chardet/encoding/unicode.js

115 lines
3.3 KiB
JavaScript
Raw Normal View History

2013-03-03 20:08:00 +00:00
var util = require('util'),
Match = require ('../match');
/**
* This class matches UTF-16 and UTF-32, both big- and little-endian. The
* BOM will be used if it is present.
*/
module.exports.UTF_16BE = function() {
2013-03-04 19:47:01 +00:00
this.name = function() {
2013-08-06 01:35:58 +00:00
return 'UTF-16BE';
2013-03-04 19:47:01 +00:00
};
2013-03-03 20:08:00 +00:00
this.match = function(det) {
var input = det.fRawInput;
2013-08-06 01:35:58 +00:00
if (input.length >= 2 && ((input[0] & 0xFF) == 0xFE && (input[1] & 0xFF) == 0xFF)) {
2013-03-03 20:08:00 +00:00
var confidence = 100;
return new Match(det, this, confidence);
}
// TODO: Do some statistics to check for unsigned UTF-16BE
return null;
}
};
module.exports.UTF_16LE = function() {
2013-03-04 19:47:01 +00:00
this.name = function() {
2013-08-06 01:35:58 +00:00
return 'UTF-16LE';
2013-03-04 19:47:01 +00:00
};
2013-03-03 20:08:00 +00:00
this.match = function(det) {
var input = det.fRawInput;
if (input.length >= 2 && ((input[0] & 0xFF) == 0xFF && (input[1] & 0xFF) == 0xFE)) {
// An LE BOM is present.
2013-08-06 01:35:58 +00:00
if (input.length >= 4 && input[2] == 0x00 && input[3] == 0x00) {
2013-03-03 20:08:00 +00:00
// It is probably UTF-32 LE, not UTF-16
return null;
}
var confidence = 100;
return new Match(det, this, confidence);
}
// TODO: Do some statistics to check for unsigned UTF-16LE
return null;
}
};
function UTF_32() {};
UTF_32.prototype.match = function(det) {
var input = det.fRawInput,
limit = (det.fRawLength / 4) * 4,
numValid = 0,
numInvalid = 0,
hasBOM = false,
confidence = 0;
if (limit==0) {
return null;
}
if (this.getChar(input, 0) == 0x0000FEFF) {
hasBOM = true;
}
for(var i = 0; i < limit; i += 4) {
var ch = this.getChar(input, i);
if (ch < 0 || ch >= 0x10FFFF || (ch >= 0xD800 && ch <= 0xDFFF)) {
numInvalid += 1;
} else {
numValid += 1;
}
}
// Cook up some sort of confidence score, based on presence of a BOM
// and the existence of valid and/or invalid multi-byte sequences.
2013-08-06 01:35:58 +00:00
if (hasBOM && numInvalid == 0) {
2013-03-03 20:08:00 +00:00
confidence = 100;
2013-08-06 01:35:58 +00:00
} else if (hasBOM && numValid > numInvalid * 10) {
2013-03-03 20:08:00 +00:00
confidence = 80;
} else if (numValid > 3 && numInvalid == 0) {
confidence = 100;
} else if (numValid > 0 && numInvalid == 0) {
confidence = 80;
} else if (numValid > numInvalid*10) {
// Probably corrupt UTF-32BE data. Valid sequences aren't likely by chance.
confidence = 25;
}
// return confidence == 0 ? null : new CharsetMatch(det, this, confidence);
return confidence == 0 ? null : new Match(det, this, confidence);
};
module.exports.UTF_32BE = function() {
2013-03-04 19:47:01 +00:00
this.name = function() {
2013-08-06 01:35:58 +00:00
return 'UTF-32BE';
2013-03-04 19:47:01 +00:00
};
2013-03-03 20:08:00 +00:00
this.getChar = function(input, index) {
return (input[index + 0] & 0xFF) << 24 | (input[index + 1] & 0xFF) << 16 |
(input[index + 2] & 0xFF) << 8 | (input[index + 3] & 0xFF);
};
};
util.inherits(module.exports.UTF_32BE, UTF_32);
module.exports.UTF_32LE = function() {
2013-03-04 19:47:01 +00:00
this.name = function() {
2013-08-06 01:35:58 +00:00
return 'UTF-32LE';
2013-03-04 19:47:01 +00:00
};
2013-03-03 20:08:00 +00:00
this.getChar = function(input, index) {
return (input[index + 3] & 0xFF) << 24 | (input[index + 2] & 0xFF) << 16 |
(input[index + 1] & 0xFF) << 8 | (input[index + 0] & 0xFF);
};
};
util.inherits(module.exports.UTF_32LE, UTF_32);