跨平台utf8与unicode的互转

这几天帮同事看一个聊天敏感词过滤问题，需求是在ndk环境下，把用户输入的句子中不健康的词语替换成*，开始时需要先将用户输入转成wchar_t*，用了mbstowcs，在本地linux环境下测试没有问题，但是比较坑的是ndk安卓真机环境下不行，后来干脆自己写了utf8与unicode相互转换的一组函数，utf8sToUnicodes是将utf8字符串转成unicode数组，unicodesToUtf8s是将unicode数组转成utf8字符串。

\\根据多字节字符串的第一个字符返回该unicode占的字节数
int getUtf8Len(const char* utfs) {
  unsigned int in = utfs[0];
  in = in & 0x000000ff;
  if (in >= 0x00 && in < 0xC0) return 1;
  else if (in >= 0xC0 && in < 0xE0) return 2;
  else if (in >= 0xE0 && in < 0xF0) return 3;
  else return 0;
}

\\将多个字符组成的一个字转换成unicode
int utf8ToUnicode(const char* utf, int len) {
  unsigned int rst;
  if (1 == len) rst = utf[0]; 
  else if (2 == len) {
    unsigned int part1 = utf[0] & 0x1F;
    part1 << 6;
    unsigned int part2 = utf[1] & 0x3F;
    rst = part1 | part2;
  } else if (3 == len) {
    unsigned int part1 = utf[0] & 0x0F;
    part1 = part1 << 12;
    unsigned int part2 = utf[1] & 0x3F;
    part2 = part2 << 6;
    unsigned int part3 = utf[2] & 0x3F;
    rst = part1 | part2 | part3;
  } else {
    rst = 0;
  }
  return rst;
}

\\将Unicode转换成多字节字符
void unicodeToUtf(const unsigned int unicode, char* utf) {
  if (unicode < 0x80) {
    utf[0] = unicode >> 0 & 0x7F | 0x00;
    utf[1] = '\0';
  } else if (unicode < 0x0800) {
    utf[0] = unicode >> 6 & 0x1F | 0xC0;
    utf[1] = unicode >> 0 & 0x3F | 0x80;
    utf[2] = '\0';
  } else if (unicode < 0x010000) {
    utf[0] = unicode >> 12 & 0x0F | 0xE0;
    utf[1] = unicode >> 6 & 0x3F | 0x80;
    utf[2] = unicode >> 0 & 0x3F | 0x80;
    utf[3] = '\0';
  }
}

\\将多字节字符串转换成unicode数组
int utf8sToUnicodes(const char* utfs, unsigned int * unicodes) {
  int len = strlen(utfs);
  int mindex = 0;
  int windex = 0;
  while (mindex < len) {
    int bytes = getUtf8Len(utfs + mindex);
    unicodes[windex] = utf8ToUnicode(utfs + mindex, bytes);
    mindex += bytes;
    windex ++;
  }
  unicodes[windex] = 0;
  return windex;
}

\\将unicode数组转换成多字节字符串
void unicodesToUtf8s(const unsigned int* unicodes, char* utfs) {
  sprintf(utfs, "%s", "");
  int index = 0;
  while(unicodes[index] != 0) {
    char word[4];
    unicodeToUtf(unicodes[index], word);
    sprintf(utfs, "%s%s", utfs, word);
    index ++;
  }
}