LumeWeb
/
chardet
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

major style changes

This commit is contained in:
Dmitry Shirokov 2013-11-22 15:37:30 +11:00
parent c429ac9026
commit e5ba182d7f
5 changed files with 1354 additions and 1354 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

194
encoding/iso2022.js
View File

@ -1,5 +1,5 @@
var util = require('util'), var util = require('util'),
Match = require ('../match'); Match = require ('../match');
/** /**
@ -12,130 +12,130 @@ function ISO_2022() {}
ISO_2022.prototype.match = function(det) { ISO_2022.prototype.match = function(det) {
/** /**
* Matching function shared among the 2022 detectors JP, CN and KR * Matching function shared among the 2022 detectors JP, CN and KR
* Counts up the number of legal an unrecognized escape sequences in * Counts up the number of legal an unrecognized escape sequences in
* the sample of text, and computes a score based on the total number & * the sample of text, and computes a score based on the total number &
* the proportion that fit the encoding. * the proportion that fit the encoding.
* *
* *
* @param text the byte buffer containing text to analyse * @param text the byte buffer containing text to analyse
* @param textLen the size of the text in the byte. * @param textLen the size of the text in the byte.
* @param escapeSequences the byte escape sequences to test for. * @param escapeSequences the byte escape sequences to test for.
* @return match quality, in the range of 0-100. * @return match quality, in the range of 0-100.
*/ */
var i, j; var i, j;
var escN; var escN;
var hits = 0; var hits = 0;
var misses = 0; var misses = 0;
var shifts = 0; var shifts = 0;
var quality; var quality;
// TODO: refactor me // TODO: refactor me
var text = det.fInputBytes; var text = det.fInputBytes;
var textLen = det.fInputLen; var textLen = det.fInputLen;
scanInput: scanInput:
for (i = 0; i < textLen; i++) { for (i = 0; i < textLen; i++) {
if (text[i] == 0x1b) { if (text[i] == 0x1b) {
checkEscapes: checkEscapes:
for (escN = 0; escN < this.escapeSequences.length; escN++) { for (escN = 0; escN < this.escapeSequences.length; escN++) {
var seq = this.escapeSequences[escN]; var seq = this.escapeSequences[escN];
if ((textLen - i) < seq.length) if ((textLen - i) < seq.length)
continue checkEscapes; continue checkEscapes;
for (j = 1; j < seq.length; j++) for (j = 1; j < seq.length; j++)
if (seq[j] != text[i + j]) if (seq[j] != text[i + j])
continue checkEscapes; continue checkEscapes;
hits++; hits++;
i += seq.length - 1; i += seq.length - 1;
continue scanInput; continue scanInput;
} }
misses++; misses++;
} }
// Shift in/out // Shift in/out
if (text[i] == 0x0e || text[i] == 0x0f) if (text[i] == 0x0e || text[i] == 0x0f)
shifts++; shifts++;
} }
if (hits == 0) if (hits == 0)
return null; return null;
// //
// Initial quality is based on relative proportion of recongized vs. // Initial quality is based on relative proportion of recongized vs.
// unrecognized escape sequences. // unrecognized escape sequences.
// All good: quality = 100; // All good: quality = 100;
// half or less good: quality = 0; // half or less good: quality = 0;
// linear inbetween. // linear inbetween.
quality = (100 * hits - 100 * misses) / (hits + misses); quality = (100 * hits - 100 * misses) / (hits + misses);
// Back off quality if there were too few escape sequences seen. // Back off quality if there were too few escape sequences seen.
// Include shifts in this computation, so that KR does not get penalized // Include shifts in this computation, so that KR does not get penalized
// for having only a single Escape sequence, but many shifts. // for having only a single Escape sequence, but many shifts.
if (hits + shifts < 5) if (hits + shifts < 5)
quality -= (5 - (hits + shifts)) * 10; quality -= (5 - (hits + shifts)) * 10;
return quality <= 0 ? null : new Match(det, this, quality); return quality <= 0 ? null : new Match(det, this, quality);
}; };
module.exports.ISO_2022_JP = function() { module.exports.ISO_2022_JP = function() {
this.name = function() { this.name = function() {
return 'ISO-2022-JP'; return 'ISO-2022-JP';
}; };
this.escapeSequences = [ this.escapeSequences = [
[ 0x1b, 0x24, 0x28, 0x43 ], // KS X 1001:1992 [ 0x1b, 0x24, 0x28, 0x43 ], // KS X 1001:1992
[ 0x1b, 0x24, 0x28, 0x44 ], // JIS X 212-1990 [ 0x1b, 0x24, 0x28, 0x44 ], // JIS X 212-1990
[ 0x1b, 0x24, 0x40 ], // JIS C 6226-1978 [ 0x1b, 0x24, 0x40 ], // JIS C 6226-1978
[ 0x1b, 0x24, 0x41 ], // GB 2312-80 [ 0x1b, 0x24, 0x41 ], // GB 2312-80
[ 0x1b, 0x24, 0x42 ], // JIS X 208-1983 [ 0x1b, 0x24, 0x42 ], // JIS X 208-1983
[ 0x1b, 0x26, 0x40 ], // JIS X 208 1990, 1997 [ 0x1b, 0x26, 0x40 ], // JIS X 208 1990, 1997
[ 0x1b, 0x28, 0x42 ], // ASCII [ 0x1b, 0x28, 0x42 ], // ASCII
[ 0x1b, 0x28, 0x48 ], // JIS-Roman [ 0x1b, 0x28, 0x48 ], // JIS-Roman
[ 0x1b, 0x28, 0x49 ], // Half-width katakana [ 0x1b, 0x28, 0x49 ], // Half-width katakana
[ 0x1b, 0x28, 0x4a ], // JIS-Roman [ 0x1b, 0x28, 0x4a ], // JIS-Roman
[ 0x1b, 0x2e, 0x41 ], // ISO 8859-1 [ 0x1b, 0x2e, 0x41 ], // ISO 8859-1
[ 0x1b, 0x2e, 0x46 ] // ISO 8859-7 [ 0x1b, 0x2e, 0x46 ] // ISO 8859-7
]; ];
}; };
util.inherits(module.exports.ISO_2022_JP, ISO_2022); util.inherits(module.exports.ISO_2022_JP, ISO_2022);
module.exports.ISO_2022_KR = function() { module.exports.ISO_2022_KR = function() {
this.name = function() { this.name = function() {
return 'ISO-2022-KR'; return 'ISO-2022-KR';
}; };
this.escapeSequences = [ this.escapeSequences = [
[ 0x1b, 0x24, 0x29, 0x43 ] [ 0x1b, 0x24, 0x29, 0x43 ]
]; ];
}; };
util.inherits(module.exports.ISO_2022_KR, ISO_2022); util.inherits(module.exports.ISO_2022_KR, ISO_2022);
module.exports.ISO_2022_CN = function() { module.exports.ISO_2022_CN = function() {
this.name = function() { this.name = function() {
