mirror of
https://github.com/alchemy-fr/Phraseanet.git
synced 2025-10-10 11:33:17 +00:00
2159 lines
62 KiB
PHP
2159 lines
62 KiB
PHP
<?php
|
|
class qparser {
|
|
|
|
var $ops = array( "et" => array("NODETYPE"=>PHRASEA_OP_AND, "CANNUM"=>false),
|
|
"and" => array("NODETYPE"=>PHRASEA_OP_AND, "CANNUM"=>false),
|
|
"ou" => array("NODETYPE"=>PHRASEA_OP_OR, "CANNUM"=>false),
|
|
"or" => array("NODETYPE"=>PHRASEA_OP_OR, "CANNUM"=>false),
|
|
"sauf" => array("NODETYPE"=>PHRASEA_OP_EXCEPT, "CANNUM"=>false),
|
|
"except" => array("NODETYPE"=>PHRASEA_OP_EXCEPT, "CANNUM"=>false),
|
|
"pres" => array("NODETYPE"=>PHRASEA_OP_NEAR, "CANNUM"=>true ),
|
|
"near" => array("NODETYPE"=>PHRASEA_OP_NEAR, "CANNUM"=>true ),
|
|
"avant" => array("NODETYPE"=>PHRASEA_OP_BEFORE, "CANNUM"=>true ),
|
|
"before" => array("NODETYPE"=>PHRASEA_OP_BEFORE, "CANNUM"=>true ),
|
|
"apres" => array("NODETYPE"=>PHRASEA_OP_AFTER, "CANNUM"=>true ),
|
|
"after" => array("NODETYPE"=>PHRASEA_OP_AFTER, "CANNUM"=>true ),
|
|
"dans" => array("NODETYPE"=>PHRASEA_OP_IN, "CANNUM"=>false),
|
|
"in" => array("NODETYPE"=>PHRASEA_OP_IN, "CANNUM"=>false) );
|
|
var $opk = array( "<" => array("NODETYPE"=>PHRASEA_OP_LT, "CANNUM"=>false),
|
|
">" => array("NODETYPE"=>PHRASEA_OP_GT, "CANNUM"=>false),
|
|
"<=" => array("NODETYPE"=>PHRASEA_OP_LEQT, "CANNUM"=>false),
|
|
">=" => array("NODETYPE"=>PHRASEA_OP_GEQT, "CANNUM"=>false),
|
|
"<>" => array("NODETYPE"=>PHRASEA_OP_NOTEQU, "CANNUM"=>false),
|
|
"=" => array("NODETYPE"=>PHRASEA_OP_EQUAL, "CANNUM"=>false),
|
|
":" => array("NODETYPE"=>PHRASEA_OP_COLON, "CANNUM"=>false) );
|
|
var $spw = array( "all" => array("CLASS"=>"PHRASEA_KW_ALL", "NODETYPE"=>PHRASEA_KW_ALL, "CANNUM"=>false),
|
|
"last" => array("CLASS"=>"PHRASEA_KW_LAST", "NODETYPE"=>PHRASEA_KW_LAST, "CANNUM"=>true),
|
|
// "first" => array("CLASS"=>PHRASEA_KW_FIRST, "CANNUM"=>true),
|
|
// "premiers" => array("CLASS"=>PHRASEA_KW_FIRST, "CANNUM"=>true),
|
|
"tout" => array("CLASS"=>"PHRASEA_KW_ALL", "NODETYPE"=>PHRASEA_KW_ALL, "CANNUM"=>false),
|
|
"derniers" => array("CLASS"=>"PHRASEA_KW_LAST", "NODETYPE"=>PHRASEA_KW_LAST, "CANNUM"=>true) );
|
|
|
|
var $quoted_defaultop = array("VALUE"=>"default_avant", "NODETYPE"=>PHRASEA_OP_BEFORE, "PNUM"=>0 );
|
|
var $defaultop = array("VALUE"=>"and", "NODETYPE"=>PHRASEA_OP_AND, "PNUM"=>NULL );
|
|
var $defaultlast = 12;
|
|
var $phq;
|
|
|
|
// var $noaccent = array();
|
|
// var $str_in = "àáâãäåçèéêëìíîïðñòóôõöùúûüýÿøæå";
|
|
// var $str_out = "aaaaaaceeeeiiiionooooouuuuyyo@a";
|
|
var $errmsg ="";
|
|
|
|
var $debug = false;
|
|
|
|
// var $proposals = Array("QRY"=>"","TERMS"=>array()); // un tableau qui contiendra des propositions de thesaurus pour les termes de l'arbre simple
|
|
var $proposals = Array("QRY"=>"","BASES"=>array()); // un tableau qui contiendra des propositions de thesaurus pour les termes de l'arbre simple
|
|
|
|
var $lng = null; // le skin, utile pour afficher le thesaurus dans la bonne langue
|
|
|
|
// le constructor
|
|
function qparser($lng = "???")
