function getWordInfo($word, $MODEL_CORE, $MODEL_SEARCH, $MODEL_QAC, $fast = FALSE, $exactWord = FALSE)
{
$wordInfoArr = array();
$word = trim($word);
$wordUthmani = "";
$wordSimple = "";
if (isSimpleQuranWord($word)) {
$wordUthmani = getItemFromUthmaniToSimpleMappingTable($word);
$wordSimple = $word;
} else {
$wordUthmani = $word;
//preprint_r($UTHMANI_TO_SIMPLE_WORD_MAP_VS);
// tashkeel of last char is significant, ex: lemmas will probably not be in the MAP because of that
$wordSimple = getItemFromUthmaniToSimpleMappingTable($wordUthmani);
}
$WORDS_FREQUENCY = getModelEntryFromMemory("AR", "MODEL_CORE", "WORDS_FREQUENCY", "");
$freqArr = $WORDS_FREQUENCY['WORDS_TFIDF'][$wordSimple];
//preprint_r($freqArr);
$wordInfoArr['WORD_SIMPLE'] = $wordSimple;
$wordInfoArr['WORD_UTHMANI'] = $wordUthmani;
/*echoN("Simple:".$wordSimple);
echoN("Uthmani:".$wordUthmani);
echoN("Repetition:".$freqArr['TF']);
echoN("TF-IDF Weight:".round($freqArr['TFIDF'],2));
*/
$wordInfoArr['TF'] = $freqArr['TF'];
$wordInfoArr['TFIDF'] = round($freqArr['TFIDF'], 2);
//preprint_r($MODEL_QAC['QAC_MASTERTABLE']);
//preprint_r(array_keys($MODEL_QAC['QAC_FEATURES']));
$buckwalterTransliteration = "";
$posTagsArr = array();
$lemmasArr = array();
$wordRoot = "";
$featuresArr = array();
$versesArr = array();
$versesTagsArr = array();
$buckwalterTransliteration = "";
$wordRoot = "";
if (empty($wordSimple) || !modelEntryExistsInMemory("AR", "MODEL_SEARCH", "INVERTED_INDEX", $wordSimple)) {
return null;
}
//preprint_r($MODEL_SEARCH['INVERTED_INDEX'][$wordSimple]);
$invertedIndexEntry = getModelEntryFromMemory("AR", "MODEL_SEARCH", "INVERTED_INDEX", $wordSimple);
$QURAN_TEXT = getModelEntryFromMemory("AR", "MODEL_CORE", "QURAN_TEXT", "");
$TOTALS = getModelEntryFromMemory("AR", "MODEL_CORE", "TOTALS", "");
$PAUSEMARKS = $TOTALS['PAUSEMARKS'];
foreach ($invertedIndexEntry as $documentArrInIndex) {
$SURA = $documentArrInIndex['SURA'];
$AYA = $documentArrInIndex['AYA'];
$INDEX_IN_AYA_EMLA2Y = $documentArrInIndex['INDEX_IN_AYA_EMLA2Y'];
$INDEX_IN_AYA_UTHMANI = $documentArrInIndex['INDEX_IN_AYA_UTHMANI'];
$WORD_TYPE = $documentArrInIndex['WORD_TYPE'];
$EXTRA_WORD_TYPE_INFO = $documentArrInIndex['EXTRA_INFO'];
// INGORE ROOT SOURCES AND PRONOUNS, WE ONLY NEED THE NROMAL CORRESPONDING WORD
if ($WORD_TYPE == "PRONOUN_ANTECEDENT" || $WORD_TYPE == "ROOT") {
continue;
}
$qacLocation = getQACLocationStr($SURA + 1, $AYA + 1, $INDEX_IN_AYA_UTHMANI);
//echoN($qacLocation);exit;;
//preprint_r($MODEL_QAC['QAC_MASTERTABLE'][$qacLocation]);
//exit;
$qacMasterTableEntryArr = getModelEntryFromMemory("AR", "MODEL_QAC", "QAC_MASTERTABLE", $qacLocation);
