/*
 * BRLTTY - A background process providing access to the console screen (when in
 *          text mode) for a blind person using a refreshable braille display.
 *
 * Copyright (C) 1995-2023 by The BRLTTY Developers.
 *
 * BRLTTY comes with ABSOLUTELY NO WARRANTY.
 *
 * This is free software, placed under the terms of the
 * GNU Lesser General Public License, as published by the Free Software
 * Foundation; either version 2.1 of the License, or (at your option) any
 * later version. Please see the file LICENSE-LGPL for details.
 *
 * Web Page: http://brltty.app/
 *
 * This software is maintained by Dave Mielke <dave@mielke.cc>.
 */

#include "prologue.h"

#include <string.h>

#include "log.h"
#include "unicode.h"
#include "ascii.h"

#ifdef HAVE_ICU
#include <unicode/uversion.h>
#include <unicode/uchar.h>

#ifdef HAVE_UNICODE_UNORM2_H
#include <unicode/unorm2.h>
#else /* unorm */
#include <unicode/unorm.h>
#endif /* unorm */

static int
isUcharCompatible (wchar_t character) {
  UChar uc = character;
  return uc == character;
}

static int
getName (wchar_t character, char *buffer, size_t size, UCharNameChoice choice) {
  UErrorCode error = U_ZERO_ERROR;
  u_charName(character, choice, buffer, size, &error);
  return U_SUCCESS(error) && *buffer;
}

static int
getByName (wchar_t *character, const char *name, UCharNameChoice choice) {
  UErrorCode error = U_ZERO_ERROR;
  UChar uc = u_charFromName(choice, name, &error);
  if (!U_SUCCESS(error)) return 0;

  *character = uc;
  return 1;
}

static int
nextBaseCharacter (const UChar **current, const UChar *end) {
  do {
    if (*current == end) return 0;
  } while (u_getCombiningClass(*(*current)++));

  return 1;
}
#endif /* HAVE_ICU */

#ifdef HAVE_ICONV_H
#include <iconv.h>
#endif /* HAVE_ICONV_H */

int
getCharacterName (wchar_t character, char *buffer, size_t size) {
#ifdef HAVE_ICU
  return getName(character, buffer, size, U_EXTENDED_CHAR_NAME);
#else /* HAVE_ICU */
  return 0;
#endif /* HAVE_ICU */
}

int
getCharacterByName (wchar_t *character, const char *name) {
#ifdef HAVE_ICU
  return getByName(character, name, U_EXTENDED_CHAR_NAME);
#else /* HAVE_ICU */
  return 0;
#endif /* HAVE_ICU */
}

int
getCharacterAlias (wchar_t character, char *buffer, size_t size) {
#ifdef HAVE_ICU
  return getName(character, buffer, size, U_CHAR_NAME_ALIAS);
#else /* HAVE_ICU */
  return 0;
#endif /* HAVE_ICU */
}

int
getCharacterByAlias (wchar_t *character, const char *alias) {
#ifdef HAVE_ICU
  return getByName(character, alias, U_CHAR_NAME_ALIAS);
#else /* HAVE_ICU */
  return 0;
#endif /* HAVE_ICU */
}

int
getCharacterWidth (wchar_t character) {
#if defined(HAVE_WCWIDTH)
  return wcwidth(character);
#elif defined(HAVE_ICU)
  UCharCategory category = u_getIntPropertyValue(character, UCHAR_GENERAL_CATEGORY);
  UEastAsianWidth width = u_getIntPropertyValue(character, UCHAR_EAST_ASIAN_WIDTH);

  if (character == 0) return 0;
  if (category == U_CONTROL_CHAR) return -1;

  if (category == U_NON_SPACING_MARK) return 0;
  if (category == U_ENCLOSING_MARK) return 0;

  /* Hangul Jamo medial vowels and final consonants */
  if ((character >= 0X1160) && (character <= 0X11FF) && (category == U_OTHER_LETTER)) return 0;