return 'ISO-2022-CN'; return 'ISO-2022-CN';
}; };
this.escapeSequences = [ this.escapeSequences = [
[ 0x1b, 0x24, 0x29, 0x41 ], // GB 2312-80 [ 0x1b, 0x24, 0x29, 0x41 ], // GB 2312-80
[ 0x1b, 0x24, 0x29, 0x47 ], // CNS 11643-1992 Plane 1 [ 0x1b, 0x24, 0x29, 0x47 ], // CNS 11643-1992 Plane 1
[ 0x1b, 0x24, 0x2A, 0x48 ], // CNS 11643-1992 Plane 2 [ 0x1b, 0x24, 0x2A, 0x48 ], // CNS 11643-1992 Plane 2
[ 0x1b, 0x24, 0x29, 0x45 ], // ISO-IR-165 [ 0x1b, 0x24, 0x29, 0x45 ], // ISO-IR-165
[ 0x1b, 0x24, 0x2B, 0x49 ], // CNS 11643-1992 Plane 3 [ 0x1b, 0x24, 0x2B, 0x49 ], // CNS 11643-1992 Plane 3
[ 0x1b, 0x24, 0x2B, 0x4A ], // CNS 11643-1992 Plane 4 [ 0x1b, 0x24, 0x2B, 0x4A ], // CNS 11643-1992 Plane 4
[ 0x1b, 0x24, 0x2B, 0x4B ], // CNS 11643-1992 Plane 5 [ 0x1b, 0x24, 0x2B, 0x4B ], // CNS 11643-1992 Plane 5
[ 0x1b, 0x24, 0x2B, 0x4C ], // CNS 11643-1992 Plane 6 [ 0x1b, 0x24, 0x2B, 0x4C ], // CNS 11643-1992 Plane 6
[ 0x1b, 0x24, 0x2B, 0x4D ], // CNS 11643-1992 Plane 7 [ 0x1b, 0x24, 0x2B, 0x4D ], // CNS 11643-1992 Plane 7
[ 0x1b, 0x4e ], // SS2 [ 0x1b, 0x4e ], // SS2
[ 0x1b, 0x4f ] // SS3 [ 0x1b, 0x4f ] // SS3
]; ];
}; };
util.inherits(module.exports.ISO_2022_CN, ISO_2022); util.inherits(module.exports.ISO_2022_CN, ISO_2022);

686
encoding/mbcs.js
View File

@ -1,30 +1,30 @@
var util = require('util'), var util = require('util'),
Match = require ('../match'); Match = require ('../match');
/** /**
* Binary search implementation (recursive) * Binary search implementation (recursive)
*/ */
function binarySearch(arr, searchValue) { function binarySearch(arr, searchValue) {
function find(arr, searchValue, left, right) { function find(arr, searchValue, left, right) {
if (right < left) if (right < left)
return -1; return -1;
/* /*
int mid = mid = (left + right) / 2; int mid = mid = (left + right) / 2;
There is a bug in the above line; There is a bug in the above line;
Joshua Bloch suggests the following replacement: Joshua Bloch suggests the following replacement:
*/ */
var mid = Math.floor((left + right) >>> 1); var mid = Math.floor((left + right) >>> 1);
if (searchValue > arr[mid]) if (searchValue > arr[mid])
return find(arr, searchValue, mid + 1, right); return find(arr, searchValue, mid + 1, right);
if (searchValue < arr[mid]) if (searchValue < arr[mid])
return find(arr, searchValue, left, mid - 1); return find(arr, searchValue, left, mid - 1);
return mid; return mid;
}; };
return find(arr, searchValue, 0, arr.length - 1); return find(arr, searchValue, 0, arr.length - 1);
}; };
// 'Character' iterated character class. // 'Character' iterated character class.
@ -40,28 +40,28 @@ function binarySearch(arr, searchValue) {
// //
function IteratedChar() { function IteratedChar() {
this.charValue = 0; // 1-4 bytes from the raw input data this.charValue = 0; // 1-4 bytes from the raw input data
this.index = 0; this.index = 0;
this.nextIndex = 0;
this.error = false;
this.done = false;
this.reset = function() {
this.charValue = 0;
this.index = -1;
this.nextIndex = 0; this.nextIndex = 0;
this.error = false; this.error = false;
this.done = false; this.done = false;
};
this.reset = function() { this.nextByte = function(det) {
this.charValue = 0; if (this.nextIndex >= det.fRawLength) {
this.index = -1; this.done = true;
this.nextIndex = 0; return -1;
this.error = false; }
this.done = false; var byteValue = det.fRawInput[this.nextIndex++] & 0x00ff;
}; return byteValue;
};
this.nextByte = function(det) {
if (this.nextIndex >= det.fRawLength) {
this.done = true;
return -1;
}
var byteValue = det.fRawInput[this.nextIndex++] & 0x00ff;
return byteValue;
};
}; };
@ -87,87 +87,87 @@ function mbcs() {};
*/ */
mbcs.prototype.match = function(det) { mbcs.prototype.match = function(det) {
var singleByteCharCount = 0, //TODO Do we really need this? var singleByteCharCount = 0, //TODO Do we really need this?
doubleByteCharCount = 0, doubleByteCharCount = 0,
commonCharCount = 0, commonCharCount = 0,
badCharCount = 0, badCharCount = 0,
totalCharCount = 0, totalCharCount = 0,
confidence = 0; confidence = 0;
var iter = new IteratedChar(); var iter = new IteratedChar();
detectBlock: { detectBlock: {
for (iter.reset(); this.nextChar(iter, det);) { for (iter.reset(); this.nextChar(iter, det);) {
totalCharCount++; totalCharCount++;
if (iter.error) { if (iter.error) {
badCharCount++; badCharCount++;
} else { } else {
var cv = iter.charValue & 0xFFFFFFFF; var cv = iter.charValue & 0xFFFFFFFF;
if (cv <= 0xff) { if (cv <= 0xff) {
singleByteCharCount++; singleByteCharCount++;
} else {
doubleByteCharCount++;
if (this.commonChars != null) {
// NOTE: This assumes that there are no 4-byte common chars.
if (binarySearch(this.commonChars, cv) >= 0) {
commonCharCount++;
}
}
}
}
if (badCharCount >= 2 && badCharCount * 5 >= doubleByteCharCount) {
// console.log('its here!')
// Bail out early if the byte data is not matching the encoding scheme.
break detectBlock;
}
}
if (doubleByteCharCount <= 10 && badCharCount== 0) {
// Not many multi-byte chars.
if (doubleByteCharCount == 0 && totalCharCount < 10) {
// There weren't any multibyte sequences, and there was a low density of non-ASCII single bytes.
// We don't have enough data to have any confidence.
// Statistical analysis of single byte non-ASCII charcters would probably help here.
confidence = 0;
}
else {
// ASCII or ISO file? It's probably not our encoding,
// but is not incompatible with our encoding, so don't give it a zero.
confidence = 10;
}
break detectBlock;
}
//
// No match if there are too many characters that don't fit the encoding scheme.
// (should we have zero tolerance for these?)
//
if (doubleByteCharCount < 20 * badCharCount) {
confidence = 0;
break detectBlock;
}
if (this.commonChars == null) {
// We have no statistics on frequently occuring characters.