// search QAC for roots and LEMMAS for this word
foreach ($qacMasterTableEntryArr as $segmentIndex => $segmentDataArr) {
$tag = $segmentDataArr['TAG'];
$segmentWord = $segmentDataArr['FORM_AR'];
//echoN($segmentWord);
//preprint_r($segmentDataArr);
$segmentWordSimple = "";
$simpleRepresentation = getItemFromUthmaniToSimpleMappingTable($segmentWord);
// !empty() produced error = Can't use function return value in write context
if (strlen(trim($simpleRepresentation)) > 0) {
$segmentWordSimple = $simpleRepresentation;
}
$buckwalterTransliteration = $segmentDataArr['FORM_EN'];
if (isset($segmentDataArr['FEATURES']['LEM'])) {
$lemma = $segmentDataArr['FEATURES']['LEM'];
}
$featuresArr = array_merge($segmentDataArr['FEATURES']);
$verseText = getVerseByQACLocation($QURAN_TEXT, $qacLocation);
$wordId = getWordIndexFromQACLocation($qacLocation);
if ($exactWord == TRUE) {
$wordFromVerseAtLocation = getWordFromVerseByIndex($PAUSEMARKS, $verseText, $wordId);
if ($wordSimple !== $wordFromVerseAtLocation) {
continue;
}
}
//echoN("$segmentWord|$tag");
//for segments like ุงู no corresponding simple words to compare, not our target segment, so continue
//if ( empty($segmentWordSimple)) continue;
if (isset($segmentDataArr['FEATURES']['ROOT']) && $segmentDataArr['FEATURES']['ROOT'] != -1) {
$wordRoot = $segmentDataArr['FEATURES']['ROOT'];
}
$posTagsArr[$tag] = 1;
$lemmasArr[$lemma] = 1;
//echoN("|$segmentWordSimple|$wordSimple|$segmentWord");
//$verseText = markSpecificWordInText($verseText,$wordId,$segmentWordSimple,"marked_fg");
$qacVerseLocation = substr($qacLocation, 0, strrpos($qacLocation, ":"));
if (!isset($versesArr[$qacVerseLocation])) {
$versesArr[$qacVerseLocation] = $verseText;
}
if (!isset($versesTagsArr[$qacVerseLocation])) {
$versesTagsArr[$qacVerseLocation] = "";
}
$versesTagsArr[$qacVerseLocation] = $versesTagsArr[$qacVerseLocation] . " " . $tag;
}
// we don't need all inverted index list except for verses, only break if we found at least one word
if ($fast == true && !empty($versesArr)) {
break;
}
}
$wordInfoArr['BUCKWALTER'] = $buckwalterTransliteration;
$wordInfoArr['ROOT'] = $wordRoot;
$wordInfoArr['LEM'] = $lemmasArr;
$wordInfoArr['POS'] = $posTagsArr;
$wordInfoArr['VERSES'] = $versesArr;
$wordInfoArr['VERSES_POS_TAGS'] = $versesTagsArr;
$wordInfoArr['FEATURES'] = $featuresArr;
return $wordInfoArr;
}
function getScoredDocumentsFromInveretdIndex($extendedQueryWordsArr, $query, $isPhraseSearch, $isQuestion, $isColumnSearch, $columnSearchKeyValParams, $isConceptSearch, $lang, $isTransliterationSearch)
{
global $MODEL_CORE, $MODEL_SEARCH;
