js编码函数encodeURIComponent源码解析

function URIAddEncodedOctetToBuffer(octet, result, index) {  result[index++] = 37; // Char code of '%'.  result[index++] = hexCharCodeArray[octet >> 4];  result[index++] = hexCharCodeArray[octet & 0x0F];  return index;}function URIEncodeOctets(octets, result, index) {  if (hexCharCodeArray === 0) {    // 0-F    hexCharCodeArray = [48, 49, 50, 51, 52, 53, 54, 55, 56, 57,                        65, 66, 67, 68, 69, 70];  }  index = URIAddEncodedOctetToBuffer(octets[0], result, index);  if (octets[1]) index = URIAddEncodedOctetToBuffer(octets[1], result, index);  if (octets[2]) index = URIAddEncodedOctetToBuffer(octets[2], result, index);  if (octets[3]) index = URIAddEncodedOctetToBuffer(octets[3], result, index);  return index;}// 对unicode中除了代理对外的字符编码function URIEncodeSingle(cc, result, index) {  var x = (cc >> 12) & 0xF;  var y = (cc >> 6) & 63;  var z = cc & 63;  // unicode最多值是三个字节  var octets = new InternalArray(3);  // ascii码  if (cc <= 0x007F) {    octets[0] = cc;  } else if (cc <= 0x07FF) {    octets[0] = y + 192;    octets[1] = z + 128;  } else {    octets[0] = x + 224;    octets[1] = y + 128;    octets[2] = z + 128;  }  return URIEncodeOctets(octets, result, index);}function URIEncodePair(cc1 , cc2, result, index) {  var u = ((cc1 >> 6) & 0xF) + 1;  var w = (cc1 >> 2) & 0xF;  var x = cc1 & 3;  var y = (cc2 >> 6) & 0xF;  var z = cc2 & 63;  var octets = new InternalArray(4);  octets[0] = (u >> 2) + 240;  octets[1] = (((u & 3) << 4) | w) + 128;  octets[2] = ((x << 4) | y) + 128;  octets[3] = z + 128;  return URIEncodeOctets(octets, result, index);}// ECMA-262, section 15.1.3function Encode(uri, unescape) {  uri = TO_STRING(uri);  var uriLength = uri.length;  var array = new InternalArray(uriLength);  var index = 0;  for (var k = 0; k < uriLength; k++) {    var cc1 = %_StringCharCodeAt(uri, k);    if (unescape(cc1)) {      array[index++] = cc1;    } else {      if (cc1 >= 0xDC00 && cc1 <= 0xDFFF) throw MakeURIError();      // 非高代理项      if (cc1 < 0xD800 || cc1 > 0xDBFF) {        index = URIEncodeSingle(cc1, array, index);      } else {        // 高代理项        k++;        if (k == uriLength) throw MakeURIError();        var cc2 = %_StringCharCodeAt(uri, k);        // 不是合法的低代理项        if (cc2 < 0xDC00 || cc2 > 0xDFFF) throw MakeURIError();        index = URIEncodePair(cc1, cc2, array, index);      }    }  }  var result = %NewString(array.length, NEW_ONE_BYTE_STRING);  for (var i = 0; i < array.length; i++) {    %_OneByteSeqStringSetChar(i, array[i], result);  }  return result;}// ECMA-262 - 15.1.3.4function URIEncodeComponent(component) {  var unescapePredicate = function(cc) {    if (isAlphaNumeric(cc)) return true;    // !    if (cc == 33) return true;    // '()*    if (39 <= cc && cc <= 42) return true;    // -.    if (45 <= cc && cc <= 46) return true;    // _    if (cc == 95) return true;    // ~    if (cc == 126) return true;    return false;  };  return Encode(component, unescapePredicate);}

URIEncodeComponent实现过程为，直接调用Encode函数进行编码，首先利用unescapePredicate 函数跳过一些特殊的字符，这些字符不需要进行编码，具体看代码。基本工程为，逐个字符遍历，如果当前字符为低代理项，则报错（cc1 >= 0xDC00 && cc1 <= 0xDFFF）；因为低代理项必须跟在高代理项后面。接着判断如果当前字节不是高代理项(cc1 < 0xD800 || cc1 > 0xDBFF)，则直接调用URIEncodeSingle函数进行编码，如果当前字符编码为高代理项，则取下一位，如果下一位超过了字符串长度，则报错，如果下一位不是低代理项则报错，因为高代理项后面必须跟着低代理项，如果下一位是低代理项，则调用URIEncodePair函数对代理对进行编码。Encode函数主要是判断，字符的是不是不需要编码的字符，是不是代理对，是不是一般的字符三种情形。
URIEncodeSingle函数基本流程是对一个字符进行某种运算，最后调用URIEncodeOctets函数按字节进行编码。unicode码最大为三个字节，所以URIEncodeSingle函数里定义了数组长度为3。URIEncodePair函数基本流程为，对高低代理项进行某种运算，高低代理项一共是四个字节，然后调用URIEncodeOctets函数对四个字节进行编码。