#pragma region CPL License
/*
Nuclex Native Framework
Copyright (C) 2002-2023 Nuclex Development Labs

This library is free software; you can redistribute it and/or
modify it under the terms of the IBM Common Public License as
published by the IBM Corporation; either version 1.0 of the
License, or (at your option) any later version.

This library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
IBM Common Public License for more details.

You should have received a copy of the IBM Common Public
License along with this library
*/
#pragma endregion // CPL License

// If the library is compiled as a DLL, this ensures symbols are exported
#define NUCLEX_SUPPORT_SOURCE 1

#include "Nuclex/Support/Text/StringConverter.h"

#include <gtest/gtest.h>

namespace Nuclex { namespace Support { namespace Text {

  // ------------------------------------------------------------------------------------------- //

  TEST(StringConverterTest, CanConvertUtf8ToWide) {
    std::string text = u8"ăѣ𝔠ծềſģȟᎥ𝒋ǩľḿꞑȯ𝘱𝑞𝗋𝘴ȶ𝞄𝜈ψ𝒙𝘆𝚣";
    EXPECT_EQ(StringConverter::WideFromUtf8(text), L"ăѣ𝔠ծềſģȟᎥ𝒋ǩľḿꞑȯ𝘱𝑞𝗋𝘴ȶ𝞄𝜈ψ𝒙𝘆𝚣");

    text = u8"𝘈Ḇ𝖢𝕯٤ḞԍНǏ𝙅ƘԸⲘ𝙉০Ρ𝗤Ɍ𝓢ȚЦ𝒱Ѡ𝓧ƳȤѧᖯć𝗱ễ𝑓𝙜Ⴙ𝞲𝑗𝒌ļṃʼnо𝞎𝒒ᵲꜱ𝙩ừ𝗏ŵ𝒙𝒚ź";
    EXPECT_EQ(
      StringConverter::WideFromUtf8(text),
      L"𝘈Ḇ𝖢𝕯٤ḞԍНǏ𝙅ƘԸⲘ𝙉০Ρ𝗤Ɍ𝓢ȚЦ𝒱Ѡ𝓧ƳȤѧᖯć𝗱ễ𝑓𝙜Ⴙ𝞲𝑗𝒌ļṃʼnо𝞎𝒒ᵲꜱ𝙩ừ𝗏ŵ𝒙𝒚ź"
    );

    text = u8"АḂⲤ𝗗𝖤𝗙ꞠꓧȊ𝐉𝜥ꓡ𝑀𝑵Ǭ𝙿𝑄Ŗ𝑆𝒯𝖴𝘝𝘞ꓫŸ𝜡ả𝘢ƀ𝖼ḋếᵮℊ𝙝Ꭵ𝕛кιṃդⱺ𝓅𝘲𝕣𝖘ŧ𝑢ṽẉ𝘅ყž";
    EXPECT_EQ(
      StringConverter::WideFromUtf8(text),
      L"АḂⲤ𝗗𝖤𝗙ꞠꓧȊ𝐉𝜥ꓡ𝑀𝑵Ǭ𝙿𝑄Ŗ𝑆𝒯𝖴𝘝𝘞ꓫŸ𝜡ả𝘢ƀ𝖼ḋếᵮℊ𝙝Ꭵ𝕛кιṃդⱺ𝓅𝘲𝕣𝖘ŧ𝑢ṽẉ𝘅ყž"
    );

    text = u8"Ѧ𝙱ƇᗞΣℱԍҤ١𝔍К𝓛𝓜ƝȎ𝚸𝑄Ṛ𝓢ṮṺƲᏔꓫ𝚈𝚭𝜶Ꮟçძ𝑒𝖿𝗀ḧ𝗂𝐣ҝɭḿ𝕟𝐨𝝔𝕢ṛ𝓼тú𝔳ẃ⤬𝝲𝗓";
    EXPECT_EQ(
      StringConverter::WideFromUtf8(text),
      L"Ѧ𝙱ƇᗞΣℱԍҤ١𝔍К𝓛𝓜ƝȎ𝚸𝑄Ṛ𝓢ṮṺƲᏔꓫ𝚈𝚭𝜶Ꮟçძ𝑒𝖿𝗀ḧ𝗂𝐣ҝɭḿ𝕟𝐨𝝔𝕢ṛ𝓼тú𝔳ẃ⤬𝝲𝗓"
    );

    text = u8"𝖠Β𝒞𝘋𝙴𝓕ĢȞỈ𝕵ꓗʟ𝙼ℕ০𝚸𝗤ՀꓢṰǓⅤ𝔚Ⲭ𝑌𝙕𝘢𝕤";
    EXPECT_EQ(
      StringConverter::WideFromUtf8(text),
      L"𝖠Β𝒞𝘋𝙴𝓕ĢȞỈ𝕵ꓗʟ𝙼ℕ০𝚸𝗤ՀꓢṰǓⅤ𝔚Ⲭ𝑌𝙕𝘢𝕤"
    );
  }

  // ------------------------------------------------------------------------------------------- //

