Merge pull request #20 from ethz-asl/fix/use_std_traits_and_documenta…

…tion

Use std traits to remove reference and some more documentation

README.md

@@ -1,4 +1,4 @@
-eigen_checks
+Eigen check macros
 ============
 
 A repository for glog CHECK and gtest ASSERT and EXPECT macros for Eigen types.
@@ -32,6 +32,55 @@ Checks if two matrices are equal to a user-specified precision.
 ## GTEST
 The gtest macros are built to facilitate unit testing with matrix types. This library provides two pieces of functionality: an macro that defines the main function, or *entrypoint*, for a guest invocation, and several macros for testing if matrices are similar.
 
+### GTest 
+The Gtest `EXPECT()` and `ASSERT()` macros are used to verify assumptions in your tests and print useful information in case these don't match the expected outcome. These Eigen equivalents can be used to check conditions on matrices. See [the gtest documentation](https://code.google.com/p/googletest/wiki/Primer#Assertions) for a full description. The general form of use is:
+
+```c++
+EXPECT_TRUE(EIGEN_MATRIX_EQUAL(MatrixA, MatrixB));
+EXPECT_FALSE(EIGEN_MATRIX_EQUAL(MatrixA, MatrixB));
+ASSERT_TRUE(EIGEN_MATRIX_EQUAL(MatrixA, MatrixB));
+ASSERT_FALSE(EIGEN_MATRIX_EQUAL(MatrixA, MatrixB));
+```
+
+All tests happen component-wise. They fail if the matrices are different sizes. If the matrices are the same size, the test is applied to each corresponding pair of components.
+
+To get these macros, use:
+
+```c++
+#include<eigen-checks/gtest.h>
+```
+
+### `EXPECT_TRUE(EIGEN_MATRIX_EQUAL(MatrixA, MatrixB))`
+
+Succeeds if two matrices are binary equal.
+
+### `EXPECT_TRUE(EIGEN_MATRIX_EQUAL_DOUBLE(MatrixA, MatrixB))`
+
+Succeeds if two matrices are equal to floating-point precision
+
+### `EXPECT_TRUE(EIGEN_MATRIX_NEAR(MatrixA, MatrixB, Precision))`
+
+Succeeds if two matrices are equal to a user-specified precision.
+
+
 ### Entrypoint
 
-### Macros 
+Instead of writing the following code in your unit-test runner file:
+
+```c++
+#include <gtest/gtest.h>
+int main(int argc, char** argv) {
+  testing::InitGoogleTest(&argc, argv);
+  google::ParseCommandLineFlags(&argc, &argv, true);
+  ::testing::FLAGS_gtest_death_test_style = "threadsafe";
+  //...
+  return RUN_ALL_TESTS();
+}
+```
+You can just use:
+
+```c++
+#include <eigen-checks/entrypoint.h>
+
+UNITTEST_ENTRYPOINT
+```

include/eigen-checks/glog.h

@@ -1,9 +1,10 @@
 #ifndef EIGEN_CHECKS_GLOG_H_
 #define EIGEN_CHECKS_GLOG_H_
-#include <glog/logging.h>
 #include <limits>
+#include <type_traits>
+
 #include <eigen-checks/internal/gtest-equal.h>
-#include <eigen-checks/internal/traits.h>
+#include <glog/logging.h>
 
 // To allow the user to append custom messages we store the result of the
 // verification inside a unique local variable to be used in the check and for
@@ -22,14 +23,14 @@
   INTERNAL_EIGEN_CHECKS_MAKE_GLOG_CHECK(                                     \
     eigen_checks::internal::MatricesNear(                                    \
     MatrixA, #MatrixA, MatrixB, #MatrixB,                                    \
-    static_cast<typename eigen_checks::internal::RemoveCR<                   \
+    static_cast<typename std::remove_reference<                              \
     decltype(MatrixA)>::type::Scalar>(0.0), "0.0"))
 
 #define CHECK_EIGEN_MATRIX_EQUAL_DOUBLE(MatrixA, MatrixB)                    \
   INTERNAL_EIGEN_CHECKS_MAKE_GLOG_CHECK(                                     \
       eigen_checks::internal::MatricesNear(                                  \
         MatrixA, #MatrixA, MatrixB, #MatrixB,                                \
-        static_cast<typename eigen_checks::internal::RemoveCR<               \
+        static_cast<typename std::remove_reference<                          \
         decltype(MatrixA)>::type::Scalar>(                                   \
       eigen_checks::internal::kDefaultPrecision),                            \
       "Floating point precision"))

include/eigen-checks/gtest.h

@@ -1,17 +1,18 @@
 #ifndef EIGEN_CHECKS_GTEST_H_
 #define EIGEN_CHECKS_GTEST_H_
 #include <limits>
+#include <type_traits>
+
 #include <eigen-checks/internal/gtest-equal.h>
-#include <eigen-checks/internal/traits.h>
 
 #define EIGEN_MATRIX_EQUAL(MatrixA, MatrixB)                                 \
   eigen_checks::internal::MatricesNear(MatrixA, #MatrixA, MatrixB, #MatrixB, \
-      static_cast<typename eigen_checks::internal::RemoveCR<                 \
+      static_cast<typename std::remove_reference<                            \
       decltype(MatrixA)>::type::Scalar>(0.0), "0.0")
 
 #define EIGEN_MATRIX_EQUAL_DOUBLE(MatrixA, MatrixB)                          \
   eigen_checks::internal::MatricesNear(MatrixA, #MatrixA, MatrixB, #MatrixB, \
-      static_cast<typename eigen_checks::internal::RemoveCR<                 \
+      static_cast<typename std::remove_reference<                            \
       decltype(MatrixA)>::type::Scalar>(                                     \
       eigen_checks::internal::kDefaultPrecision), "Floating point precision")