$verseLocation = substr($location, 0, strlen($location) - 2);
//$segmentWord = removeTashkeel($segmentWord);
if ($POS == "DET") {
// second segment PoS
$segmentPoStag = $qacMasterTableEntry[$segmentId]['TAG'];
//number of segments
$numberOfSegmentsInWord = count($qacMasterTableEntry);
if (($segmentPoStag == "N" || $segmentPoStag == "ADJ") && $numberOfSegmentsInWord == 2) {
continue;
}
}
// get word index in verse
$wordIndex = getWordIndexFromQACLocation($location);
//echoN($segmentFormARimla2y);
// get simple version of the word index
$imla2yWordIndex = getImla2yWordIndexByUthmaniLocation($location);
// get verse text
$verseText = getVerseByQACLocation($QURAN_TEXT, $location);
$imla2yWord = getWordFromVerseByIndex($PAUSEMARKS, $verseText, $imla2yWordIndex);
//echoN($imla2yWord);
$stopWordsFromQuran[$imla2yWord] = 1;
echoN($imla2yWord);
}
}
$ya = "يا";
//add يا
$stopWordsFromQuran[$ya] = 1;
echoN(count($stopWordsFromQuran));
//preprint_r($stopWordsFromQuran);
//exit;
file_put_contents(dirname(__FILE__) . "/../data/quran-stop-words.strict.l2.ar", implode("\n", array_keys($stopWordsFromQuran)));
/** Returns words from QAC by PoS tags - grouped by lemma **/
function getWordsByPos(&$finalTerms, $POS)
{
global $LEMMA_TO_SIMPLE_WORD_MAP;
$qacPosEntryArr = getModelEntryFromMemory("AR", "MODEL_QAC", "QAC_POS", $POS);
$QURAN_TEXT = getModelEntryFromMemory("AR", "MODEL_CORE", "QURAN_TEXT", "");
$TOTALS = getModelEntryFromMemory("AR", "MODEL_CORE", "TOTALS", "");
$PAUSEMARKS = $TOTALS['PAUSEMARKS'];
$WORDS_FREQUENCY = getModelEntryFromMemory("AR", "MODEL_CORE", "WORDS_FREQUENCY", "");
// Get all segment in QAC for that PoS
foreach ($qacPosEntryArr as $location => $segmentId) {
$qacMasterTableEntry = getModelEntryFromMemory("AR", "MODEL_QAC", "QAC_MASTERTABLE", $location);
// get Word, Lema and root
$segmentWord = $qacMasterTableEntry[$segmentId - 1]['FORM_AR'];
$segmentWordLema = $qacMasterTableEntry[$segmentId - 1]['FEATURES']['LEM'];
$segmentWordRoot = $qacMasterTableEntry[$segmentId - 1]['FEATURES']['ROOT'];
$verseLocation = substr($location, 0, strlen($location) - 2);
//$segmentWord = removeTashkeel($segmentWord);
// get word index in verse
$wordIndex = getWordIndexFromQACLocation($location);
//$segmentFormARimla2y = $UTHMANI_TO_SIMPLE_WORD_MAP_AND_VS[$segmentWord];
// get simple version of the word index
$imla2yWordIndex = getImla2yWordIndexByUthmaniLocation($location);
// get verse text
$verseText = getVerseByQACLocation($QURAN_TEXT, $location);
//echoN("|$segmentWord|$imla2yWord");
$segmentWordNoTashkeel = removeTashkeel($segmentWordLema);
$superscriptAlef = json_decode('"\\u0670"');
$alefWasla = "ٱ";
//U+0671
//$imla2yWord = $LEMMA_TO_SIMPLE_WORD_MAP[$segmentWordLema];
// this block is important since $LEMMA_TO_SIMPLE_WORD_MAP is not good for non $superscriptAlef words
// ex زيت lemma is converted to زيتها which spoiled the ontology concept list results
if (mb_strpos($segmentWordLema, $superscriptAlef) !== false || mb_strpos($segmentWordLema, $alefWasla) !== false) {
$imla2yWord = getItemFromUthmaniToSimpleMappingTable($segmentWordLema);
if (empty($imla2yWord)) {
$imla2yWord = $LEMMA_TO_SIMPLE_WORD_MAP[$segmentWordLema];
}
} else {
$imla2yWord = getItemFromUthmaniToSimpleMappingTable($segmentWordLema);
if (empty($imla2yWord)) {
$imla2yWord = shallowUthmaniToSimpleConversion($segmentWordLema);
//$segmentWordNoTashkeel;
}
}
/// in case the word was not found after removing tashkeel, try the lema mappign table
$termWeightArr = $MODEL_CORE['WORDS_FREQUENCY']['WORDS_TFIDF'][$imla2yWord];
// NOT WORKING BECAUSE LEMMAS WILL NOT BE IN SIMPLE WORDS LIST و الصابيئن =>صَّٰبِـِٔين
// if the word after removing tashkeel is not found in quran simple words list, then try lemma table
/*if (!isset($MODEL_CORE['WORDS_FREQUENCY']['WORDS'][$imla2yWord]) )
{
$imla2yWord = $LEMMA_TO_SIMPLE_WORD_MAP[$segmentWordLema];
if ( empty($imla2yWord) )
{
echoN($segmentWordLema);
echoN($imla2yWord);
preprint_r($LEMMA_TO_SIMPLE_WORD_MAP);
preprint_r($MODEL_CORE['WORDS_FREQUENCY']['WORDS']);
exit;
}
}*/
if (empty($termWeightArr)) {
//only for weight since the lema table decrease qurana matching
$imla2yWordForWeight = $LEMMA_TO_SIMPLE_WORD_MAP[$segmentWordLema];
$termWeightArr = $WORDS_FREQUENCY['WORDS_TFIDF'][$imla2yWordForWeight];
}
$termWeight = $termWeightArr['TFIDF'];
////////////////////////////////////////////
$termWord = $segmentWordLema;
//$imla2yWord;//"|$segmentWord| ".$imla2yWord ." - $location:$segmentId - $wordIndex=$imla2yWordIndex";
if (!isset($finalTerms[$termWord])) {
$finalTerms[$termWord] = generateEmptyConceptMetadata();
$finalTerms[$termWord]['LEM'] = $segmentWordLema;
$finalTerms[$termWord]['POS'] = $POS;
$finalTerms[$termWord]['SIMPLE_WORD'] = $imla2yWord;
$finalTerms[$termWord]['ROOT'] = $segmentWordRoot;
$finalTerms[$termWord]['WEIGHT'] = $termWeight;
}
$finalTerms[$termWord]["FREQ"] = $finalTerms[$termWord]["FREQ"] + 1;
if (!isset($finalTerms[$termWord]["SEG"][$segmentWord])) {
$finalTerms[$termWord]["SEG"][$segmentWord] = $imla2yWord;
}
if (!isset($finalTerms[$termWord]["POSES"][$POS])) {
$finalTerms[$termWord]["POSES"][$POS] = 1;
}
}
return $finalTerms;
}