// Assess confidence purely on having a reasonable number of
// multi-byte characters (the more the better
confidence = 30 + doubleByteCharCount - 20 * badCharCount;
if (confidence > 100) {
confidence = 100;
}
} else { } else {
// doubleByteCharCount++;
// Frequency of occurence statistics exist. if (this.commonChars != null) {
// // NOTE: This assumes that there are no 4-byte common chars.
var maxVal = Math.log(parseFloat(doubleByteCharCount) / 4); if (binarySearch(this.commonChars, cv) >= 0) {
var scaleFactor = 90.0 / maxVal; commonCharCount++;
confidence = Math.floor(Math.log(commonCharCount + 1) * scaleFactor + 10); }
confidence = Math.min(confidence, 100); }
} }
} // end of detectBlock: }
if (badCharCount >= 2 && badCharCount * 5 >= doubleByteCharCount) {
// console.log('its here!')
// Bail out early if the byte data is not matching the encoding scheme.
break detectBlock;
}
}
return confidence == 0 ? null : new Match(det, this, confidence); if (doubleByteCharCount <= 10 && badCharCount== 0) {
// Not many multi-byte chars.
if (doubleByteCharCount == 0 && totalCharCount < 10) {
// There weren't any multibyte sequences, and there was a low density of non-ASCII single bytes.
// We don't have enough data to have any confidence.
// Statistical analysis of single byte non-ASCII charcters would probably help here.
confidence = 0;
}
else {
// ASCII or ISO file? It's probably not our encoding,
// but is not incompatible with our encoding, so don't give it a zero.
confidence = 10;
}
break detectBlock;
}
//
// No match if there are too many characters that don't fit the encoding scheme.
// (should we have zero tolerance for these?)
//
if (doubleByteCharCount < 20 * badCharCount) {
confidence = 0;
break detectBlock;
}
if (this.commonChars == null) {
// We have no statistics on frequently occuring characters.
// Assess confidence purely on having a reasonable number of
// multi-byte characters (the more the better
confidence = 30 + doubleByteCharCount - 20 * badCharCount;
if (confidence > 100) {
confidence = 100;
}
} else {
//
// Frequency of occurence statistics exist.
//
var maxVal = Math.log(parseFloat(doubleByteCharCount) / 4);
var scaleFactor = 90.0 / maxVal;
confidence = Math.floor(Math.log(commonCharCount + 1) * scaleFactor + 10);
confidence = Math.min(confidence, 100);
}
} // end of detectBlock:
return confidence == 0 ? null : new Match(det, this, confidence);
}; };
/** /**
@ -191,48 +191,48 @@ mbcs.prototype.nextChar = function(iter, det) {};
* Shift-JIS charset recognizer. * Shift-JIS charset recognizer.
*/ */
module.exports.sjis = function() { module.exports.sjis = function() {
this.name = function() { this.name = function() {
return 'Shift-JIS'; return 'Shift-JIS';
}; };
this.language = function() { this.language = function() {
return 'ja'; return 'ja';
}; };
// TODO: This set of data comes from the character frequency- // TODO: This set of data comes from the character frequency-
// of-occurence analysis tool. The data needs to be moved // of-occurence analysis tool. The data needs to be moved
// into a resource and loaded from there. // into a resource and loaded from there.
this.commonChars = [ this.commonChars = [
0x8140, 0x8141, 0x8142, 0x8145, 0x815b, 0x8169, 0x816a, 0x8175, 0x8176, 0x82a0, 0x8140, 0x8141, 0x8142, 0x8145, 0x815b, 0x8169, 0x816a, 0x8175, 0x8176, 0x82a0,
0x82a2, 0x82a4, 0x82a9, 0x82aa, 0x82ab, 0x82ad, 0x82af, 0x82b1, 0x82b3, 0x82b5, 0x82a2, 0x82a4, 0x82a9, 0x82aa, 0x82ab, 0x82ad, 0x82af, 0x82b1, 0x82b3, 0x82b5,
0x82b7, 0x82bd, 0x82be, 0x82c1, 0x82c4, 0x82c5, 0x82c6, 0x82c8, 0x82c9, 0x82cc, 0x82b7, 0x82bd, 0x82be, 0x82c1, 0x82c4, 0x82c5, 0x82c6, 0x82c8, 0x82c9, 0x82cc,
0x82cd, 0x82dc, 0x82e0, 0x82e7, 0x82e8, 0x82e9, 0x82ea, 0x82f0, 0x82f1, 0x8341, 0x82cd, 0x82dc, 0x82e0, 0x82e7, 0x82e8, 0x82e9, 0x82ea, 0x82f0, 0x82f1, 0x8341,
0x8343, 0x834e, 0x834f, 0x8358, 0x835e, 0x8362, 0x8367, 0x8375, 0x8376, 0x8389, 0x8343, 0x834e, 0x834f, 0x8358, 0x835e, 0x8362, 0x8367, 0x8375, 0x8376, 0x8389,
0x838a, 0x838b, 0x838d, 0x8393, 0x8e96, 0x93fa, 0x95aa 0x838a, 0x838b, 0x838d, 0x8393, 0x8e96, 0x93fa, 0x95aa
]; ];
this.nextChar = function(iter, det) { this.nextChar = function(iter, det) {
iter.index = iter.nextIndex; iter.index = iter.nextIndex;
iter.error = false; iter.error = false;
var firstByte; var firstByte;
firstByte = iter.charValue = iter.nextByte(det); firstByte = iter.charValue = iter.nextByte(det);
if (firstByte < 0) if (firstByte < 0)
return false; return false;
if (firstByte <= 0x7f || (firstByte > 0xa0 && firstByte <= 0xdf)) if (firstByte <= 0x7f || (firstByte > 0xa0 && firstByte <= 0xdf))
return true; return true;
var secondByte = iter.nextByte(det); var secondByte = iter.nextByte(det);
if (secondByte < 0) if (secondByte < 0)
return false; return false;
iter.charValue = (firstByte << 8) | secondByte; iter.charValue = (firstByte << 8) | secondByte;
if (! ((secondByte >= 0x40 && secondByte <= 0x7f) || (secondByte >= 0x80 && secondByte <= 0xff))) { if (! ((secondByte >= 0x40 && secondByte <= 0x7f) || (secondByte >= 0x80 && secondByte <= 0xff))) {
// Illegal second byte value. // Illegal second byte value.
iter.error = true; iter.error = true;
} }
return true; return true;
}; };
}; };
util.inherits(module.exports.sjis, mbcs); util.inherits(module.exports.sjis, mbcs);
@ -242,52 +242,52 @@ util.inherits(module.exports.sjis, mbcs);
* Big5 charset recognizer. * Big5 charset recognizer.
