Change new line format back to unix style

This commit is contained in:
mossroy 2014-02-28 17:18:42 +01:00
parent 8dd14095e6
commit 166c87118e
4 changed files with 1797 additions and 1797 deletions

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

View File

@ -1,126 +1,126 @@
/** /**
* titleIterators.js : Various classes to iterate over titles, for example as a * titleIterators.js : Various classes to iterate over titles, for example as a
* result of searching. * result of searching.
* *
* Copyright 2014 Evopedia developers * Copyright 2014 Evopedia developers
* License GPL v3: * License GPL v3:
* *
* This file is part of Evopedia. * This file is part of Evopedia.
* *
* Evopedia is free software: you can redistribute it and/or modify * Evopedia is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by * it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or * the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version. * (at your option) any later version.
* *
* Evopedia is distributed in the hope that it will be useful, * Evopedia is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of * but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details. * GNU General Public License for more details.
* *
* You should have received a copy of the GNU General Public License * You should have received a copy of the GNU General Public License
* along with Evopedia (file LICENSE-GPLv3.txt). If not, see <http://www.gnu.org/licenses/> * along with Evopedia (file LICENSE-GPLv3.txt). If not, see <http://www.gnu.org/licenses/>
*/ */
define(['utf8', 'title', 'util', 'jquery'], function(utf8, evopediaTitle, util, jQuery) { define(['utf8', 'title', 'util', 'jquery'], function(utf8, evopediaTitle, util, jQuery) {
// Maximum length of a title // Maximum length of a title
// 300 bytes is arbitrary : we actually do not really know how long the titles will be // 300 bytes is arbitrary : we actually do not really know how long the titles will be
// But mediawiki titles seem to be limited to ~200 bytes, so 300 should be more than enough // But mediawiki titles seem to be limited to ~200 bytes, so 300 should be more than enough
var MAX_TITLE_LENGTH = 300; var MAX_TITLE_LENGTH = 300;
/** /**
* Iterates over all titles starting at the given offset. * Iterates over all titles starting at the given offset.
* The asynchronous method advance has to be called before this.title is * The asynchronous method advance has to be called before this.title is
* valid. * valid.
* @param archive * @param archive
* @param offset * @param offset
*/ */
function SequentialTitleIterator(archive, offset) { function SequentialTitleIterator(archive, offset) {
this._titleFile = archive.titleFile; this._titleFile = archive.titleFile;
this._archive = archive; this._archive = archive;
this._offset = offset; this._offset = offset;
this.title = null; this.title = null;
}; };
/** /**
* Advances to the next title (or the first), if possible. * Advances to the next title (or the first), if possible.
* @returns jQuery promise containing the next title or null if there is no * @returns jQuery promise containing the next title or null if there is no
* next title * next title
*/ */
SequentialTitleIterator.prototype.advance = function() { SequentialTitleIterator.prototype.advance = function() {
if (this._offset >= this._titleFile.size) { if (this._offset >= this._titleFile.size) {
this.title = null; this.title = null;
return jQuery.when(this.title); return jQuery.when(this.title);
} }
var that = this; var that = this;
return util.readFileSlice(this._titleFile, this._offset, return util.readFileSlice(this._titleFile, this._offset,
this._offset + MAX_TITLE_LENGTH).then(function(byteArray) { this._offset + MAX_TITLE_LENGTH).then(function(byteArray) {
var newLineIndex = 15; var newLineIndex = 15;
while (newLineIndex < byteArray.length && byteArray[newLineIndex] !== 10) { while (newLineIndex < byteArray.length && byteArray[newLineIndex] !== 10) {
newLineIndex++; newLineIndex++;
} }
var encodedTitle = byteArray.subarray(0, newLineIndex); var encodedTitle = byteArray.subarray(0, newLineIndex);
that._title = evopediaTitle.Title.parseTitle(encodedTitle, that._archive, that._offset); that._title = evopediaTitle.Title.parseTitle(encodedTitle, that._archive, that._offset);
that._offset += newLineIndex + 1; that._offset += newLineIndex + 1;
return that._title; return that._title;
}); });
}; };
/** /**
* Searches for the offset into the given title file where the first title * Searches for the offset into the given title file where the first title
* with the given prefix (or lexicographically larger) is located. * with the given prefix (or lexicographically larger) is located.