  /*  */
  if (character == 0XAD) return 1; /* soft hyphen */
  if (category == U_FORMAT_CHAR) return 0;

  if (width == U_EA_FULLWIDTH) return 2;
  if (width == U_EA_HALFWIDTH) return 1;

  if (width == U_EA_WIDE) return 2;
  if (width == U_EA_NARROW) return 1;

  if (width == U_EA_AMBIGUOUS) {
    /* CJK Unified Ideographs block */
    if ((character >= 0X4E00) && (character <= 0X9FFF)) return 2;

    /* CJK Unified Ideographs Externsion A block */
    if ((character >= 0X3400) && (character <= 0X4DBF)) return 2;

    /* CJK Compatibility Ideographs block */
    if ((character >= 0XF900) && (character <= 0XFAFF)) return 2;

    /* Supplementary Ideographic Plane */
  //if ((character >= 0X20000) && (character <= 0X2FFFF)) return 2;

    /* Tertiary Ideographic Plane */
  //if ((character >= 0X30000) && (character <= 0X3FFFF)) return 2;
  }

  if (category == U_UNASSIGNED) return -1;
  return 1;
#else /* character width */
  if (character == ASCII_NUL) return 0;
  if (character == ASCII_DEL) return -1;
  if (!(character & 0X60)) return -1;
  return 1;
#endif /* character width */
}

int
isBrailleCharacter (wchar_t character) {
  return (character & ~UNICODE_CELL_MASK) == UNICODE_BRAILLE_ROW;
}

int
isIdeographicCharacter (wchar_t character) {
#ifdef HAVE_ICU
  if (u_hasBinaryProperty(character, UCHAR_IDEOGRAPHIC)) return 1;
#endif /* HAVE_ICU */

  return 0;
}

int
isEmojiSequence (const wchar_t *characters, size_t count) {
#ifdef HAVE_ICU
  const wchar_t *character = characters;
  const wchar_t *end = character + count;

  while (character < end) {
    #if U_ICU_VERSION_MAJOR_NUM >= 57
    if (u_hasBinaryProperty(*character, UCHAR_EMOJI)) {
      if (u_hasBinaryProperty(*character, UCHAR_EMOJI_PRESENTATION)) {
        return 1;
      }
    }
    #endif /* U_ICU_VERSION_MAJOR_NUM >= 57 */

    character += 1;
  }
#endif /* HAVE_ICU */

  return 0;
}

wchar_t
getReplacementCharacter (void) {
#ifdef HAVE_WCHAR_H
 return UNICODE_REPLACEMENT_CHARACTER;
#else /* HAVE_WCHAR_H */
 return ASCII_SUB;
#endif /* HAVE_WCHAR_H */
}

int
composeCharacters (
  size_t *length, const wchar_t *characters,
  wchar_t *buffer, unsigned int *map
) {
#ifdef HAVE_ICU
  if (*length < 2) return 0;

  UChar source[*length];
  UChar target[*length];
  int32_t count;

  {
    const wchar_t *src = characters;
    const wchar_t *end = src + *length;
    UChar *trg = source;

    while (src < end) {
      *trg++ = *src++;
    }
  }

  {
    UErrorCode error = U_ZERO_ERROR;

#ifdef HAVE_UNICODE_UNORM2_H
    static const UNormalizer2 *normalizer = NULL;

    if (!normalizer) {
      normalizer = unorm2_getNFCInstance(&error);
      if (!U_SUCCESS(error)) return 0;
    }

    count = unorm2_normalize(normalizer,
                             source, ARRAY_COUNT(source),
                             target, ARRAY_COUNT(target),
                             &error);
#else /* unorm */
    count = unorm_normalize(source, ARRAY_COUNT(source),
                            UNORM_NFC, 0,
                            target, ARRAY_COUNT(target),
                            &error);
#endif /* unorm */

    if (!U_SUCCESS(error)) return 0;
  }

  if (count == *length) {
    if (memcmp(source, target, (*length * sizeof(source[0]))) == 0) {
      return 0;
    }
  }