*/ */
module.exports.big5 = function() { module.exports.big5 = function() {
this.name = function() { this.name = function() {
return 'Big5'; return 'Big5';
}; };
this.language = function() { this.language = function() {
return 'zh'; return 'zh';
}; };
// TODO: This set of data comes from the character frequency- // TODO: This set of data comes from the character frequency-
// of-occurence analysis tool. The data needs to be moved // of-occurence analysis tool. The data needs to be moved
// into a resource and loaded from there. // into a resource and loaded from there.
this.commonChars = [ this.commonChars = [
0xa140, 0xa141, 0xa142, 0xa143, 0xa147, 0xa149, 0xa175, 0xa176, 0xa440, 0xa446, 0xa140, 0xa141, 0xa142, 0xa143, 0xa147, 0xa149, 0xa175, 0xa176, 0xa440, 0xa446,
0xa447, 0xa448, 0xa451, 0xa454, 0xa457, 0xa464, 0xa46a, 0xa46c, 0xa477, 0xa4a3, 0xa447, 0xa448, 0xa451, 0xa454, 0xa457, 0xa464, 0xa46a, 0xa46c, 0xa477, 0xa4a3,
0xa4a4, 0xa4a7, 0xa4c1, 0xa4ce, 0xa4d1, 0xa4df, 0xa4e8, 0xa4fd, 0xa540, 0xa548, 0xa4a4, 0xa4a7, 0xa4c1, 0xa4ce, 0xa4d1, 0xa4df, 0xa4e8, 0xa4fd, 0xa540, 0xa548,
0xa558, 0xa569, 0xa5cd, 0xa5e7, 0xa657, 0xa661, 0xa662, 0xa668, 0xa670, 0xa6a8, 0xa558, 0xa569, 0xa5cd, 0xa5e7, 0xa657, 0xa661, 0xa662, 0xa668, 0xa670, 0xa6a8,
0xa6b3, 0xa6b9, 0xa6d3, 0xa6db, 0xa6e6, 0xa6f2, 0xa740, 0xa751, 0xa759, 0xa7da, 0xa6b3, 0xa6b9, 0xa6d3, 0xa6db, 0xa6e6, 0xa6f2, 0xa740, 0xa751, 0xa759, 0xa7da,
0xa8a3, 0xa8a5, 0xa8ad, 0xa8d1, 0xa8d3, 0xa8e4, 0xa8fc, 0xa9c0, 0xa9d2, 0xa9f3, 0xa8a3, 0xa8a5, 0xa8ad, 0xa8d1, 0xa8d3, 0xa8e4, 0xa8fc, 0xa9c0, 0xa9d2, 0xa9f3,
0xaa6b, 0xaaba, 0xaabe, 0xaacc, 0xaafc, 0xac47, 0xac4f, 0xacb0, 0xacd2, 0xad59, 0xaa6b, 0xaaba, 0xaabe, 0xaacc, 0xaafc, 0xac47, 0xac4f, 0xacb0, 0xacd2, 0xad59,
0xaec9, 0xafe0, 0xb0ea, 0xb16f, 0xb2b3, 0xb2c4, 0xb36f, 0xb44c, 0xb44e, 0xb54c, 0xaec9, 0xafe0, 0xb0ea, 0xb16f, 0xb2b3, 0xb2c4, 0xb36f, 0xb44c, 0xb44e, 0xb54c,
0xb5a5, 0xb5bd, 0xb5d0, 0xb5d8, 0xb671, 0xb7ed, 0xb867, 0xb944, 0xbad8, 0xbb44, 0xb5a5, 0xb5bd, 0xb5d0, 0xb5d8, 0xb671, 0xb7ed, 0xb867, 0xb944, 0xbad8, 0xbb44,
0xbba1, 0xbdd1, 0xc2c4, 0xc3b9, 0xc440, 0xc45f 0xbba1, 0xbdd1, 0xc2c4, 0xc3b9, 0xc440, 0xc45f
]; ];
this.nextChar = function(iter, det) { this.nextChar = function(iter, det) {
iter.index = iter.nextIndex; iter.index = iter.nextIndex;
iter.error = false; iter.error = false;
var firstByte = iter.charValue = iter.nextByte(det); var firstByte = iter.charValue = iter.nextByte(det);
if (firstByte < 0) if (firstByte < 0)
return false; return false;
// single byte character. // single byte character.
if (firstByte <= 0x7f || firstByte == 0xff) if (firstByte <= 0x7f || firstByte == 0xff)
return true; return true;
var secondByte = iter.nextByte(det); var secondByte = iter.nextByte(det);
if (secondByte < 0) if (secondByte < 0)
return false; return false;
iter.charValue = (iter.charValue << 8) | secondByte; iter.charValue = (iter.charValue << 8) | secondByte;
if (secondByte < 0x40 || secondByte == 0x7f || secondByte == 0xff) if (secondByte < 0x40 || secondByte == 0x7f || secondByte == 0xff)
iter.error = true; iter.error = true;
return true; return true;
}; };
}; };
util.inherits(module.exports.big5, mbcs); util.inherits(module.exports.big5, mbcs);
@ -303,55 +303,55 @@ util.inherits(module.exports.big5, mbcs);
* packed into an int. * packed into an int.
*/ */
function eucNextChar(iter, det) { function eucNextChar(iter, det) {
iter.index = iter.nextIndex; iter.index = iter.nextIndex;
iter.error = false; iter.error = false;
var firstByte = 0; var firstByte = 0;
var secondByte = 0; var secondByte = 0;
var thirdByte = 0; var thirdByte = 0;
//int fourthByte = 0; //int fourthByte = 0;
buildChar: { buildChar: {
firstByte = iter.charValue = iter.nextByte(det); firstByte = iter.charValue = iter.nextByte(det);
if (firstByte < 0) { if (firstByte < 0) {
// Ran off the end of the input data // Ran off the end of the input data
iter.done = true; iter.done = true;
break buildChar; break buildChar;
}
if (firstByte <= 0x8d) {
// single byte char
break buildChar;
}
secondByte = iter.nextByte(det);
iter.charValue = (iter.charValue << 8) | secondByte;
if (firstByte >= 0xA1 && firstByte <= 0xfe) {
// Two byte Char
if (secondByte < 0xa1) {
iter.error = true;
}
break buildChar;
}
if (firstByte == 0x8e) {
// Code Set 2.
// In EUC-JP, total char size is 2 bytes, only one byte of actual char value.
// In EUC-TW, total char size is 4 bytes, three bytes contribute to char value.
// We don't know which we've got.
// Treat it like EUC-JP. If the data really was EUC-TW, the following two
// bytes will look like a well formed 2 byte char.
if (secondByte < 0xa1) {
iter.error = true;
}
break buildChar;
}
if (firstByte == 0x8f) {
// Code set 3.
// Three byte total char size, two bytes of actual char value.
thirdByte = iter.nextByte(det);
iter.charValue = (iter.charValue << 8) | thirdByte;
if (thirdByte < 0xa1) {
iter.error = true;
}
}
} }
return iter.done == false; if (firstByte <= 0x8d) {
// single byte char
break buildChar;
}
secondByte = iter.nextByte(det);
iter.charValue = (iter.charValue << 8) | secondByte;
if (firstByte >= 0xA1 && firstByte <= 0xfe) {
// Two byte Char
if (secondByte < 0xa1) {
iter.error = true;
}
break buildChar;
}
if (firstByte == 0x8e) {
// Code Set 2.
// In EUC-JP, total char size is 2 bytes, only one byte of actual char value.
// In EUC-TW, total char size is 4 bytes, three bytes contribute to char value.