* The given function normalize is applied to every title before comparison. * The given function normalize is applied to every title before comparison.
* @param titleFile * @param titleFile
* @param prefix * @param prefix
* @param normalize function to be applied to every title before comparison * @param normalize function to be applied to every title before comparison
* @returns jQuery promise giving the offset * @returns jQuery promise giving the offset
*/ */
function findPrefixOffset(titleFile, prefix, normalize) { function findPrefixOffset(titleFile, prefix, normalize) {
prefix = normalize(prefix); prefix = normalize(prefix);
var lo = 0; var lo = 0;
var hi = titleFile.size; var hi = titleFile.size;
var iterate = function() { var iterate = function() {
if (lo >= hi - 1) { if (lo >= hi - 1) {
if (lo > 0) if (lo > 0)
lo += 2; // Let lo point to the start of an entry lo += 2; // Let lo point to the start of an entry
return jQuery.when(lo); return jQuery.when(lo);
} else { } else {
var mid = Math.floor((lo + hi) / 2); var mid = Math.floor((lo + hi) / 2);
return util.readFileSlice(titleFile, mid, mid + MAX_TITLE_LENGTH).then(function(byteArray) { return util.readFileSlice(titleFile, mid, mid + MAX_TITLE_LENGTH).then(function(byteArray) {
// Look for the index of the next NewLine // Look for the index of the next NewLine
var newLineIndex = 0; var newLineIndex = 0;
while (newLineIndex < byteArray.length && byteArray[newLineIndex] !== 10) { while (newLineIndex < byteArray.length && byteArray[newLineIndex] !== 10) {
newLineIndex++; newLineIndex++;
} }
var startIndex = 0; var startIndex = 0;
if (mid > 0) { if (mid > 0) {
startIndex = newLineIndex + 16; startIndex = newLineIndex + 16;
newLineIndex = startIndex; newLineIndex = startIndex;
// Look for the index of the next NewLine // Look for the index of the next NewLine
while (newLineIndex < byteArray.length && byteArray[newLineIndex] !== 10) { while (newLineIndex < byteArray.length && byteArray[newLineIndex] !== 10) {
newLineIndex++; newLineIndex++;
} }
} }
if (newLineIndex === startIndex) { if (newLineIndex === startIndex) {
// End of file reached // End of file reached
hi = mid; hi = mid;
} else { } else {
var normalizedTitle = normalize(utf8.parse(byteArray.subarray(startIndex, newLineIndex))); var normalizedTitle = normalize(utf8.parse(byteArray.subarray(startIndex, newLineIndex)));
if (normalizedTitle < prefix) { if (normalizedTitle < prefix) {
lo = mid + newLineIndex - 1; lo = mid + newLineIndex - 1;
} else { } else {
hi = mid; hi = mid;
} }
} }
return iterate(); return iterate();
}); });
} }
}; };
return iterate(); return iterate();
} }
/** /**
* Functions and classes exposed by this module * Functions and classes exposed by this module
*/ */
return { return {
SequentialTitleIterator : SequentialTitleIterator, SequentialTitleIterator : SequentialTitleIterator,
findPrefixOffset : findPrefixOffset findPrefixOffset : findPrefixOffset
}; };
}); });

View File

@ -1,151 +1,151 @@
/** /**
* util.js : Utility functions * util.js : Utility functions
* *
* Copyright 2013 Mossroy * Copyright 2013 Mossroy
* License GPL v3: * License GPL v3:
* *
* This file is part of Evopedia. * This file is part of Evopedia.