  {
    const UChar *src = source;
    const UChar *srcEnd = src + ARRAY_COUNT(source);
    const UChar *trg = target;
    const UChar *trgEnd = target + count;
    wchar_t *out = buffer;

    while (trg < trgEnd) {
      if (!nextBaseCharacter(&src, srcEnd)) return 0;
      if (map) *map++ = src - source - 1;
      *out++ = *trg++;
    }

    if (nextBaseCharacter(&src, srcEnd)) return 0;
    if (map) *map = src - source;
  }

  *length = count;
  return 1;
#else /* HAVE_ICU */
  return 0;
#endif /* HAVE_ICU */
}

size_t
decomposeCharacter (
  wchar_t character, wchar_t *buffer, size_t length
) {
#ifdef HAVE_ICU
  if (isUcharCompatible(character)) {
    UChar source[1] = {character};
    UChar target[length];
    int32_t count;

    {
      UErrorCode error = U_ZERO_ERROR;

#ifdef HAVE_UNICODE_UNORM2_H
      static const UNormalizer2 *normalizer = NULL;

      if (!normalizer) {
        normalizer = unorm2_getNFDInstance(&error);
        if (!U_SUCCESS(error)) return 0;
      }

      count = unorm2_normalize(normalizer,
                               source, ARRAY_COUNT(source),
                               target, ARRAY_COUNT(target),
                               &error);
#else /* unorm */
      count = unorm_normalize(source, ARRAY_COUNT(source),
                              UNORM_NFD, 0,
                              target, ARRAY_COUNT(target),
                              &error);
#endif /* unorm */

      if (!U_SUCCESS(error)) return 0;
    }

    {
      const UChar *trg = target;
      const UChar *end = target + count;
      wchar_t *out = buffer;

      while (trg < end) {
        *out++ = *trg++;
      }
    }

    return count;
  }
#endif /* HAVE_ICU */

  return 0;
}

wchar_t
getBaseCharacter (wchar_t character) {
  wchar_t decomposed[0X10];
  size_t count = decomposeCharacter(character, decomposed, sizeof(decomposed));
  if (count) return decomposed[0];
  return 0;
}

wchar_t
getTransliteratedCharacter (wchar_t character) {
#ifdef HAVE_ICONV_H
  static iconv_t handle = NULL;
  if (!handle) handle = iconv_open("ASCII//TRANSLIT", "WCHAR_T");

  if (handle != (iconv_t)-1) {
    char *inputAddress = (char *)&character;
    size_t inputSize = sizeof(character);
    size_t outputSize = 0X10;
    char outputBuffer[outputSize];
    char *outputAddress = outputBuffer;

    if (iconv(handle, &inputAddress, &inputSize, &outputAddress, &outputSize) != (size_t)-1) {
      if ((outputAddress - outputBuffer) == 1) {
        wchar_t result = outputBuffer[0] & 0XFF;

        if (result != character) {
          if (result == WC_C('?')) {
            return 0;
          }
        }

        return result;
      }
    }
  }
#endif /* HAVE_ICONV_H */

  return 0;
}

int
handleBestCharacter (wchar_t character, CharacterHandler handleCharacter, void *data) {
  if (isBrailleCharacter(character)) return 0;

  typedef wchar_t CharacterTranslator (wchar_t character);
  static CharacterTranslator *const characterTranslators[] = {
    getBaseCharacter,
    getTransliteratedCharacter,
    NULL
  };

  CharacterTranslator *const *translateCharacter = characterTranslators;
  while (!handleCharacter(character, data)) {
    if (!*translateCharacter) return 0;

    {
      wchar_t alternate = (*translateCharacter++)(character);
      if (alternate) character = alternate;
    }
  }

  return 1;
}