|
|
{
|
|
$this->lng = $lng;
|
|
// $l = strlen($this->str_in);
|
|
// for($i=0; $i<$l; $i++)
|
|
// $this->noaccent[substr($this->str_in, $i, 1)] = substr($this->str_out, $i, 1);
|
|
// $this->debug = true;
|
|
}
|
|
|
|
// simulation du trim manquant dans mb_xxxx, ne fonctionne qu'en utf8
|
|
function mb_trim($s, $encoding)
|
|
{
|
|
return(trim($s));
|
|
}
|
|
function mb_ltrim($s, $encoding)
|
|
{
|
|
return(ltrim($s));
|
|
}
|
|
|
|
function parsequery($phq)
|
|
{
|
|
// $phq = utf8_encode($phq);
|
|
// $phq0 = $phq;
|
|
// $phq = "";
|
|
// for($i=0; $i<strlen($phq0); $i++)
|
|
// $phq .= ord($c=$phq0[$i])==160 ? " " : $c;
|
|
if($this->debug)
|
|
{
|
|
for($i=0; $i<mb_strlen($phq, 'UTF-8'); $i++)
|
|
{
|
|
$c = mb_substr($phq, $i, 1, 'UTF-8');
|
|
printf("// %s : '%s' (%d octets)\n", $i, $c, strlen($c) );
|
|
}
|
|
}
|
|
// printf("<!-- PARSING $phq -->\n\n");
|
|
// $this->proposals = Array("QRY"=>"","TERMS"=>array());
|
|
$this->proposals = Array("QRY"=>"","BASES"=>array());
|
|
$this->phq = $this->mb_trim($phq, 'UTF-8');
|
|
if($this->phq != "")
|
|
return($this->maketree(0));
|
|
else
|
|
{
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= sprintf("\\n");
|
|
$this->errmsg .= _('qparser::la question est vide');
|
|
return(null);
|
|
}
|
|
}
|
|
|
|
function astext($tree)
|
|
{
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "SIMPLE":
|
|
if(is_array($tree["VALUE"]))
|
|
return(implode(" ", $tree["VALUE"]));
|
|
else
|
|
return($tree["VALUE"]);
|
|
break;
|
|
case "QSIMPLE":
|
|
if(is_array($tree["VALUE"]))
|
|
return("\"" . implode(" ", $tree["VALUE"]) . "\"");
|
|
else
|
|
return("\"" . $tree["VALUE"] . "\"");
|
|
break;
|
|
case "PHRASEA_KW_ALL":
|
|
return($tree["VALUE"][0]);
|
|
break;
|
|
case "PHRASEA_KW_LAST":
|
|
if($tree["PNUM"]!==null)
|
|
return("" . $tree["VALUE"][0] . "[" . $tree["PNUM"] . "]");
|
|
else
|
|
return($tree["VALUE"][0]);
|
|
break;
|
|
case "OPS":
|
|
case "OPK":
|
|
if(isset($tree["PNUM"]))
|
|
return("(" . $this->astext($tree["LB"]) . " " . $tree["VALUE"] . "[" . $tree["PNUM"] . "] " . $this->astext($tree["RB"]) . ")");
|
|
else
|
|
return("(" . $this->astext($tree["LB"]) . " " . $tree["VALUE"] . " " . $this->astext($tree["RB"]) . ")");
|
|
break;
|
|
}
|
|
}
|
|
|
|
function astable(&$tree)
|
|
{
|
|
$this->calc_complexity($tree);
|
|
// var_dump($tree);
|
|
$txt = "";
|
|
$this->astable2($txt, $tree);
|
|
$txt = "<table border=\"1\">\n<tr>\n" . $txt . "</tr>\n</table>\n";
|
|
return($txt);
|
|
}
|
|
function calc_complexity(&$tree)
|
|
{
|
|
if($tree)
|
|
{
|
|
if($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
return($tree["COMPLEXITY"] = $this->calc_complexity($tree["LB"]) + $this->calc_complexity($tree["RB"]));
|
|
else
|
|
return($tree["COMPLEXITY"] = 1);
|
|
}
|
|
}
|
|
|
|
function astable2(&$out, &$tree, $depth=0)
|
|
{
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "SIMPLE":
|
|
if(is_array($tree["VALUE"]))
|
|
$txt = implode(" ", $tree["VALUE"]);
|
|
else
|
|
$txt = $tree["VALUE"];
|
|
$out .= "\t<td>".$txt."</td>\n";
|
|
break;
|
|
case "QSIMPLE":
|
|
if(is_array($tree["VALUE"]))
|
|
$txt = implode(" ", $tree["VALUE"]);