* *
* Evopedia is free software: you can redistribute it and/or modify * Evopedia is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by * it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or * the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version. * (at your option) any later version.
* *
* Evopedia is distributed in the hope that it will be useful, * Evopedia is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of * but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details. * GNU General Public License for more details.
* *
* You should have received a copy of the GNU General Public License * You should have received a copy of the GNU General Public License
* along with Evopedia (file LICENSE-GPLv3.txt). If not, see <http://www.gnu.org/licenses/> * along with Evopedia (file LICENSE-GPLv3.txt). If not, see <http://www.gnu.org/licenses/>
*/ */
define(function(require) { define(function(require) {
var jQuery = require('jquery'); var jQuery = require('jquery');
/** /**
* Utility function : return true if the given string ends with the suffix * Utility function : return true if the given string ends with the suffix
* @param str * @param str
* @param suffix * @param suffix
* @returns {Boolean} * @returns {Boolean}
*/ */
function endsWith(str, suffix) { function endsWith(str, suffix) {
return str.indexOf(suffix, str.length - suffix.length) !== -1; return str.indexOf(suffix, str.length - suffix.length) !== -1;
} }
/** /**
* Read an integer encoded in 4 bytes, little endian * Read an integer encoded in 4 bytes, little endian
* @param {type} byteArray * @param {type} byteArray
* @param {type} firstIndex * @param {type} firstIndex
* @returns {Number} * @returns {Number}
*/ */
function readIntegerFrom4Bytes(byteArray, firstIndex) { function readIntegerFrom4Bytes(byteArray, firstIndex) {
var dataView = new DataView(byteArray.buffer, firstIndex, 4); var dataView = new DataView(byteArray.buffer, firstIndex, 4);
var integer = dataView.getUint32(0, true); var integer = dataView.getUint32(0, true);
return integer; return integer;
} }
/** /**
* Read an integer encoded in 2 bytes, little endian * Read an integer encoded in 2 bytes, little endian
* @param {type} byteArray * @param {type} byteArray
* @param {type} firstIndex * @param {type} firstIndex
* @returns {Number} * @returns {Number}
*/ */
function readIntegerFrom2Bytes(byteArray, firstIndex) { function readIntegerFrom2Bytes(byteArray, firstIndex) {
var dataView = new DataView(byteArray.buffer, firstIndex, 2); var dataView = new DataView(byteArray.buffer, firstIndex, 2);
var integer = dataView.getUint16(0, true); var integer = dataView.getUint16(0, true);
return integer; return integer;
} }
/** /**
* Read a float encoded in 2 bytes * Read a float encoded in 2 bytes
* @param {type} byteArray * @param {type} byteArray
* @param {type} firstIndex * @param {type} firstIndex
* @param {bool} littleEndian (optional) * @param {bool} littleEndian (optional)
* @returns {Number} * @returns {Number}
*/ */
function readFloatFrom4Bytes(byteArray, firstIndex, littleEndian) { function readFloatFrom4Bytes(byteArray, firstIndex, littleEndian) {
var dataView = new DataView(byteArray.buffer, firstIndex, 4); var dataView = new DataView(byteArray.buffer, firstIndex, 4);
var float = dataView.getFloat32(0, littleEndian); var float = dataView.getFloat32(0, littleEndian);
return float; return float;
} }
/** /**
* Convert a Uint8Array to a lowercase hex string * Convert a Uint8Array to a lowercase hex string
* @param {type} byteArray * @param {type} byteArray
* @returns {String} * @returns {String}