  TEST(StringConverterTest, CanConvertWideToUtf8) {
    std::wstring text = L"ăѣ𝔠ծềſģȟᎥ𝒋ǩľḿꞑȯ𝘱𝑞𝗋𝘴ȶ𝞄𝜈ψ𝒙𝘆𝚣";
    EXPECT_EQ(StringConverter::Utf8FromWide(text), u8"ăѣ𝔠ծềſģȟᎥ𝒋ǩľḿꞑȯ𝘱𝑞𝗋𝘴ȶ𝞄𝜈ψ𝒙𝘆𝚣");

    text = L"𝘈Ḇ𝖢𝕯٤ḞԍНǏ𝙅ƘԸⲘ𝙉০Ρ𝗤Ɍ𝓢ȚЦ𝒱Ѡ𝓧ƳȤѧᖯć𝗱ễ𝑓𝙜Ⴙ𝞲𝑗𝒌ļṃʼnо𝞎𝒒ᵲꜱ𝙩ừ𝗏ŵ𝒙𝒚ź";
    EXPECT_EQ(
      StringConverter::Utf8FromWide(text),
      u8"𝘈Ḇ𝖢𝕯٤ḞԍНǏ𝙅ƘԸⲘ𝙉০Ρ𝗤Ɍ𝓢ȚЦ𝒱Ѡ𝓧ƳȤѧᖯć𝗱ễ𝑓𝙜Ⴙ𝞲𝑗𝒌ļṃʼnо𝞎𝒒ᵲꜱ𝙩ừ𝗏ŵ𝒙𝒚ź"
    );

    text = L"АḂⲤ𝗗𝖤𝗙ꞠꓧȊ𝐉𝜥ꓡ𝑀𝑵Ǭ𝙿𝑄Ŗ𝑆𝒯𝖴𝘝𝘞ꓫŸ𝜡ả𝘢ƀ𝖼ḋếᵮℊ𝙝Ꭵ𝕛кιṃդⱺ𝓅𝘲𝕣𝖘ŧ𝑢ṽẉ𝘅ყž";
    EXPECT_EQ(
      StringConverter::Utf8FromWide(text),
      u8"АḂⲤ𝗗𝖤𝗙ꞠꓧȊ𝐉𝜥ꓡ𝑀𝑵Ǭ𝙿𝑄Ŗ𝑆𝒯𝖴𝘝𝘞ꓫŸ𝜡ả𝘢ƀ𝖼ḋếᵮℊ𝙝Ꭵ𝕛кιṃդⱺ𝓅𝘲𝕣𝖘ŧ𝑢ṽẉ𝘅ყž"
    );

    text = L"Ѧ𝙱ƇᗞΣℱԍҤ١𝔍К𝓛𝓜ƝȎ𝚸𝑄Ṛ𝓢ṮṺƲᏔꓫ𝚈𝚭𝜶Ꮟçძ𝑒𝖿𝗀ḧ𝗂𝐣ҝɭḿ𝕟𝐨𝝔𝕢ṛ𝓼тú𝔳ẃ⤬𝝲𝗓";
    EXPECT_EQ(
      StringConverter::Utf8FromWide(text),
      u8"Ѧ𝙱ƇᗞΣℱԍҤ١𝔍К𝓛𝓜ƝȎ𝚸𝑄Ṛ𝓢ṮṺƲᏔꓫ𝚈𝚭𝜶Ꮟçძ𝑒𝖿𝗀ḧ𝗂𝐣ҝɭḿ𝕟𝐨𝝔𝕢ṛ𝓼тú𝔳ẃ⤬𝝲𝗓"
    );

    text = L"𝖠Β𝒞𝘋𝙴𝓕ĢȞỈ𝕵ꓗʟ𝙼ℕ০𝚸𝗤ՀꓢṰǓⅤ𝔚Ⲭ𝑌𝙕𝘢𝕤";
    EXPECT_EQ(
      StringConverter::Utf8FromWide(text),
      u8"𝖠Β𝒞𝘋𝙴𝓕ĢȞỈ𝕵ꓗʟ𝙼ℕ০𝚸𝗤ՀꓢṰǓⅤ𝔚Ⲭ𝑌𝙕𝘢𝕤"
    );
  }

  // ------------------------------------------------------------------------------------------- //

  TEST(StringConverterTest, Utf8StringsCanBeCaseFolded) {
    std::string variant1 = u8"HeLlO wOrLd Ä ö Ü λ Φ δ ẞ";
    std::string variant2 = u8"hElLo WoRlD ä Ö ü Λ φ Δ ß";
    std::string wrong1 = u8"hElLo WoRlD A o U λ Φ Δ B";
    std::string wrong2 = u8"hElLo WoRlD ä ö ü ^ & ∩ b";

    EXPECT_EQ(
      StringConverter::FoldedLowercaseFromUtf8(variant1),
      StringConverter::FoldedLowercaseFromUtf8(variant2)
    );
    EXPECT_NE(
      StringConverter::FoldedLowercaseFromUtf8(variant1),
      StringConverter::FoldedLowercaseFromUtf8(wrong1)
    );
    EXPECT_NE(
      StringConverter::FoldedLowercaseFromUtf8(variant2),
      StringConverter::FoldedLowercaseFromUtf8(wrong2)
    );
  }

  // ------------------------------------------------------------------------------------------- //

  TEST(StringConverterTest, Utf8CharactersCanBeCounted) {

    // Symbols that require 4 bytes to represent in UTF-8. If your IDE doesn't display this,
    // that's fine. Just, maybe, don't save this file!
    // (it's a fly, a snake, an alligator and a fish symbol from egyptian hieroglyphs)
    std::string fourByteSymbols = u8"𓆦𓆓𓆌𓆟";
    EXPECT_EQ(StringConverter::CountUtf8Letters(fourByteSymbols), 4U);

    // Also try with some less exotic UTF-8 characters using 2 or 3 bytes.
    std::string otherSymbols = u8"𝖠Β𝒞𝘋𝙴𝓕ĢȞỈ𝕵ꓗʟ𝙼ℕ০𝚸𝗤ՀꓢṰǓⅤ𝔚Ⲭ𝑌𝙕𝘢𝕤";
    EXPECT_EQ(StringConverter::CountUtf8Letters(otherSymbols), 28U);

  }

  // ------------------------------------------------------------------------------------------- //

}}} // namespace Nuclex::Support::Text