// We don't know which we've got.
// Treat it like EUC-JP. If the data really was EUC-TW, the following two
// bytes will look like a well formed 2 byte char.
if (secondByte < 0xa1) {
iter.error = true;
}
break buildChar;
}
if (firstByte == 0x8f) {
// Code set 3.
// Three byte total char size, two bytes of actual char value.
thirdByte = iter.nextByte(det);
iter.charValue = (iter.charValue << 8) | thirdByte;
if (thirdByte < 0xa1) {
iter.error = true;
}
}
}
return iter.done == false;
}; };
@ -361,30 +361,30 @@ function eucNextChar(iter, det) {
* is created and kept by the public CharsetDetector class * is created and kept by the public CharsetDetector class
*/ */
module.exports.euc_jp = function() { module.exports.euc_jp = function() {
this.name = function() { this.name = function() {
return 'EUC-JP'; return 'EUC-JP';
}; };
this.language = function() { this.language = function() {
return 'ja'; return 'ja';
}; };
// TODO: This set of data comes from the character frequency- // TODO: This set of data comes from the character frequency-
// of-occurence analysis tool. The data needs to be moved // of-occurence analysis tool. The data needs to be moved
// into a resource and loaded from there. // into a resource and loaded from there.
this.commonChars = [ this.commonChars = [
0xa1a1, 0xa1a2, 0xa1a3, 0xa1a6, 0xa1bc, 0xa1ca, 0xa1cb, 0xa1d6, 0xa1d7, 0xa4a2, 0xa1a1, 0xa1a2, 0xa1a3, 0xa1a6, 0xa1bc, 0xa1ca, 0xa1cb, 0xa1d6, 0xa1d7, 0xa4a2,
0xa4a4, 0xa4a6, 0xa4a8, 0xa4aa, 0xa4ab, 0xa4ac, 0xa4ad, 0xa4af, 0xa4b1, 0xa4b3, 0xa4a4, 0xa4a6, 0xa4a8, 0xa4aa, 0xa4ab, 0xa4ac, 0xa4ad, 0xa4af, 0xa4b1, 0xa4b3,
0xa4b5, 0xa4b7, 0xa4b9, 0xa4bb, 0xa4bd, 0xa4bf, 0xa4c0, 0xa4c1, 0xa4c3, 0xa4c4, 0xa4b5, 0xa4b7, 0xa4b9, 0xa4bb, 0xa4bd, 0xa4bf, 0xa4c0, 0xa4c1, 0xa4c3, 0xa4c4,
0xa4c6, 0xa4c7, 0xa4c8, 0xa4c9, 0xa4ca, 0xa4cb, 0xa4ce, 0xa4cf, 0xa4d0, 0xa4de, 0xa4c6, 0xa4c7, 0xa4c8, 0xa4c9, 0xa4ca, 0xa4cb, 0xa4ce, 0xa4cf, 0xa4d0, 0xa4de,
0xa4df, 0xa4e1, 0xa4e2, 0xa4e4, 0xa4e8, 0xa4e9, 0xa4ea, 0xa4eb, 0xa4ec, 0xa4ef, 0xa4df, 0xa4e1, 0xa4e2, 0xa4e4, 0xa4e8, 0xa4e9, 0xa4ea, 0xa4eb, 0xa4ec, 0xa4ef,
0xa4f2, 0xa4f3, 0xa5a2, 0xa5a3, 0xa5a4, 0xa5a6, 0xa5a7, 0xa5aa, 0xa5ad, 0xa5af, 0xa4f2, 0xa4f3, 0xa5a2, 0xa5a3, 0xa5a4, 0xa5a6, 0xa5a7, 0xa5aa, 0xa5ad, 0xa5af,
0xa5b0, 0xa5b3, 0xa5b5, 0xa5b7, 0xa5b8, 0xa5b9, 0xa5bf, 0xa5c3, 0xa5c6, 0xa5c7, 0xa5b0, 0xa5b3, 0xa5b5, 0xa5b7, 0xa5b8, 0xa5b9, 0xa5bf, 0xa5c3, 0xa5c6, 0xa5c7,
0xa5c8, 0xa5c9, 0xa5cb, 0xa5d0, 0xa5d5, 0xa5d6, 0xa5d7, 0xa5de, 0xa5e0, 0xa5e1, 0xa5c8, 0xa5c9, 0xa5cb, 0xa5d0, 0xa5d5, 0xa5d6, 0xa5d7, 0xa5de, 0xa5e0, 0xa5e1,
0xa5e5, 0xa5e9, 0xa5ea, 0xa5eb, 0xa5ec, 0xa5ed, 0xa5f3, 0xb8a9, 0xb9d4, 0xbaee, 0xa5e5, 0xa5e9, 0xa5ea, 0xa5eb, 0xa5ec, 0xa5ed, 0xa5f3, 0xb8a9, 0xb9d4, 0xbaee,
0xbbc8, 0xbef0, 0xbfb7, 0xc4ea, 0xc6fc, 0xc7bd, 0xcab8, 0xcaf3, 0xcbdc, 0xcdd1 0xbbc8, 0xbef0, 0xbfb7, 0xc4ea, 0xc6fc, 0xc7bd, 0xcab8, 0xcaf3, 0xcbdc, 0xcdd1
]; ];
this.nextChar = eucNextChar; this.nextChar = eucNextChar;
}; };
util.inherits(module.exports.euc_jp, mbcs); util.inherits(module.exports.euc_jp, mbcs);
@ -395,30 +395,30 @@ util.inherits(module.exports.euc_jp, mbcs);
* is created and kept by the public CharsetDetector class * is created and kept by the public CharsetDetector class
*/ */
module.exports.euc_kr = function() { module.exports.euc_kr = function() {
this.name = function() { this.name = function() {
return 'EUC-KR'; return 'EUC-KR';
}; };
this.language = function() { this.language = function() {
return 'ko'; return 'ko';
}; };
// TODO: This set of data comes from the character frequency- // TODO: This set of data comes from the character frequency-
// of-occurence analysis tool. The data needs to be moved // of-occurence analysis tool. The data needs to be moved
// into a resource and loaded from there. // into a resource and loaded from there.