$QURAN_TEXT = getModelEntryFromMemory($lang, "MODEL_CORE", "QURAN_TEXT", "");
$TOTALS = getModelEntryFromMemory($lang, "MODEL_CORE", "TOTALS", "");
$PAUSEMARKS = $TOTALS['PAUSEMARKS'];
if ($isColumnSearch) {
$SURA = $columnSearchKeyValParams['KEY'] - 1;
$isFullChapter = $columnSearchKeyValParams['VAL'] == "ALL";
if ($isFullChapter) {
$suraSize = count($QURAN_TEXT[$SURA]);
for ($AYA = 0; $AYA < $suraSize; $AYA++) {
$scoringTable[$SURA . ":" . $AYA] = array();
$scoringTable[$SURA . ":" . $AYA]['SCORE'] = 1;
$scoringTable[$SURA . ":" . $AYA]['SURA'] = $SURA;
$scoringTable[$SURA . ":" . $AYA]['AYA'] = $AYA;
}
} else {
$AYA = $columnSearchKeyValParams['VAL'] - 1;
// VERSE VALIDITY CHECK
$qacLocation = getQACLocationStr($SURA + 1, $AYA + 1, 0);
$verseText = getVerseByQACLocation($QURAN_TEXT, $qacLocation);
if (empty($verseText)) {
return array();
}
//////////////////////
$scoringTable[$SURA . ":" . $AYA] = array();
$scoringTable[$SURA . ":" . $AYA]['SCORE'] = 1;
$scoringTable[$SURA . ":" . $AYA]['SURA'] = $SURA;
$scoringTable[$SURA . ":" . $AYA]['AYA'] = $AYA;
}
return $scoringTable;
}
// $MODEL_QURANA = apc_fetch("MODEL_QURANA");
//preprint_r($extendedQueryWordsArr);
//$isOneWordQuery = preg_match("/ /", $query)==0;
//preprint_r($extendedQueryWordsArr);
/**
* GET ALL RESULT FORM INDEX USING EXTENDED QUERY WORD (WHICH INCLUDES ALL VARIATIONS AND PRONOUNS)
*/
foreach ($extendedQueryWordsArr as $word => $targetQACLocation) {
//echoN("|$word|");
//echoN($lang);
//echoN($isConceptSearch);
/*if ($lang=="EN" && $isConceptSearch )
{
}*/
//preprint_r($MODEL_SEARCH['INVERTED_INDEX'][$word]);
$invertedIndexEntry = getModelEntryFromMemory($lang, "MODEL_SEARCH", "INVERTED_INDEX", $word);
foreach ($invertedIndexEntry as $documentArrInIndex) {
//echoN("$word");
//preprint_r($documentArrInIndex);;
$SURA = $documentArrInIndex['SURA'];
$AYA = $documentArrInIndex['AYA'];
$INDEX_IN_AYA_EMLA2Y = $documentArrInIndex['INDEX_IN_AYA_EMLA2Y'];
$INDEX_IN_AYA_UTHMANI = $documentArrInIndex['INDEX_IN_AYA_UTHMANI'];
$WORD_TYPE = $documentArrInIndex['WORD_TYPE'];
$EXTRA_INFO = $documentArrInIndex['EXTRA_INFO'];
//echo getQACLocationStr($SURA,$AYA,$INDEX_IN_AYA_EMLA2Y);
$qacLocation = getQACLocationStr($SURA + 1, $AYA + 1, $INDEX_IN_AYA_UTHMANI);
$verseText = getVerseByQACLocation($QURAN_TEXT, $qacLocation);
/*
*
* NOTE: A DECISION SHOULD BE TAKEN TO SERACH AROUND AND REMOVE PAUSE MARKS OR NOT
*/
if ($lang == "AR") {
$verseTextWithoutPauseMarks = removePauseMarkFromVerse($verseText);
} else {
$verseTextWithoutPauseMarks = removeSpecialCharactersFromMidQuery($verseText);
}
//echoN("|$query|$verseTextWithoutPauseMarks");
