/**
* @return DOMDocument DOM to manipulate
*/
public function getDoc()
{
if (!$this->doc) {
// DOMDocument::loadHTML apparently isn't very good with encodings, so
// convert input to ASCII by encoding everything above 128 as entities.
if (function_exists('mb_convert_encoding')) {
$html = mb_convert_encoding($this->html, 'HTML-ENTITIES', 'UTF-8');
} else {
$html = preg_replace_callback('/[\\x{80}-\\x{10ffff}]/u', function ($m) {
return '&#' . UtfNormal\Utils::utf8ToCodepoint($m[0]) . ';';
}, $this->html);
}
// Workaround for bug that caused spaces before references
// to disappear during processing: https://phabricator.wikimedia.org/T55086
// TODO: Please replace with a better fix if one can be found.
$html = str_replace(' <', ' <', $html);
libxml_use_internal_errors(true);
$loader = libxml_disable_entity_loader();
$this->doc = new DOMDocument();
$this->doc->strictErrorChecking = false;
$this->doc->loadHTML($html);
libxml_disable_entity_loader($loader);
libxml_use_internal_errors(false);
$this->doc->encoding = 'UTF-8';
}
return $this->doc;
}
public function execute()
{
$hexPairs = array('0D23 0D4D 200D' => '0D7A', '0D28 0D4D 200D' => '0D7B', '0D30 0D4D 200D' => '0D7C', '0D32 0D4D 200D' => '0D7D', '0D33 0D4D 200D' => '0D7E', '0D15 0D4D 200D' => '0D7F');
$pairs = array();
foreach ($hexPairs as $hexSource => $hexDest) {
$source = UtfNormal\Utils::hexSequenceToUtf8($hexSource);
$dest = UtfNormal\Utils::hexSequenceToUtf8($hexDest);
$pairs[$source] = $dest;
}
global $IP;
file_put_contents("{$IP}/serialized/normalize-ml.ser", serialize($pairs));
echo "ml: " . count($pairs) . " pairs written.\n";
}
/**
* Get the first character of a string.
*
* @param string $s
* @return string
*/
function firstChar($s)
{
$matches = array();
preg_match('/^([\\x00-\\x7f]|[\\xc0-\\xdf][\\x80-\\xbf]|' . '[\\xe0-\\xef][\\x80-\\xbf]{2}|[\\xf0-\\xf7][\\x80-\\xbf]{3})/', $s, $matches);
if (isset($matches[1])) {
if (strlen($matches[1]) != 3) {
return $matches[1];
}
// Break down Hangul syllables to grab the first jamo
$code = UtfNormal\Utils::utf8ToCodepoint($matches[1]);
if ($code < 0xac00 || 0xd7a4 <= $code) {
return $matches[1];
} elseif ($code < 0xb098) {
return "ㄱ";
} elseif ($code < 0xb2e4) {
return "ㄴ";
} elseif ($code < 0xb77c) {
return "ㄷ";
} elseif ($code < 0xb9c8) {
return "ㄹ";
} elseif ($code < 0xbc14) {
return "ㅁ";
} elseif ($code < 0xc0ac) {
return "ㅂ";
} elseif ($code < 0xc544) {
return "ㅅ";
} elseif ($code < 0xc790) {
return "ㅇ";
} elseif ($code < 0xcc28) {
return "ㅈ";
} elseif ($code < 0xce74) {
return "ㅊ";
} elseif ($code < 0xd0c0) {
return "ㅋ";
} elseif ($code < 0xd30c) {
return "ㅌ";
} elseif ($code < 0xd558) {
return "ㅍ";
} else {
return "ㅎ";
}
} else {
return '';
}
}
function generateFirstChars()
{
$file = fopen("{$this->dataDir}/allkeys.txt", 'r');
if (!$file) {
$this->error("Unable to open allkeys.txt");
exit(1);
}
global $IP;
$outFile = fopen("{$IP}/serialized/first-letters-root.ser", 'w');
if (!$outFile) {
$this->error("Unable to open output file first-letters-root.ser");
exit(1);
}
$goodTertiaryChars = array();
// For each character with an entry in allkeys.txt, overwrite the implicit
// entry in $this->weights that came from the UCD.
