Move unicode unit test to CATCH and enable it

Author: tautschnig

unit/CMakeLists.txt

@@ -13,7 +13,6 @@ list(REMOVE_ITEM sources
     ${CMAKE_CURRENT_SOURCE_DIR}/json.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/cpp_parser.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/osx_fat_reader.cpp
-    ${CMAKE_CURRENT_SOURCE_DIR}/unicode.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/wp.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/cpp_scanner.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/float_utils.cpp

unit/Makefile

@@ -37,6 +37,7 @@ SRC += unit_tests.cpp \
        util/string_utils/split_string.cpp \
        util/string_utils/strip_string.cpp \
        util/symbol_table.cpp \
+       util/unicode.cpp \
        catch_example.cpp \
        # Empty last line
 

unit/unicode.cpp renamed to unit/util/unicode.cpp

@@ -6,20 +6,19 @@ Author: Vojtech Forejt, forejtv@diffblue.com
 
 \*******************************************************************/
 
-#include <cassert>
+#include <testing-utils/catch.hpp>
+
 #include <vector>
 #include <string>
 #include <codecvt>
-#include <iomanip>
-#include <iostream>
 #include <locale>
 
 #include <util/unicode.h>
 
 // This unit test compares our implementation with codecvt implementation,
 // checking bit-by-bit equivalence of results.
 
-bool paranoid_wstr_equals(const std::wstring &a, const std::wstring &b)
+static bool paranoid_wstr_equals(const std::wstring &a, const std::wstring &b)
 {
   if(a.size() != b.size())
     return false;
@@ -35,7 +34,10 @@ bool paranoid_wstr_equals(const std::wstring &a, const std::wstring &b)
 }
 
 // helper print function, can be called for debugging problem
-void wstr_print(const std::wstring &a, const std::wstring &b)
+#if 0
+#include <iostream>
+
+static void wstr_print(const std::wstring &a, const std::wstring &b)
 {
   int endi=(a.size()>b.size())?a.size():b.size();
   const unsigned char
@@ -49,19 +51,23 @@ void wstr_print(const std::wstring &a, const std::wstring &b)
   }
   std::cout << '\n';
 }
+#endif
 
-void compare_utf8_to_utf16_big_endian(std::string& in)
+#if 0
+// big-endian test is broken, will be fixed in subsequent commit
+static bool compare_utf8_to_utf16_big_endian(const std::string &in)
 {
   std::wstring s1=utf8_to_utf16_big_endian(in);
 
   typedef std::codecvt_utf8_utf16<wchar_t> codecvt_utf8_utf16t;
   std::wstring_convert<codecvt_utf8_utf16t> converter;
   std::wstring s2=converter.from_bytes(in);
 
-  assert(paranoid_wstr_equals(s1, s2));
+  return paranoid_wstr_equals(s1, s2);
 }
+#endif
 
-void compare_utf8_to_utf16_little_endian(std::string& in)
+static bool compare_utf8_to_utf16_little_endian(const std::string &in)
 {
   std::wstring s1=utf8_to_utf16_little_endian(in);
 
@@ -72,23 +78,40 @@ void compare_utf8_to_utf16_little_endian(std::string& in)
   std::wstring_convert<codecvt_utf8_utf16t> converter;
   std::wstring s2=converter.from_bytes(in);
 
-  assert(paranoid_wstr_equals(s1, s2));
+  return paranoid_wstr_equals(s1, s2);
 }
 
-int main()
+TEST_CASE("unicode0", "[core][util][unicode]")
 {
-  std::string s;
-  s=u8"\u0070\u00DF\u00E0\u00EF\u00F0\u00F7\u00F8";
-  compare_utf8_to_utf16_big_endian(s);
-  compare_utf8_to_utf16_little_endian(s);
-  s=u8"$¢€𐍈";
-  compare_utf8_to_utf16_big_endian(s);
-  compare_utf8_to_utf16_little_endian(s);
-  s=u8"𐐏𤭢";
-  compare_utf8_to_utf16_big_endian(s);
-  compare_utf8_to_utf16_little_endian(s);
-  s=u8"дȚȨɌṡʒʸͼἨѶݔݺ→⅒⅀▤▞╢◍⛳⻥龍ンㄗㄸ";
-  compare_utf8_to_utf16_big_endian(s);
-  compare_utf8_to_utf16_little_endian(s);
+  const std::string s = u8"abc";
+  // REQUIRE(compare_utf8_to_utf16_big_endian(s));
+  REQUIRE(compare_utf8_to_utf16_little_endian(s));
 }
 
+TEST_CASE("unicode1", "[core][util][unicode]")
+{
+  const std::string s = u8"\u0070\u00DF\u00E0\u00EF\u00F0\u00F7\u00F8";
+  // REQUIRE(compare_utf8_to_utf16_big_endian(s));
+  REQUIRE(compare_utf8_to_utf16_little_endian(s));
+}
+
+TEST_CASE("unicode2", "[core][util][unicode]")
+{
+  const std::string s = u8"$¢€𐍈";
+  // REQUIRE(compare_utf8_to_utf16_big_endian(s));
+  REQUIRE(compare_utf8_to_utf16_little_endian(s));
+}
+
+TEST_CASE("unicode3", "[core][util][unicode]")
+{
+  const std::string s = u8"𐐏𤭢";
+  // REQUIRE(compare_utf8_to_utf16_big_endian(s));
+  REQUIRE(compare_utf8_to_utf16_little_endian(s));
+}
+
+TEST_CASE("unicode4", "[core][util][unicode]")
+{
+  const std::string s = u8"дȚȨɌṡʒʸͼἨѶݔݺ→⅒⅀▤▞╢◍⛳⻥龍ンㄗㄸ";
+  // REQUIRE(compare_utf8_to_utf16_big_endian(s));
+  REQUIRE(compare_utf8_to_utf16_little_endian(s));
+}