$fullQueryIsFoundInVerseCount = preg_match_all("/(^|[ ]){$query}([ ]|\$)/umi", $verseTextWithoutPauseMarks);
//echoN("$query | $word");
if ($isPhraseSearch && $WORD_TYPE != "PRONOUN_ANTECEDENT") {
$numberOfOccurencesForWord = $fullQueryIsFoundInVerseCount;
if ($numberOfOccurencesForWord == 0) {
continue;
}
} else {
$numberOfOccurencesForWord = preg_match_all("/{$word}/um", $verseText);
/*if ( $numberOfOccurencesForWord> 100)
{
echoN($word);
echoN($verseText);
preprint_r($extendedQueryWordsArr);
exit;
}*/
}
//echoN($numberOfOccurencesForWord);
//echoN("$qacLocation|$targetQACLocation|$word|$EXTRA_INFO|$WORD_TYPE");
// incase of non normal word ( QAC/QURANA) .. translate WordIndex from Uthmani script to Imla2y script
/*if ( $WORD_TYPE!="NORMAL_WORD" )
{
//echoN("OLD:$INDEX_IN_AYA_EMLA2Y");
$INDEX_IN_AYA_EMLA2Y = getImla2yWordIndexByUthmaniLocation($qacLocation,$UTHMANI_TO_SIMPLE_LOCATION_MAP);
//echoN("NEW:$INDEX_IN_AYA_EMLA2Y");
}*/
//echoN($word);
//preprint_r($documentArrInIndex);
//preprint_r($MODEL_QAC['QAC_MASTERTABLE'][$qacLocation]);
if (!isset($scoringTable[$SURA . ":" . $AYA])) {
$scoringTable[$SURA . ":" . $AYA] = array();
$scoringTable[$SURA . ":" . $AYA]['SCORE'] = 0;
$scoringTable[$SURA . ":" . $AYA]['FREQ'] = 0;
$scoringTable[$SURA . ":" . $AYA]['DISTANCE'] = 0;
$scoringTable[$SURA . ":" . $AYA]['WORD_OCCURENCES_COUNT'] = 0;
$scoringTable[$SURA . ":" . $AYA]['QUERY_WORDS_IN_VERSE'] = 0;
$scoringTable[$SURA . ":" . $AYA]['IS_FULL_QUERY_IN_VERSE'] = 0;
$scoringTable[$SURA . ":" . $AYA]['SURA'] = $SURA;
$scoringTable[$SURA . ":" . $AYA]['AYA'] = $AYA;
$scoringTable[$SURA . ":" . $AYA]['POSSIBLE_HIGHLIGHTABLE_WORDS'] = array();
$scoringTable[$SURA . ":" . $AYA]['WORD_TYPE'] = $WORD_TYPE;
$scoringTable[$SURA . ":" . $AYA]['EXTRA_INFO'] = $EXTRA_INFO;
$scoringTable[$SURA . ":" . $AYA]['INDEX_IN_AYA_EMLA2Y'] = $INDEX_IN_AYA_EMLA2Y;
$scoringTable[$SURA . ":" . $AYA]['INDEX_IN_AYA_UTHMANI'] = $INDEX_IN_AYA_UTHMANI;
$scoringTable[$SURA . ":" . $AYA]['PRONOUNS'] = array();
}
$scoringTable[$SURA . ":" . $AYA]['WORD_OCCURENCES_COUNT'] = $numberOfOccurencesForWord;
//echoN($numberOfOccurencesForWord);
if (!isset($scoringTable[$SURA . ":" . $AYA]['POSSIBLE_HIGHLIGHTABLE_WORDS'][$word]) && $numberOfOccurencesForWord > 0 && $scoringTable[$SURA . ":" . $AYA]['FREQ'] > 0) {
//TODO: seems duplicate of WORD_OCCURENCES_COUNT
// Raise the frequency (score) of ayas containing more than one of the query items
$scoringTable[$SURA . ":" . $AYA]['FREQ']++;
//=$numberOfOccurencesForWord;
} else {
$scoringTable[$SURA . ":" . $AYA]['FREQ']++;
}
/*$verseArr = preg_split("/ /",$MODEL_CORE['QURAN_TEXT'][$SURA][$AYA]);