// Also gather a list of tertiary weights, for use in selecting the group header
while (false !== ($line = fgets($file))) {
// We're only interested in single-character weights, pick them out with a regex
$line = trim($line);
if (!preg_match('/^([0-9A-F]+)\\s*;\\s*([^#]*)/', $line, $m)) {
continue;
}
$cp = hexdec($m[1]);
$allWeights = trim($m[2]);
$primary = '';
$tertiary = '';
if (!isset($this->weights[$cp])) {
// Non-printable, ignore
continue;
}
foreach (StringUtils::explode('[', $allWeights) as $weightStr) {
preg_match_all('/[*.]([0-9A-F]+)/', $weightStr, $m);
if (!empty($m[1])) {
if ($m[1][0] !== '0000') {
$primary .= '.' . $m[1][0];
}
if ($m[1][2] !== '0000') {
$tertiary .= '.' . $m[1][2];
}
}
}
$this->weights[$cp] = $primary;
if ($tertiary === '.0008' || $tertiary === '.000E') {
$goodTertiaryChars[$cp] = true;
}
}
fclose($file);
// Identify groups of characters with the same primary weight
$this->groups = array();
asort($this->weights, SORT_STRING);
$prevWeight = reset($this->weights);
$group = array();
foreach ($this->weights as $cp => $weight) {
if ($weight !== $prevWeight) {
$this->groups[$prevWeight] = $group;
$prevWeight = $weight;
if (isset($this->groups[$weight])) {
$group = $this->groups[$weight];
} else {
$group = array();
}
}
$group[] = $cp;
}
if ($group) {
$this->groups[$prevWeight] = $group;
}
// If one character has a given primary weight sequence, and a second
// character has a longer primary weight sequence with an initial
// portion equal to the first character, then remove the second
// character. This avoids having characters like U+A732 (double A)
// polluting the basic latin sort area.
foreach ($this->groups as $weight => $group) {
if (preg_match('/(\\.[0-9A-F]*)\\./', $weight, $m)) {
if (isset($this->groups[$m[1]])) {
unset($this->groups[$weight]);
}
}
}
ksort($this->groups, SORT_STRING);
// Identify the header character in each group
$headerChars = array();
$prevChar = "";
$tertiaryCollator = new Collator('root');
$primaryCollator = new Collator('root');
$primaryCollator->setStrength(Collator::PRIMARY);
$numOutOfOrder = 0;
foreach ($this->groups as $weight => $group) {
$uncomposedChars = array();
$goodChars = array();
foreach ($group as $cp) {
if (isset($goodTertiaryChars[$cp])) {
$goodChars[] = $cp;
}
if (!isset($this->mappedChars[$cp])) {
$uncomposedChars[] = $cp;
}
}
$x = array_intersect($goodChars, $uncomposedChars);
if (!$x) {
$x = $uncomposedChars;
if (!$x) {
$x = $group;
}
}
// Use ICU to pick the lowest sorting character in the selection
$tertiaryCollator->sort($x);
$cp = $x[0];
$char = UtfNormal\Utils::codepointToUtf8($cp);
$headerChars[] = $char;
if ($primaryCollator->compare($char, $prevChar) <= 0) {
$numOutOfOrder++;
/*
printf( "Out of order: U+%05X > U+%05X\n",
utf8ToCodepoint( $prevChar ),
utf8ToCodepoint( $char ) );
*/
}
$prevChar = $char;
if ($this->debugOutFile) {
fwrite($this->debugOutFile, sprintf("%05X %s %s (%s)\n", $cp, $weight, $char, implode(' ', array_map('UtfNormal\\Utils::codepointToUtf8', $group))));
}
}
print "Out of order: {$numOutOfOrder} / " . count($headerChars) . "\n";
fwrite($outFile, serialize($headerChars));
}
/**
* If the named entity is defined in the HTML 4.0/XHTML 1.0 DTD,
* return the UTF-8 encoding of that character. Otherwise, returns
* pseudo-entity source (eg "&foo;")
*
* @param string $name
* @return string
*/
static function decodeEntity($name)
{
if (isset(self::$htmlEntityAliases[$name])) {
$name = self::$htmlEntityAliases[$name];
}
if (isset(self::$htmlEntities[$name])) {
return UtfNormal\Utils::codepointToUtf8(self::$htmlEntities[$name]);
} else {
return "&{$name};";
}
}
public function execute()
{
if (!$this->hasOption('unicode-data-file')) {
$dataFile = 'UnicodeData.txt';
if (!file_exists($dataFile)) {