this.commonChars = [ this.commonChars = [
0xb0a1, 0xb0b3, 0xb0c5, 0xb0cd, 0xb0d4, 0xb0e6, 0xb0ed, 0xb0f8, 0xb0fa, 0xb0fc, 0xb0a1, 0xb0b3, 0xb0c5, 0xb0cd, 0xb0d4, 0xb0e6, 0xb0ed, 0xb0f8, 0xb0fa, 0xb0fc,
0xb1b8, 0xb1b9, 0xb1c7, 0xb1d7, 0xb1e2, 0xb3aa, 0xb3bb, 0xb4c2, 0xb4cf, 0xb4d9, 0xb1b8, 0xb1b9, 0xb1c7, 0xb1d7, 0xb1e2, 0xb3aa, 0xb3bb, 0xb4c2, 0xb4cf, 0xb4d9,
0xb4eb, 0xb5a5, 0xb5b5, 0xb5bf, 0xb5c7, 0xb5e9, 0xb6f3, 0xb7af, 0xb7c2, 0xb7ce, 0xb4eb, 0xb5a5, 0xb5b5, 0xb5bf, 0xb5c7, 0xb5e9, 0xb6f3, 0xb7af, 0xb7c2, 0xb7ce,
0xb8a6, 0xb8ae, 0xb8b6, 0xb8b8, 0xb8bb, 0xb8e9, 0xb9ab, 0xb9ae, 0xb9cc, 0xb9ce, 0xb8a6, 0xb8ae, 0xb8b6, 0xb8b8, 0xb8bb, 0xb8e9, 0xb9ab, 0xb9ae, 0xb9cc, 0xb9ce,
0xb9fd, 0xbab8, 0xbace, 0xbad0, 0xbaf1, 0xbbe7, 0xbbf3, 0xbbfd, 0xbcad, 0xbcba, 0xb9fd, 0xbab8, 0xbace, 0xbad0, 0xbaf1, 0xbbe7, 0xbbf3, 0xbbfd, 0xbcad, 0xbcba,
0xbcd2, 0xbcf6, 0xbdba, 0xbdc0, 0xbdc3, 0xbdc5, 0xbec6, 0xbec8, 0xbedf, 0xbeee, 0xbcd2, 0xbcf6, 0xbdba, 0xbdc0, 0xbdc3, 0xbdc5, 0xbec6, 0xbec8, 0xbedf, 0xbeee,
0xbef8, 0xbefa, 0xbfa1, 0xbfa9, 0xbfc0, 0xbfe4, 0xbfeb, 0xbfec, 0xbff8, 0xc0a7, 0xbef8, 0xbefa, 0xbfa1, 0xbfa9, 0xbfc0, 0xbfe4, 0xbfeb, 0xbfec, 0xbff8, 0xc0a7,
0xc0af, 0xc0b8, 0xc0ba, 0xc0bb, 0xc0bd, 0xc0c7, 0xc0cc, 0xc0ce, 0xc0cf, 0xc0d6, 0xc0af, 0xc0b8, 0xc0ba, 0xc0bb, 0xc0bd, 0xc0c7, 0xc0cc, 0xc0ce, 0xc0cf, 0xc0d6,
0xc0da, 0xc0e5, 0xc0fb, 0xc0fc, 0xc1a4, 0xc1a6, 0xc1b6, 0xc1d6, 0xc1df, 0xc1f6, 0xc0da, 0xc0e5, 0xc0fb, 0xc0fc, 0xc1a4, 0xc1a6, 0xc1b6, 0xc1d6, 0xc1df, 0xc1f6,
0xc1f8, 0xc4a1, 0xc5cd, 0xc6ae, 0xc7cf, 0xc7d1, 0xc7d2, 0xc7d8, 0xc7e5, 0xc8ad 0xc1f8, 0xc4a1, 0xc5cd, 0xc6ae, 0xc7cf, 0xc7d1, 0xc7d2, 0xc7d8, 0xc7e5, 0xc8ad
]; ];
this.nextChar = eucNextChar; this.nextChar = eucNextChar;
}; };
util.inherits(module.exports.euc_kr, mbcs); util.inherits(module.exports.euc_kr, mbcs);
@ -428,75 +428,75 @@ util.inherits(module.exports.euc_kr, mbcs);
* GB-18030 recognizer. Uses simplified Chinese statistics. * GB-18030 recognizer. Uses simplified Chinese statistics.
*/ */
module.exports.gb_18030 = function() { module.exports.gb_18030 = function() {
this.name = function() { this.name = function() {
return 'GB18030'; return 'GB18030';
}; };
this.language = function() { this.language = function() {
return 'zh'; return 'zh';
}; };
/* /*
* Get the next character value for EUC based encodings. * Get the next character value for EUC based encodings.
* Character 'value' is simply the raw bytes that make up the character * Character 'value' is simply the raw bytes that make up the character
* packed into an int. * packed into an int.
*/ */
this.nextChar = function(iter, det) { this.nextChar = function(iter, det) {
iter.index = iter.nextIndex; iter.index = iter.nextIndex;
iter.error = false; iter.error = false;
var firstByte = 0; var firstByte = 0;
var secondByte = 0; var secondByte = 0;
var thirdByte = 0; var thirdByte = 0;
var fourthByte = 0; var fourthByte = 0;
buildChar: { buildChar: {
firstByte = iter.charValue = iter.nextByte(det); firstByte = iter.charValue = iter.nextByte(det);
if (firstByte < 0) { if (firstByte < 0) {
// Ran off the end of the input data // Ran off the end of the input data
iter.done = true; iter.done = true;
break buildChar; break buildChar;
} }
if (firstByte <= 0x80) { if (firstByte <= 0x80) {
// single byte char // single byte char
break buildChar; break buildChar;
} }
secondByte = iter.nextByte(det); secondByte = iter.nextByte(det);
iter.charValue = (iter.charValue << 8) | secondByte; iter.charValue = (iter.charValue << 8) | secondByte;
if (firstByte >= 0x81 && firstByte <= 0xFE) { if (firstByte >= 0x81 && firstByte <= 0xFE) {
// Two byte Char // Two byte Char
if ((secondByte >= 0x40 && secondByte <= 0x7E) || (secondByte >=80 && secondByte <= 0xFE)) { if ((secondByte >= 0x40 && secondByte <= 0x7E) || (secondByte >=80 && secondByte <= 0xFE)) {
break buildChar; break buildChar;
}
// Four byte char
if (secondByte >= 0x30 && secondByte <= 0x39) {
thirdByte = iter.nextByte(det);
if (thirdByte >= 0x81 && thirdByte <= 0xFE) {
fourthByte = iter.nextByte(det);
if (fourthByte >= 0x30 && fourthByte <= 0x39) {
iter.charValue = (iter.charValue << 16) | (thirdByte << 8) | fourthByte;
break buildChar;
}
}
}
iter.error = true;
break buildChar;
}
} }
return iter.done == false; // Four byte char
}; if (secondByte >= 0x30 && secondByte <= 0x39) {
thirdByte = iter.nextByte(det);
if (thirdByte >= 0x81 && thirdByte <= 0xFE) {
fourthByte = iter.nextByte(det);
if (fourthByte >= 0x30 && fourthByte <= 0x39) {
iter.charValue = (iter.charValue << 16) | (thirdByte << 8) | fourthByte;
break buildChar;
}
}
}
iter.error = true;
break buildChar;
}
}
return iter.done == false;
};
// TODO: This set of data comes from the character frequency- // TODO: This set of data comes from the character frequency-
// of-occurence analysis tool. The data needs to be moved // of-occurence analysis tool. The data needs to be moved
// into a resource and loaded from there. // into a resource and loaded from there.
