Unit test and fixed bug

CMakeLists.txt

@@ -33,6 +33,9 @@ project(fuzzywuzzy LANGUAGES C CXX)
 
 message(STATUS "Build type: ${CMAKE_BUILD_TYPE}")
 
+set(CMAKE_CXX_STANDARD 14)
+set(CMAKE_CXX_STANDARD_REQUIRED ON)
+
 set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/bin)
 
 if (CMAKE_COMPILER_IS_GNUCXX OR (CMAKE_CXX_COMPILER_ID MATCHES "Clang"))
@@ -46,7 +49,9 @@ set(CMAKE_MODULE_PATH
 add_subdirectory(src)
 add_subdirectory(test)
 
-if (FMT_TEST)
-  enable_testing()
-  #add_subdirectory(test)
-endif()
+#if (FMT_TEST)
+enable_testing()
+add_test(main_test ${CMAKE_BINARY_DIR}/bin/main)
+#add_subdirectory(test)
+#endif()
+

README.md

@@ -11,6 +11,30 @@ and been wrapped around some C++ code.
 | `utils.{c,h}pp` | Utility functions, translated from the Python library's `utils.py`. |
 | `levenshtein.{c,h}` | The underlaying C functions, copied verbatim. |
 
+Build
+-----
+
+Build this project and the unit test
+
+```
+mkdir -p build
+cd build
+cmake ..
+make -j$(nproc)
+```
+
+Test
+-----
+
+The unit test of C++ fuzzywuzzy is based on the [Catch2](https://github.com/catchorg/Catch2) framework. Install Catch2 into system path or place the single-header version into the include path as `catch2/catch.hpp`
+
+To run the unit test:
+
+```
+cd build
+./bin/main
+```
+
 Usage
 -----
 ```cpp
@@ -39,4 +63,4 @@ fuzz::token_sort_ratio("fuzzy wuzzy was a bear", "wuzzy fuzzy was a bear"); // r
 fuzz::token_sort_ratio("fuzzy was a bear", "fuzzy fuzzy was a bear"); // returns 83 (this should be 84)
 
 fuzz::token_set_ratio("fuzzy was a bear", "fuzzy fuzzy was a bear"); // returns 100
-```
+```

include/fuzzywuzzy.hpp

@@ -9,13 +9,13 @@ namespace /* I'm in your mind... */ fuzz {
 /*                          */
 
 /* Calculates a Levenshtein simple ratio between the string. */
-unsigned int ratio(const string &s1, const string &s2, const bool full_process = true);
+unsigned int ratio(const string &s1, const string &s2, const bool full_process = false);
 
 /*
  * Return the ratio of the most similar substring
  * as a number between 0 and 100.
  */
-unsigned int partial_ratio(const string &s1, const string &s2, const bool full_process = true);
+unsigned int partial_ratio(const string &s1, const string &s2, const bool full_process = false);
 
 /*                             */
 /* Advanced scoring functions. */

src/fuzzywuzzy.cpp

@@ -13,6 +13,12 @@ unsigned int ratio(const string &s1, const string &s2, const bool full_process)
     string p1 = full_process ? utils::full_process(s1) : s1;
     string p2 = full_process ? utils::full_process(s2) : s2;
 
+    // handle empty string cases
+    if (p1.empty() && p2.empty())
+        return 100;
+    else if (p1.empty() || p2.empty())
+        return 0;
+
     auto m = string_matcher(p1, p2);
     return utils::percent_round(m.ratio());
 }
@@ -22,6 +28,38 @@ unsigned int partial_ratio(const string &s1, const string &s2, const bool full_p
     string p1 = full_process ? utils::full_process(s1) : s1;
     string p2 = full_process ? utils::full_process(s2) : s2;
 
+    // handle empty string cases
+    if (p1.empty() && p2.empty())
+        return 100;
+    else if (p1.empty() || p2.empty())
+        return 0;
+
+    /**
+     * The core part of `partial_ratio` fails to handle the matching of two
+     * identical strings, such that:
+     *
+     * ``` partial_ratio('a, b', 'a, b') == 0;
+     * ```
+     *
+     * In this case, I found the root cause is that the `m.get_matching_blocks()`
+     * return zero block.
+     *
+     * ``` auto blocks = m.get_matching_blocks();
+     *     blocks.empty() == true;
+     * ```
+     *
+     * Wheras in the orignal Python implementation, it returns a list of two
+     * elements, as following:
+     *
+     * ``` [Match(a=0, b=0, size=4), Match(a=4, b=4, size=0)]
+     * ```
+     *
+     * As a result, a temporary workaround is placed here below.
+     *
+     */
+    if (p1 == p2)
+        return 100;
+
     string shorter, longer;
 
     if (p1.length() <= p2.length()) {

test/CMakeLists.txt

@@ -3,4 +3,4 @@ include_directories(${LIB_INCLUDE_DIRS})
 
 file(GLOB_RECURSE SOURCES RELATIVE ${PROJECT_SOURCE_DIR}/test *.c[p]*)
 add_executable(main ${SOURCES})
-target_link_libraries(main fuzzywuzzy)
+target_link_libraries(main fuzzywuzzy)

test/main.cpp

@@ -1,13 +1,5 @@
 #include <iostream>
 
-#include "fuzzywuzzy.hpp"
+#define CATCH_CONFIG_MAIN
 
-int main()
-{
-    const string a = "I'm in your mind", b = "I'm in your mind fuzz";
-    const string c = "fuzzy wuzzy was a bear", d = "wuzzy fuzzy was a bear";
-
-    std::cout << fuzz::ratio(a, b) << '\n';
-    std::cout << fuzz::partial_ratio(a, b) << '\n';
-    std::cout << fuzz::token_sort_ratio(c, d) << '\n';
-}
+#include "catch2/catch.hpp"