$this->error("Unable to find UnicodeData.txt. Please specify " . "its location with --unicode-data-file=<FILE>");
exit(1);
}
} else {
$dataFile = $this->getOption('unicode-data-file');
if (!file_exists($dataFile)) {
$this->error('Unable to find the specified data file.');
exit(1);
}
}
$file = fopen($dataFile, 'r');
if (!$file) {
$this->error('Unable to open the data file.');
exit(1);
}
// For the file format, see http://www.unicode.org/reports/tr44/
$fieldNames = array('Code', 'Name', 'General_Category', 'Canonical_Combining_Class', 'Bidi_Class', 'Decomposition_Type_Mapping', 'Numeric_Type_Value_6', 'Numeric_Type_Value_7', 'Numeric_Type_Value_8', 'Bidi_Mirrored', 'Unicode_1_Name', 'ISO_Comment', 'Simple_Uppercase_Mapping', 'Simple_Lowercase_Mapping', 'Simple_Titlecase_Mapping');
$pairs = array();
$lineNum = 0;
while (false !== ($line = fgets($file))) {
++$lineNum;
# Strip comments
$line = trim(substr($line, 0, strcspn($line, '#')));
if ($line === '') {
continue;
}
# Split fields
$numberedData = explode(';', $line);
$data = array();
foreach ($fieldNames as $number => $name) {
$data[$name] = $numberedData[$number];
}
$code = base_convert($data['Code'], 16, 10);
if ($code >= 0xfb50 && $code <= 0xfdff || $code >= 0xfe70 && $code <= 0xfeff) {
if ($data['Decomposition_Type_Mapping'] === '') {
// No decomposition
continue;
}
if (!preg_match('/^ *(<\\w*>) +([0-9A-F ]*)$/', $data['Decomposition_Type_Mapping'], $m)) {
$this->error("Can't parse Decomposition_Type/Mapping on line {$lineNum}");
$this->error($line);
continue;
}
$source = UtfNormal\Utils::hexSequenceToUtf8($data['Code']);
$dest = UtfNormal\Utils::hexSequenceToUtf8($m[2]);
$pairs[$source] = $dest;
}
}
global $IP;
file_put_contents("{$IP}/serialized/normalize-ar.ser", serialize($pairs));
echo "ar: " . count($pairs) . " pairs written.\n";
}
function getFirstLetter($string)
{
$string = strval($string);
if ($string === '') {
return '';
}
// Check for CJK
$firstChar = mb_substr($string, 0, 1, 'UTF-8');
if (ord($firstChar) > 0x7f && self::isCjk(UtfNormal\Utils::utf8ToCodepoint($firstChar))) {
return $firstChar;
}
$sortKey = $this->getPrimarySortKey($string);
// Do a binary search to find the correct letter to sort under
$min = ArrayUtils::findLowerBound(array($this, 'getSortKeyByLetterIndex'), $this->getFirstLetterCount(), 'strcmp', $sortKey);
if ($min === false) {
// Before the first letter
return '';
}
return $this->getLetterByIndex($min);
}
public function getFirstLetter($string)
{
$string = strval($string);
if ($string === '') {
return '';
}
$firstChar = mb_substr($string, 0, 1, 'UTF-8');
// If the first character is a CJK character, just return that character.
if (ord($firstChar) > 0x7f && self::isCjk(UtfNormal\Utils::utf8ToCodepoint($firstChar))) {
return $firstChar;
}
$sortKey = $this->getPrimarySortKey($string);
// Do a binary search to find the correct letter to sort under
$min = ArrayUtils::findLowerBound([$this, 'getSortKeyByLetterIndex'], $this->getFirstLetterCount(), 'strcmp', $sortKey);
if ($min === false) {
// Before the first letter
return '';
}
$sortLetter = $this->getLetterByIndex($min);
if ($this->useNumericCollation) {
// If the sort letter is a number, return '0–9' (or localized equivalent).
// ASCII value of 0 is 48. ASCII value of 9 is 57.
// Note that this also applies to non-Arabic numerals since they are
// mapped to Arabic numeral sort letters. For example, ২ sorts as 2.
if (ord($sortLetter) >= 48 && ord($sortLetter) <= 57) {
$sortLetter = wfMessage('category-header-numerals')->numParams(0, 9)->text();
}
}
return $sortLetter;
}
/**
* Reverse the previously applied transliteration of non-ASCII characters
* back to UTF-8. Used to protect data from corruption by broken web browsers
* as listed in $wgBrowserBlackList.
