chardet/encoding/unicode.js

113 lines
3.0 KiB
JavaScript
Raw Normal View History

'use strict';
2013-03-03 20:08:00 +00:00
var util = require('util'),
Match = require ('../match');
2013-03-03 20:08:00 +00:00
/**
* 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() {
this.name = function() {
return 'UTF-16BE';
};
this.match = function(det) {
var input = det.fRawInput;
if (input.length >= 2 && ((input[0] & 0xff) == 0xfe && (input[1] & 0xff) == 0xff)) {
2015-11-29 21:51:34 +00:00
return new Match(det, this, 100); // confidence = 100
}
// TODO: Do some statistics to check for unsigned UTF-16BE
return null;
};
2013-03-03 20:08:00 +00:00
};
module.exports.UTF_16LE = function() {
this.name = function() {
return 'UTF-16LE';
};
this.match = function(det) {
var input = det.fRawInput;
if (input.length >= 2 && ((input[0] & 0xff) == 0xff && (input[1] & 0xff) == 0xfe)) {
2015-11-29 21:51:34 +00:00
// LE BOM is present.
if (input.length >= 4 && input[2] == 0x00 && input[3] == 0x00) {
// It is probably UTF-32 LE, not UTF-16
return null;
}
2015-11-29 21:51:34 +00:00
return new Match(det, this, 100); // confidence = 100
2013-03-03 20:08:00 +00:00
}
// TODO: Do some statistics to check for unsigned UTF-16LE
return null;
}
2013-03-03 20:08:00 +00:00
};
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;
}
}
2013-11-22 04:37:30 +00:00
// Cook up some sort of confidence score, based on presence of a BOM
// and the existence of valid and/or invalid multi-byte sequences.
if (hasBOM && numInvalid == 0) {
confidence = 100;
} else if (hasBOM && numValid > numInvalid * 10) {
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;
}
2013-11-22 04:37:30 +00:00
// return confidence == 0 ? null : new CharsetMatch(det, this, confidence);
return confidence == 0 ? null : new Match(det, this, confidence);
2013-03-03 20:08:00 +00:00
};
module.exports.UTF_32BE = function() {
this.name = function() {
return 'UTF-32BE';
};
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);
};
2013-03-03 20:08:00 +00:00
};
util.inherits(module.exports.UTF_32BE, UTF_32);
module.exports.UTF_32LE = function() {
this.name = function() {
return 'UTF-32LE';
};
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);
};
2013-03-03 20:08:00 +00:00
};
util.inherits(module.exports.UTF_32LE, UTF_32);