$verseArr = removePauseMarksFromArr($MODEL_CORE['TOTALS']['PAUSEMARKS'],$verseArr);
$simpleWordFromText = $verseArr[$INDEX_IN_AYA_EMLA2Y-1];
*/
/*
if ( empty($simpleWordFromText))
{
echoN($INDEX_IN_AYA_EMLA2Y);
preprint_r($verseArr);
}
echoN($qacLocation);
echoN($word);
echoN($INDEX_IN_AYA_EMLA2Y);
echoN($MODEL_CORE['QURAN_TEXT'][$SURA][$AYA]);
echoN($simpleWordFromText);
preprint_r($verseArr);
*/
// STEM or PRONOUN
if ($WORD_TYPE == "PRONOUN_ANTECEDENT") {
$scoringTable[$SURA . ":" . $AYA]['PRONOUNS'][$EXTRA_INFO] = $INDEX_IN_AYA_EMLA2Y;
} else {
if ($WORD_TYPE == "ROOT" || $WORD_TYPE == "LEM") {
// for non-normal words this will get the whole segment
$scoringTable[$SURA . ":" . $AYA]['POSSIBLE_HIGHLIGHTABLE_WORDS'][$word] = $WORD_TYPE;
// needed to fix root that are sometimes converted by uthmani/simple map below
$scoringTable[$SURA . ":" . $AYA]['POSSIBLE_HIGHLIGHTABLE_WORDS'][removeTashkeel($EXTRA_INFO)] = $WORD_TYPE;
// try to convert QAC uthmani word to simpleimla2y using the MAP table with and withou tashkeel
$wordInAya = getItemFromUthmaniToSimpleMappingTable($EXTRA_INFO);
if (empty($wordInAya)) {
$wordInAya = getItemFromUthmaniToSimpleMappingTable(removeTashkeel($EXTRA_INFO));
}
if (empty($wordInAya)) {
$wordInAya = removeTashkeel($EXTRA_INFO);
}
/*if ( empty($wordInAya ) )
{
preprint_r($documentArrInIndex);
echoN($EXTRA_INFO);
echo"HERE";
preprint_r($scoringTable[$SURA.":".$AYA]);exit;
}*/
//echoN("$word-$wordInAya-$EXTRA_INFO");
$scoringTable[$SURA . ":" . $AYA]['POSSIBLE_HIGHLIGHTABLE_WORDS'][$wordInAya] = $WORD_TYPE;
} else {
if ($isTransliterationSearch) {
$scoringTable[$SURA . ":" . $AYA]['POSSIBLE_HIGHLIGHTABLE_WORDS'][$word] = $WORD_TYPE;
} else {
// word was in original user query, not in our extended one
///if ( in_array($word,$queryWordsArr))
//{
$scoringTable[$SURA . ":" . $AYA]['POSSIBLE_HIGHLIGHTABLE_WORDS'][$word] = $WORD_TYPE;
}
//}
}
}
$scoringTable[$SURA . ":" . $AYA]['IS_FULL_QUERY_IN_VERSE'] = $fullQueryIsFoundInVerseCount;
$scoringTable[$SURA . ":" . $AYA]['QUERY_WORDS_IN_VERSE'] = count($scoringTable[$SURA . ":" . $AYA]['POSSIBLE_HIGHLIGHTABLE_WORDS']);
$scoringTable[$SURA . ":" . $AYA]['SCORE'] = $scoringTable[$SURA . ":" . $AYA]['FREQ'] / 2 + $scoringTable[$SURA . ":" . $AYA]['DISTANCE'] * 1 + $scoringTable[$SURA . ":" . $AYA]['QUERY_WORDS_IN_VERSE'] * 10 + count($scoringTable[$SURA . ":" . $AYA]['PRONOUNS']) * 1 + $scoringTable[$SURA . ":" . $AYA]['WORD_OCCURENCES_COUNT'] * 1 + $scoringTable[$SURA . ":" . $AYA]['IS_FULL_QUERY_IN_VERSE'] * 20;
}
}
rsortBy($scoringTable, 'SCORE');
//preprint_r($scoringTable);exit;
return $scoringTable;
}