this.commonChars = [ this.commonChars = [
0xa1a1, 0xa1a2, 0xa1a3, 0xa1a4, 0xa1b0, 0xa1b1, 0xa1f1, 0xa1f3, 0xa3a1, 0xa3ac, 0xa1a1, 0xa1a2, 0xa1a3, 0xa1a4, 0xa1b0, 0xa1b1, 0xa1f1, 0xa1f3, 0xa3a1, 0xa3ac,
0xa3ba, 0xb1a8, 0xb1b8, 0xb1be, 0xb2bb, 0xb3c9, 0xb3f6, 0xb4f3, 0xb5bd, 0xb5c4, 0xa3ba, 0xb1a8, 0xb1b8, 0xb1be, 0xb2bb, 0xb3c9, 0xb3f6, 0xb4f3, 0xb5bd, 0xb5c4,
0xb5e3, 0xb6af, 0xb6d4, 0xb6e0, 0xb7a2, 0xb7a8, 0xb7bd, 0xb7d6, 0xb7dd, 0xb8b4, 0xb5e3, 0xb6af, 0xb6d4, 0xb6e0, 0xb7a2, 0xb7a8, 0xb7bd, 0xb7d6, 0xb7dd, 0xb8b4,
0xb8df, 0xb8f6, 0xb9ab, 0xb9c9, 0xb9d8, 0xb9fa, 0xb9fd, 0xbacd, 0xbba7, 0xbbd6, 0xb8df, 0xb8f6, 0xb9ab, 0xb9c9, 0xb9d8, 0xb9fa, 0xb9fd, 0xbacd, 0xbba7, 0xbbd6,
0xbbe1, 0xbbfa, 0xbcbc, 0xbcdb, 0xbcfe, 0xbdcc, 0xbecd, 0xbedd, 0xbfb4, 0xbfc6, 0xbbe1, 0xbbfa, 0xbcbc, 0xbcdb, 0xbcfe, 0xbdcc, 0xbecd, 0xbedd, 0xbfb4, 0xbfc6,
0xbfc9, 0xc0b4, 0xc0ed, 0xc1cb, 0xc2db, 0xc3c7, 0xc4dc, 0xc4ea, 0xc5cc, 0xc6f7, 0xbfc9, 0xc0b4, 0xc0ed, 0xc1cb, 0xc2db, 0xc3c7, 0xc4dc, 0xc4ea, 0xc5cc, 0xc6f7,
0xc7f8, 0xc8ab, 0xc8cb, 0xc8d5, 0xc8e7, 0xc9cf, 0xc9fa, 0xcab1, 0xcab5, 0xcac7, 0xc7f8, 0xc8ab, 0xc8cb, 0xc8d5, 0xc8e7, 0xc9cf, 0xc9fa, 0xcab1, 0xcab5, 0xcac7,
0xcad0, 0xcad6, 0xcaf5, 0xcafd, 0xccec, 0xcdf8, 0xceaa, 0xcec4, 0xced2, 0xcee5, 0xcad0, 0xcad6, 0xcaf5, 0xcafd, 0xccec, 0xcdf8, 0xceaa, 0xcec4, 0xced2, 0xcee5,
0xcfb5, 0xcfc2, 0xcfd6, 0xd0c2, 0xd0c5, 0xd0d0, 0xd0d4, 0xd1a7, 0xd2aa, 0xd2b2, 0xcfb5, 0xcfc2, 0xcfd6, 0xd0c2, 0xd0c5, 0xd0d0, 0xd0d4, 0xd1a7, 0xd2aa, 0xd2b2,
0xd2b5, 0xd2bb, 0xd2d4, 0xd3c3, 0xd3d0, 0xd3fd, 0xd4c2, 0xd4da, 0xd5e2, 0xd6d0 0xd2b5, 0xd2bb, 0xd2d4, 0xd3c3, 0xd3d0, 0xd3fd, 0xd4c2, 0xd4da, 0xd5e2, 0xd6d0
]; ];
}; };
util.inherits(module.exports.gb_18030, mbcs); util.inherits(module.exports.gb_18030, mbcs);

1542
encoding/sbcs.js
View File

File diff suppressed because it is too large Load Diff

146
encoding/unicode.js
View File

@ -1,107 +1,107 @@
var util = require('util'), var util = require('util'),
Match = require ('../match'); Match = require ('../match');
/** /**
* This class matches UTF-16 and UTF-32, both big- and little-endian. The * This class matches UTF-16 and UTF-32, both big- and little-endian. The
* BOM will be used if it is present. * BOM will be used if it is present.
*/ */
module.exports.UTF_16BE = function() { module.exports.UTF_16BE = function() {
this.name = function() { this.name = function() {
return 'UTF-16BE'; return 'UTF-16BE';
}; };
this.match = function(det) { this.match = function(det) {
var input = det.fRawInput; var input = det.fRawInput;
if (input.length >= 2 && ((input[0] & 0xff) == 0xfe && (input[1] & 0xff) == 0xff)) if (input.length >= 2 && ((input[0] & 0xff) == 0xfe && (input[1] & 0xff) == 0xff))
return new Match(det, this, confidence = 100); return new Match(det, this, confidence = 100);
// TODO: Do some statistics to check for unsigned UTF-16BE // TODO: Do some statistics to check for unsigned UTF-16BE
return null; return null;
}; };
}; };
module.exports.UTF_16LE = function() { module.exports.UTF_16LE = function() {
this.name = function() { this.name = function() {
return 'UTF-16LE'; return 'UTF-16LE';
}; };
this.match = function(det) { this.match = function(det) {
var input = det.fRawInput; var input = det.fRawInput;
if (input.length >= 2 && ((input[0] & 0xff) == 0xff && (input[1] & 0xff) == 0xfe)) { if (input.length >= 2 && ((input[0] & 0xff) == 0xff && (input[1] & 0xff) == 0xfe)) {
// An LE BOM is present. // An LE BOM is present.
if (input.length >= 4 && input[2] == 0x00 && input[3] == 0x00) if (input.length >= 4 && input[2] == 0x00 && input[3] == 0x00)
// It is probably UTF-32 LE, not UTF-16 // It is probably UTF-32 LE, not UTF-16
return null; return null;
return new Match(det, this, confidence = 100); return new Match(det, this, confidence = 100);
}
// TODO: Do some statistics to check for unsigned UTF-16LE
return null;
} }
// TODO: Do some statistics to check for unsigned UTF-16LE
return null;
}
}; };
function UTF_32() {}; function UTF_32() {};
UTF_32.prototype.match = function(det) { UTF_32.prototype.match = function(det) {
var input = det.fRawInput, var input = det.fRawInput,
limit = (det.fRawLength / 4) * 4, limit = (det.fRawLength / 4) * 4,
numValid = 0, numValid = 0,
numInvalid = 0, numInvalid = 0,
hasBOM = false, hasBOM = false,
confidence = 0; confidence = 0;
if (limit == 0) if (limit == 0)
return null; return null;
if (this.getChar(input, 0) == 0x0000FEFF) if (this.getChar(input, 0) == 0x0000FEFF)
hasBOM = true; hasBOM = true;
for (var i = 0; i < limit; i += 4) { for (var i = 0; i < limit; i += 4) {
var ch = this.getChar(input, i); var ch = this.getChar(input, i);
if (ch < 0 || ch >= 0x10FFFF || (ch >= 0xD800 && ch <= 0xDFFF)) if (ch < 0 || ch >= 0x10FFFF || (ch >= 0xD800 && ch <= 0xDFFF))
numInvalid += 1; numInvalid += 1;
else else
numValid += 1; numValid += 1;
} }
// Cook up some sort of confidence score, based on presence of a BOM // Cook up some sort of confidence score, based on presence of a BOM
// and the existence of valid and/or invalid multi-byte sequences. // and the existence of valid and/or invalid multi-byte sequences.