*/ */
function uint8ArrayToHex(byteArray) { function uint8ArrayToHex(byteArray) {
var s = ''; var s = '';
var hexDigits = '0123456789abcdef'; var hexDigits = '0123456789abcdef';
for (var i = 0; i < byteArray.length; i++) { for (var i = 0; i < byteArray.length; i++) {
var v = byteArray[i]; var v = byteArray[i];
s += hexDigits[(v & 0xff) >> 4]; s += hexDigits[(v & 0xff) >> 4];
s += hexDigits[v & 0xf]; s += hexDigits[v & 0xf];
} }
return s; return s;
} }
/** /**
* Convert a Uint8Array to base64 * Convert a Uint8Array to base64
* TODO : might be replaced by btoa() built-in function? https://developer.mozilla.org/en-US/docs/Web/API/window.btoa * TODO : might be replaced by btoa() built-in function? https://developer.mozilla.org/en-US/docs/Web/API/window.btoa
* @param {type} byteArray * @param {type} byteArray
* @returns {String} * @returns {String}
*/ */
function uint8ArrayToBase64(byteArray) { function uint8ArrayToBase64(byteArray) {
var b64 = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/="; var b64 = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=";
var bits, h1, h2, h3, h4, i = 0; var bits, h1, h2, h3, h4, i = 0;
var enc = ""; var enc = "";
for (var i = 0; i < byteArray.length; ) { for (var i = 0; i < byteArray.length; ) {
bits = byteArray[i++] << 16; bits = byteArray[i++] << 16;
bits |= byteArray[i++] << 8; bits |= byteArray[i++] << 8;
bits |= byteArray[i++]; bits |= byteArray[i++];
h1 = bits >> 18 & 0x3f; h1 = bits >> 18 & 0x3f;
h2 = bits >> 12 & 0x3f; h2 = bits >> 12 & 0x3f;
h3 = bits >> 6 & 0x3f; h3 = bits >> 6 & 0x3f;
h4 = bits & 0x3f; h4 = bits & 0x3f;
enc += b64[h1] + b64[h2] + b64[h3] + b64[h4]; enc += b64[h1] + b64[h2] + b64[h3] + b64[h4];
} }
var r = byteArray.length % 3; var r = byteArray.length % 3;
return (r > 0 ? enc.slice(0, r - 3) : enc) + '==='.slice(r || 3); return (r > 0 ? enc.slice(0, r - 3) : enc) + '==='.slice(r || 3);
} }
/** /**
* Reads a Uint8Array from the given file starting at byte offset begin and * Reads a Uint8Array from the given file starting at byte offset begin and
* not including byte offset end. * not including byte offset end.
* @param file * @param file
* @param begin * @param begin
* @param end * @param end
* @returns jQuery promise * @returns jQuery promise
*/ */
function readFileSlice(file, begin, end) { function readFileSlice(file, begin, end) {
var deferred = jQuery.Deferred(); var deferred = jQuery.Deferred();
var reader = new FileReader(); var reader = new FileReader();
reader.onload = function(e) { reader.onload = function(e) {
deferred.resolve(new Uint8Array(e.target.result)); deferred.resolve(new Uint8Array(e.target.result));
}; };
reader.onerror = reader.onabort = function(e) { reader.onerror = reader.onabort = function(e) {
deferred.reject(e); deferred.reject(e);
}; };
reader.readAsArrayBuffer(file.slice(begin, end)); reader.readAsArrayBuffer(file.slice(begin, end));
return deferred.promise(); return deferred.promise();
} }
/** /**
* Functions and classes exposed by this module * Functions and classes exposed by this module
*/ */
return { return {
endsWith: endsWith, endsWith: endsWith,
readIntegerFrom4Bytes: readIntegerFrom4Bytes, readIntegerFrom4Bytes: readIntegerFrom4Bytes,
readIntegerFrom2Bytes : readIntegerFrom2Bytes, readIntegerFrom2Bytes : readIntegerFrom2Bytes,
readFloatFrom4Bytes : readFloatFrom4Bytes, readFloatFrom4Bytes : readFloatFrom4Bytes,
uint8ArrayToHex : uint8ArrayToHex, uint8ArrayToHex : uint8ArrayToHex,
uint8ArrayToBase64 : uint8ArrayToBase64, uint8ArrayToBase64 : uint8ArrayToBase64,
readFileSlice : readFileSlice readFileSlice : readFileSlice
}; };
}); });