSE";
+}
+
+__declspec(noinline) inline std::string _ExchangeAlternateTestName(


??? noinline inline

extra special inlining. It means __declspec(selectany) for inline functions.

https://blogs.msdn.microsoft.com/freik/2005/10/26/how-to-make-a-no-inline-inline-function-and-why-you-might-want-to/ has a good summary.

The gist of it is that I'm using a global variable that needs to be shared across TUs (DrawingTest and the Test file that derives from it). This means I can't have more than one static function stamped out / use an inline function because the storage for the alternate name wouldn't be the same across TUs. However, I wanted to keep this in the header to minimize the additional goop needed by the tests (an alternate solution would be to place this stuff in the logger static lib). This means I need a "selectAny" type of behavior which you get from this hilarious syntax.

DHowett-MSFT · 2017-02-08T23:24:10Z

+                    } \
+                    catch (const ::WEX::TestExecution::VerifyFailureException&) \
+                    { \
+                    } \


Yeah, I think so. gtest tries them all (?) in their own exception handlers i think.

DHowett-MSFT · 2017-02-08T23:25:53Z

+    {                                                                            \
+        static std::wstring_convert<std::codecvt_utf8_utf16<wchar_t>> converter; \
+        std::wstring wideKey = converter.from_bytes(key);                        \
+        std::wstring wideValue = converter.from_bytes(value);                    \


Does Log::Property requiretypeof(value) == std::wstring? If not, and it's overloaded, we should pass value through unharmed for the appropriate implicit conversions to happen. This is too narrowly-specified as-is, it will only work for utf8 strings.

I believe it requires wchar_t*s

bbowman · 2017-02-08T23:35:46Z

    const char* testName = _STRINGIFY(__##test_class);                                                                  \
-    auto fullName = std::string(::testing::UnitTest::GetInstance()->current_test_info()->name()) + testName;            \
-    ::benchmark::internal::__RunCase(*m_test, fullName.c_str());                                                        \
+    ::benchmark::internal::__RunCase(*m_test, GetTestFullName().c_str());                                               \


GetTestFullName() [](start = 46, length = 17)

I'll let you figure that out later. I don't think you do / you may want to rethink that.

👍 TestFullName is the value this needs, you're right

DHowett-MSFT · 2017-02-08T23:36:55Z

(please file issues for anything you're marking as "you can deal with this later"; otherwise, we will all collectively lose track of them until the one day where we're like "hey wait this bug seems familiar ...")

bbowman · 2017-02-08T23:38:40Z

yep. Currently planning to file one on DrawingTests to integrate with the change (seems like we need to validate the arguments, perhaps adjust the ifdefs, and rework the test runner) and one for Benchmark tests

In reply to: 278498755 [](ancestors = 278498755)

rajsesh · 2017-02-09T00:37:30Z

 *.xcf filter=lfs diff=lfs merge=lfs -text
 *.gif filter=lfs diff=lfs merge=lfs -text
 *.ttf filter=lfs diff=lfs merge=lfs -text
+*.nupkg filter=lfs diff=lfs merge=lfs -text


This change does bring in the TAEF nupkg.

rajsesh · 2017-02-09T00:40:15Z


-    // Set our writable and temp folders
-    IwSetWritableFolder(Windows::Storage::ApplicationData::Current->LocalFolder->Path->Data());
+    // Set our temp folders


with the single line code, the comment is not necessary anymore (in the past, this used to be several lines :))

rajsesh · 2017-02-09T00:52:49Z

+using namespace Microsoft::WRL;
+using namespace Windows::Foundation;
+
+Wrappers::HString GetAppDataPath() {


_IwGetWritablePath() won't work here?

It will but I didn't really want to introduce that dependency to Starboard since its a weird private function anyway. I can though if we feel thats nicer? #Pending

Your choice, it was a nit from my side.

rajsesh · 2017-02-09T00:55:50Z

 # Decide where the WTL output files will live
 if ($WTLOutputDirectory -ne [string]$null)
 {
+    $WTTLogFullPath = Join-Path $WTTLogPath wttlog.dll 


Where would one get wttlog.dll?

from a valid location. We can take this offline. #Closed

pradipd · 2017-02-09T09:09:27Z

+ExecTest $argList

-exit $failureCount + $crashingTestArray.length
+exit


exit $script:exitCode

https://winobjc.visualstudio.com/WinObjC/_build/explorer?_a=summary&definitionType=2&buildId=7825

fails. It has some errors from CG.

