Javascript Base64編碼UTF8字符串在webkit/safari中失敗

Question

我想base64編碼包含泰文字符的utf8字符串。我正在使用內置的btoa功能的瀏覽器。它適用於ascii文本，但泰國正在導致它拋出一個異常INVALID_CHARACTER_ERR: DOM Exception 5。

下面是一個失敗（看起來像一個「n」是泰國的字符）的樣本

btoa('aก')

---

什麼我需要做的Base64編碼，非ASCII字符串？

Answer 1

不幸的是，btoa/atob沒有在任何標準中指定，但firefox和webkit中的實現都在多字節字符上失敗，所以即使現在指定這些內置函數將不能支持多字節字符（作爲輸入並且輸出字符串必然會改變）。

這似乎是你唯一的選擇是將推出自己的base64編碼+解碼程序

Answer 2

檢查此變通 http://ecmanaut.blogspot.com/2006/07/encoding-decoding-utf8-in-javascript.html

Answer 3

var Base64 = { 
    encode: function(s) { 
    return btoa(unescape(encodeURIComponent(s))); 
    }, 
    decode: function(s) { 
    return decodeURIComponent(escape(atob(s))); 
    } 
}; 

Answer 4

我知道這是老了，但我最近找UTF8-to-Base64編碼器也是如此。我在http://www.webtoolkit.info/javascript-base64.html找到了一個方便的小腳本，並在http://jsbase64.codeplex.com/處獲得了性能改進版本。

下面是腳本：

var B64 = { 
    alphabet: 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz/=', 
    lookup: null, 
    ie: /MSIE /.test(navigator.userAgent), 
    ieo: /MSIE [67]/.test(navigator.userAgent), 
    encode: function (s) { 
     var buffer = B64.toUtf8(s), 
      position = -1, 
      len = buffer.length, 
      nan0, nan1, nan2, enc = [, , , ]; 
     if (B64.ie) { 
      var result = []; 
      while (++position < len) { 
       nan0 = buffer[position]; 
       nan1 = buffer[++position]; 
       enc[0] = nan0 >> 2; 
       enc[1] = ((nan0 & 3) << 4) | (nan1 >> 4); 
       if (isNaN(nan1)) 
        enc[2] = enc[3] = 64; 
       else { 
        nan2 = buffer[++position]; 
        enc[2] = ((nan1 & 15) << 2) | (nan2 >> 6); 
        enc[3] = (isNaN(nan2)) ? 64 : nan2 & 63; 
       } 
       result.push(B64.alphabet.charAt(enc[0]), B64.alphabet.charAt(enc[1]), B64.alphabet.charAt(enc[2]), B64.alphabet.charAt(enc[3])); 
      } 
      return result.join(''); 
     } else { 
      var result = ''; 
      while (++position < len) { 
       nan0 = buffer[position]; 
       nan1 = buffer[++position]; 
       enc[0] = nan0 >> 2; 
       enc[1] = ((nan0 & 3) << 4) | (nan1 >> 4); 
       if (isNaN(nan1)) 
        enc[2] = enc[3] = 64; 
       else { 
        nan2 = buffer[++position]; 
        enc[2] = ((nan1 & 15) << 2) | (nan2 >> 6); 
        enc[3] = (isNaN(nan2)) ? 64 : nan2 & 63; 
       } 
       result += B64.alphabet[enc[0]] + B64.alphabet[enc[1]] + B64.alphabet[enc[2]] + B64.alphabet[enc[3]]; 
      } 
      return result; 
     } 
    }, 
    decode: function (s) { 
     if (s.length % 4) 
      throw new Error("InvalidCharacterError: 'B64.decode' failed: The string to be decoded is not correctly encoded."); 
     var buffer = B64.fromUtf8(s), 
      position = 0, 
      len = buffer.length; 
     if (B64.ieo) { 
      var result = []; 
      while (position < len) { 
       if (buffer[position] < 128) 
        result.push(String.fromCharCode(buffer[position++])); 
       else if (buffer[position] > 191 && buffer[position] < 224) 
        result.push(String.fromCharCode(((buffer[position++] & 31) << 6) | (buffer[position++] & 63))); 
       else 
        result.push(String.fromCharCode(((buffer[position++] & 15) << 12) | ((buffer[position++] & 63) << 6) | (buffer[position++] & 63))); 
      } 
      return result.join(''); 
     } else { 
      var result = ''; 
      while (position < len) { 
       if (buffer[position] < 128) 
        result += String.fromCharCode(buffer[position++]); 
       else if (buffer[position] > 191 && buffer[position] < 224) 
        result += String.fromCharCode(((buffer[position++] & 31) << 6) | (buffer[position++] & 63)); 
       else 
        result += String.fromCharCode(((buffer[position++] & 15) << 12) | ((buffer[position++] & 63) << 6) | (buffer[position++] & 63)); 
      } 
      return result; 
     } 
    }, 
    toUtf8: function (s) { 
     var position = -1, 
      len = s.length, 
      chr, buffer = []; 
     if (/^[\x00-\x7f]*$/.test(s)) while (++position < len) 
      buffer.push(s.charCodeAt(position)); 
     else while (++position < len) { 
      chr = s.charCodeAt(position); 
      if (chr < 128) 
       buffer.push(chr); 
      else if (chr < 2048) 
       buffer.push((chr >> 6) | 192, (chr & 63) | 128); 
      else 
       buffer.push((chr >> 12) | 224, ((chr >> 6) & 63) | 128, (chr & 63) | 128); 
     } 
     return buffer; 
    }, 
    fromUtf8: function (s) { 
     var position = -1, 
      len, buffer = [], 
      enc = [, , , ]; 
     if (!B64.lookup) { 
      len = B64.alphabet.length; 
      B64.lookup = {}; 
      while (++position < len) 
       B64.lookup[B64.alphabet.charAt(position)] = position; 
      position = -1; 
     } 
     len = s.length; 
     while (++position < len) { 
      enc[0] = B64.lookup[s.charAt(position)]; 
      enc[1] = B64.lookup[s.charAt(++position)]; 
      buffer.push((enc[0] << 2) | (enc[1] >> 4)); 
      enc[2] = B64.lookup[s.charAt(++position)]; 
      if (enc[2] == 64) 
       break; 
      buffer.push(((enc[1] & 15) << 4) | (enc[2] >> 2)); 
      enc[3] = B64.lookup[s.charAt(++position)]; 
      if (enc[3] == 64) 
       break; 
      buffer.push(((enc[2] & 3) << 6) | enc[3]); 
     } 
     return buffer; 
    } 
}; 

免責聲明：我沒有與泰國字符測試了這個特別的，但假設它會奏效。

Sav的