if (hasBOM && numInvalid == 0) { if (hasBOM && numInvalid == 0) {
confidence = 100; confidence = 100;
} else if (hasBOM && numValid > numInvalid * 10) { } else if (hasBOM && numValid > numInvalid * 10) {
confidence = 80; confidence = 80;
} else if (numValid > 3 && numInvalid == 0) { } else if (numValid > 3 && numInvalid == 0) {
confidence = 100; confidence = 100;
} else if (numValid > 0 && numInvalid == 0) { } else if (numValid > 0 && numInvalid == 0) {
confidence = 80; confidence = 80;
} else if (numValid > numInvalid * 10) { } else if (numValid > numInvalid * 10) {
// Probably corrupt UTF-32BE data. Valid sequences aren't likely by chance. // Probably corrupt UTF-32BE data. Valid sequences aren't likely by chance.
confidence = 25; confidence = 25;
} }
// return confidence == 0 ? null : new CharsetMatch(det, this, confidence); // return confidence == 0 ? null : new CharsetMatch(det, this, confidence);
return confidence == 0 ? null : new Match(det, this, confidence); return confidence == 0 ? null : new Match(det, this, confidence);
}; };
module.exports.UTF_32BE = function() { module.exports.UTF_32BE = function() {
this.name = function() { this.name = function() {
return 'UTF-32BE'; return 'UTF-32BE';
}; };
this.getChar = function(input, index) { this.getChar = function(input, index) {
return (input[index + 0] & 0xff) << 24 | (input[index + 1] & 0xff) << 16 | return (input[index + 0] & 0xff) << 24 | (input[index + 1] & 0xff) << 16 |
(input[index + 2] & 0xff) << 8 | (input[index + 3] & 0xff); (input[index + 2] & 0xff) << 8 | (input[index + 3] & 0xff);
}; };
}; };
util.inherits(module.exports.UTF_32BE, UTF_32); util.inherits(module.exports.UTF_32BE, UTF_32);
module.exports.UTF_32LE = function() { module.exports.UTF_32LE = function() {
this.name = function() { this.name = function() {
return 'UTF-32LE'; return 'UTF-32LE';
}; };
this.getChar = function(input, index) { this.getChar = function(input, index) {
return (input[index + 3] & 0xff) << 24 | (input[index + 2] & 0xff) << 16 | return (input[index + 3] & 0xff) << 24 | (input[index + 2] & 0xff) << 16 |
(input[index + 1] & 0xff) << 8 | (input[index + 0] & 0xff); (input[index + 1] & 0xff) << 8 | (input[index + 0] & 0xff);
}; };
}; };
util.inherits(module.exports.UTF_32LE, UTF_32); util.inherits(module.exports.UTF_32LE, UTF_32);

140
encoding/utf8.js
View File

@ -5,80 +5,80 @@ var Match = require ('../match');
* Charset recognizer for UTF-8 * Charset recognizer for UTF-8
*/ */
module.exports = function() { module.exports = function() {
this.name = function() { this.name = function() {
return 'UTF-8'; return 'UTF-8';
}; };
this.match = function(det) { this.match = function(det) {
var hasBOM = false, var hasBOM = false,
numValid = 0, numValid = 0,
numInvalid = 0, numInvalid = 0,
input = det.fRawInput, input = det.fRawInput,
trailBytes = 0, trailBytes = 0,
confidence; confidence;
if (det.fRawLength >= 3 && if (det.fRawLength >= 3 &&
(input[0] & 0xff) == 0xef && (input[1] & 0xff) == 0xbb && (input[2] & 0xff) == 0xbf) { (input[0] & 0xff) == 0xef && (input[1] & 0xff) == 0xbb && (input[2] & 0xff) == 0xbf) {
hasBOM = true; hasBOM = true;
}
// Scan for multi-byte sequences
for (var i = 0; i < det.fRawLength; i++) {
var b = input[i];
if ((b & 0x80) == 0)
continue; // ASCII
// Hi bit on char found. Figure out how long the sequence should be
if ((b & 0x0e0) == 0x0c0) {
trailBytes = 1;
} else if ((b & 0x0f0) == 0x0e0) {
trailBytes = 2;
} else if ((b & 0x0f8) == 0xf0) {
trailBytes = 3;
} else {
numInvalid++;
if (numInvalid > 5)
break;
trailBytes = 0;
}
// Verify that we've got the right number of trail bytes in the sequence
for (;;) {
i++;
if (i >= det.fRawLength)
break;
if ((input[i] & 0xc0) != 0x080) {
numInvalid++;
break;
} }
if (--trailBytes == 0) {
// Scan for multi-byte sequences numValid++;
for (var i = 0; i < det.fRawLength; i++) { break;
var b = input[i];
if ((b & 0x80) == 0)
continue; // ASCII
// Hi bit on char found. Figure out how long the sequence should be
if ((b & 0x0e0) == 0x0c0) {
trailBytes = 1;
} else if ((b & 0x0f0) == 0x0e0) {
trailBytes = 2;
} else if ((b & 0x0f8) == 0xf0) {
trailBytes = 3;
} else {
numInvalid++;
if (numInvalid > 5)
break;
trailBytes = 0;
}
// Verify that we've got the right number of trail bytes in the sequence
for (;;) {
i++;
if (i >= det.fRawLength)
break;
if ((input[i] & 0xc0) != 0x080) {
numInvalid++;
break;
}
if (--trailBytes == 0) {
numValid++;
break;
}
}
} }
}
}
// Cook up some sort of confidence score, based on presense of a BOM // Cook up some sort of confidence score, based on presense of a BOM
// and the existence of valid and/or invalid multi-byte sequences. // and the existence of valid and/or invalid multi-byte sequences.
confidence = 0; confidence = 0;
if (hasBOM && numInvalid == 0) if (hasBOM && numInvalid == 0)
confidence = 100; confidence = 100;
else if (hasBOM && numValid > numInvalid * 10) else if (hasBOM && numValid > numInvalid * 10)
confidence = 80; confidence = 80;
else if (numValid > 3 && numInvalid == 0) else if (numValid > 3 && numInvalid == 0)
confidence = 100; confidence = 100;
else if (numValid > 0 && numInvalid == 0) else if (numValid > 0 && numInvalid == 0)
confidence = 80; confidence = 80;
else if (numValid == 0 && numInvalid == 0) else if (numValid == 0 && numInvalid == 0)
// Plain ASCII. // Plain ASCII.
confidence = 10; confidence = 10;
else if (numValid > numInvalid * 10) else if (numValid > numInvalid * 10)
// Probably corruput utf-8 data. Valid sequences aren't likely by chance. // Probably corruput utf-8 data. Valid sequences aren't likely by chance.
confidence = 25; confidence = 25;
else else
return null return null
return new Match(det, this, confidence); return new Match(det, this, confidence);
}; };
}; };