In reply to: 100258200 [](ancestors = 100258200)

pradipd · 2017-02-09T18:55:40Z

DHowett-MSFT

I'd still like to see the non-testing changes moved out into another pull request. I want that so much that I'm even willing to do the work necessary to tease them apart. 😦

… covers

DHowett-MSFT · 2017-02-10T05:40:36Z

Woot

msftclas added the cla-not-required label Jan 27, 2017

bbowman requested review from DHowett-MSFT, aballway, jaredhms, ms-jihua, oliversa-msft and rajsesh January 27, 2017 02:12

bbowman assigned aballway, jaredhms, DHowett-MSFT, ms-jihua, msft-Jeyaram and oliversa-msft Jan 27, 2017

bbowman commented Jan 27, 2017

View reviewed changes

pradipd reviewed Jan 27, 2017

View reviewed changes

DHowett-MSFT reviewed Jan 27, 2017

View reviewed changes

aballway approved these changes Jan 27, 2017

View reviewed changes

jaredhms reviewed Jan 27, 2017

View reviewed changes

msft-Jeyaram reviewed Jan 27, 2017

View reviewed changes

bbowman commented Jan 27, 2017

View reviewed changes

pradipd reviewed Jan 30, 2017

View reviewed changes

aballway mentioned this pull request Feb 1, 2017

Add benchmarking test project and add text drawing benchmark tests #1876

Merged

bbowman force-pushed the bbowman/updateTAEF+RedoUTs branch from 7fe162c to 1a7732e Compare February 8, 2017 22:13

DHowett-MSFT reviewed Feb 8, 2017

View reviewed changes

DHowett-MSFT suggested changes Feb 8, 2017

View reviewed changes

bbowman commented Feb 8, 2017

View reviewed changes

rajsesh reviewed Feb 9, 2017

View reviewed changes

rajsesh approved these changes Feb 9, 2017

View reviewed changes

pradipd reviewed Feb 9, 2017

View reviewed changes

DHowett-MSFT approved these changes Feb 9, 2017

View reviewed changes

Framework changes as a prestep towards supporting TAEF tests.

70aa641

bbowman force-pushed the bbowman/updateTAEF+RedoUTs branch from e2bbce9 to eba6f61 Compare February 9, 2017 22:04

Add a common test static lib and move all tests to use TAEF under the…

9bdbefd

… covers

bbowman force-pushed the bbowman/updateTAEF+RedoUTs branch from eba6f61 to 9bdbefd Compare February 9, 2017 22:23

Small fixes in UTs for debug.

4940d88

bbowman force-pushed the bbowman/updateTAEF+RedoUTs branch from a5f35bf to 4940d88 Compare February 10, 2017 02:59

bbowman merged commit 58e461e into microsoft:develop Feb 10, 2017

		@@ -0,0 +1,572 @@
		//******************************************************************************

		@@ -54,7 +48,22 @@ class CommandLineDrawingTestConfigImpl : public DrawingTestConfigImpl {

		public:
		CommandLineDrawingTestConfigImpl(int argc, char** argv)


		#include <tuple>

		// Copying these macros from WexTestClass.h, but not including

Conversation

bbowman commented Jan 27, 2017

Uh oh!

bbowman commented Jan 27, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bbowman commented Jan 27, 2017 • edited by pradipd Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pradipd commented Jan 27, 2017

Uh oh!

pradipd commented Jan 27, 2017

Uh oh!

bbowman commented Jan 27, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bbowman commented Jan 27, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pradipd commented Jan 27, 2017

Uh oh!

pradipd Jan 27, 2017 • edited by bbowman Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

DHowett-MSFT left a comment

Choose a reason for hiding this comment

Uh oh!

DHowett-MSFT Jan 27, 2017 • edited by bbowman Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

DHowett-MSFT Jan 27, 2017 • edited by bbowman Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

DHowett-MSFT Jan 27, 2017 • edited by bbowman Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bbowman Jan 27, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

DHowett-MSFT Jan 27, 2017 • edited by bbowman Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bbowman Jan 27, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

aballway left a comment

Choose a reason for hiding this comment

Uh oh!

aballway Jan 27, 2017 • edited by bbowman Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bbowman Jan 27, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

bbowman commented Jan 27, 2017 •

edited by pradipd

Loading

bbowman commented Jan 27, 2017 •

edited

Loading

bbowman commented Jan 27, 2017 •

edited

Loading

pradipd Jan 27, 2017 •

edited by bbowman

Loading

DHowett-MSFT Jan 27, 2017 •

edited by bbowman

Loading

DHowett-MSFT Jan 27, 2017 •

edited by bbowman

Loading

DHowett-MSFT Jan 27, 2017 •

edited by bbowman

Loading

bbowman Jan 27, 2017 •

edited

Loading

DHowett-MSFT Jan 27, 2017 •

edited by bbowman

Loading

bbowman Jan 27, 2017 •

edited

Loading

aballway Jan 27, 2017 •

edited by bbowman

Loading

bbowman Jan 27, 2017 •

edited

Loading

aballway Jan 27, 2017 •

edited by bbowman

Loading

aballway Jan 27, 2017 •

edited by bbowman

Loading

bbowman Jan 27, 2017 •

edited

Loading

aballway Jan 27, 2017 •

edited by bbowman

Loading

jaredhms Jan 27, 2017 •

edited by bbowman

Loading

jaredhms Jan 27, 2017 •

edited by bbowman

Loading

bbowman Jan 27, 2017 •

edited

Loading

jaredhms Jan 27, 2017 •

edited by bbowman

Loading

bbowman Jan 27, 2017 •

edited

Loading