*
* @param string $invalue
* @return string
*/
private function unmakeSafe($invalue)
{
$result = "";
$valueLength = strlen($invalue);
for ($i = 0; $i < $valueLength; $i++) {
if (substr($invalue, $i, 3) == "&#x" && $invalue[$i + 3] != '0') {
$i += 3;
$hexstring = "";
do {
$hexstring .= $invalue[$i];
$i++;
} while (ctype_xdigit($invalue[$i]) && $i < strlen($invalue));
// Do some sanity checks. These aren't needed for reversibility,
// but should help keep the breakage down if the editor
// breaks one of the entities whilst editing.
if (substr($invalue, $i, 1) == ";" && strlen($hexstring) <= 6) {
$codepoint = hexdec($hexstring);
$result .= UtfNormal\Utils::codepointToUtf8($codepoint);
} else {
$result .= "&#x" . $hexstring . substr($invalue, $i, 1);
}
} else {
$result .= substr($invalue, $i, 1);
}
}
// reverse the transform that we made for reversibility reasons.
return strtr($result, array("�" => "&#x"));
}
/**
* Mangle XML-invalid names to be valid in XML
* @param string $name
* @param array $preserveKeys Names to not mangle
* @return string Mangled name
*/
private static function mangleName($name, $preserveKeys = [])
{
static $nsc = null, $nc = null;
if (in_array($name, $preserveKeys, true)) {
return $name;
}
if ($name === '') {
return '_';
}
if ($nsc === null) {
// Note we omit ':' from $nsc and $nc because it's reserved for XML
// namespacing, and we omit '_' from $nsc (but not $nc) because we
// reserve it.
$nsc = 'A-Za-z\\x{C0}-\\x{D6}\\x{D8}-\\x{F6}\\x{F8}-\\x{2FF}\\x{370}-\\x{37D}\\x{37F}-\\x{1FFF}' . '\\x{200C}-\\x{200D}\\x{2070}-\\x{218F}\\x{2C00}-\\x{2FEF}\\x{3001}-\\x{D7FF}' . '\\x{F900}-\\x{FDCF}\\x{FDF0}-\\x{FFFD}\\x{10000}-\\x{EFFFF}';
$nc = $nsc . '_\\-.0-9\\x{B7}\\x{300}-\\x{36F}\\x{203F}-\\x{2040}';
}
if (preg_match("/^[{$nsc}][{$nc}]*\$/uS", $name)) {
return $name;
}
return '_' . preg_replace_callback("/[^{$nc}]/uS", function ($m) {
return sprintf('.%X.', UtfNormal\Utils::utf8ToCodepoint($m[0]));
}, str_replace('.', '.2E.', $name));
}
public function execute()
{
if (!$this->hasOption('unicode-data-file')) {
$dataFile = 'UnicodeData.txt';
if (!file_exists($dataFile)) {
$this->error("Unable to find UnicodeData.txt. Please specify " . "its location with --unicode-data-file=<FILE>");
exit(1);
}
} else {
$dataFile = $this->getOption('unicode-data-file');
if (!file_exists($dataFile)) {
$this->error('Unable to find the specified data file.');
exit(1);
}
}
$file = fopen($dataFile, 'r');
if (!$file) {
$this->error('Unable to open the data file.');
exit(1);
}
// For the file format, see http://www.unicode.org/reports/tr44/
$fieldNames = array('Code', 'Name', 'General_Category', 'Canonical_Combining_Class', 'Bidi_Class', 'Decomposition_Type_Mapping', 'Numeric_Type_Value_6', 'Numeric_Type_Value_7', 'Numeric_Type_Value_8', 'Bidi_Mirrored', 'Unicode_1_Name', 'ISO_Comment', 'Simple_Uppercase_Mapping', 'Simple_Lowercase_Mapping', 'Simple_Titlecase_Mapping');
$upper = array();
$lower = array();
$lineNum = 0;
while (false !== ($line = fgets($file))) {
++$lineNum;
# Strip comments
$line = trim(substr($line, 0, strcspn($line, '#')));
if ($line === '') {
continue;
}
# Split fields
$numberedData = explode(';', $line);
$data = array();
foreach ($fieldNames as $number => $name) {
$data[$name] = $numberedData[$number];
}
$source = UtfNormal\Utils::hexSequenceToUtf8($data['Code']);
if ($data['Simple_Uppercase_Mapping']) {
$upper[$source] = UtfNormal\Utils::hexSequenceToUtf8($data['Simple_Uppercase_Mapping']);
}
if ($data['Simple_Lowercase_Mapping']) {
$lower[$source] = UtfNormal\Utils::hexSequenceToUtf8($data['Simple_Lowercase_Mapping']);
}
}
global $IP;
file_put_contents("{$IP}/serialized/Utf8Case.ser", serialize(array('wikiUpperChars' => $upper, 'wikiLowerChars' => $lower)));
}