|
|
else
|
|
$txt = $tree["VALUE"];
|
|
$out .= "\t<td>"".$txt.""</td>\n";
|
|
break;
|
|
case "PHRASEA_KW_ALL":
|
|
$out .= "\t<td>".$tree["VALUE"][0]."</td>\n";
|
|
break;
|
|
case "PHRASEA_KW_LAST":
|
|
if($tree["PNUM"]!==null)
|
|
$out .= "\t<td>".$tree["VALUE"][0] . "[" . $tree["PNUM"] . "]"."</td>\n";
|
|
else
|
|
$out .= "\t<td>".$tree["VALUE"][0]."</td>\n";
|
|
break;
|
|
case "OPS":
|
|
case "OPK":
|
|
$op = $tree["VALUE"];
|
|
if(isset($tree["PNUM"]))
|
|
$op .= "[" . $tree["PNUM"] . "]";
|
|
$out .= "\t<td colspan=\"".$tree["COMPLEXITY"]."\">$op</td>\n";
|
|
$this->astable2($out, $tree["LB"], $depth+1);
|
|
$this->astable2($out, $tree["RB"], $depth+1);
|
|
$out .= "</tr>\n<tr>\n";
|
|
break;
|
|
}
|
|
}
|
|
|
|
function dumpDiv(&$tree)
|
|
{
|
|
print("<div class=\"explain\">\n");
|
|
$this->dumpDiv2($tree);
|
|
print("</div>\n");
|
|
}
|
|
function dumpDiv2(&$tree, $depth=0)
|
|
{
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "SIMPLE":
|
|
if(is_array($tree["VALUE"]))
|
|
$s = implode("</font></b> , <b><font color='green'>", $tree["VALUE"]);
|
|
else
|
|
$s = $tree["VALUE"];
|
|
print(str_repeat("\t", $depth) . "<b><font color='green'>" . $s . "</font></b>\n");
|
|
case "QSIMPLE":
|
|
$s = "";
|
|
if(is_array($tree["VALUE"]))
|
|
$s = implode("</font></b> , <b><font color='green'>", $tree["VALUE"]);
|
|
else
|
|
$s = $tree["VALUE"];
|
|
print(str_repeat("\t", $depth) . ""<b><font color='green'>" . $s . "</font></b>"\n");
|
|
break;
|
|
case "PHRASEA_KW_ALL":
|
|
printf(str_repeat("\t", $depth) . "<b><font color='red'>%s</font></b>\n", $tree["VALUE"][0]);
|
|
break;
|
|
case "PHRASEA_KW_LAST":
|
|
if($tree["PNUM"]!==null)
|
|
printf(str_repeat("\t", $depth) . "<b><font color='blue'>%s <i>%s</i></font></b>\n", $tree["VALUE"][0], $tree["PNUM"]);
|
|
else
|
|
printf(str_repeat("\t", $depth) . "<b><font color='blue'>%s</font></b>\n", $tree["VALUE"][0]);
|
|
break;
|
|
// case PHRASEA_KW_FIRST:
|
|
// if($tree["PNUM"]!==null)
|
|
// printf("<b><font color='blue'>%s <i>%s</i></font></b>", $tree["VALUE"], $tree["PNUM"]);
|
|
// else
|
|
// printf("<b><font color='blue'>%s</font></b>", $tree["VALUE"]);
|
|
// break;
|
|
case "OPS":
|
|
case "OPK":
|
|
print(str_repeat("\t", $depth) . "<div>\n");
|
|
$this->dumpDiv2($tree["LB"], $depth+1);
|
|
print(str_repeat("\t", $depth) . "</div>\n");
|
|
print(str_repeat("\t", $depth) . "<div>\n");
|
|
if(isset($tree["PNUM"]))
|
|
printf(str_repeat("\t", $depth+1) . " %s[%s]\n", $tree["VALUE"], $tree["PNUM"]);
|
|
else
|
|
printf(str_repeat("\t", $depth+1) . " %s\n", $tree["VALUE"]);
|
|
print(str_repeat("\t", $depth) . "</div>\n");
|
|
print(str_repeat("\t", $depth) . "<div>\n");
|
|
$this->dumpDiv2($tree["RB"], $depth+1);
|
|
print(str_repeat("\t", $depth) . "</div>\n");
|
|
|
|
break;
|
|
}
|
|
}
|
|
|
|
function dump($tree)
|
|
{
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "SIMPLE":
|
|
if(is_array($tree["VALUE"]))
|
|
$s = implode("</font></b> , <b><font color='green'>", $tree["VALUE"]);
|
|
else
|
|
$s = $tree["VALUE"];
|
|
print("<b><font color='green'>" . $s . "</font></b>");
|
|
break;
|
|
case "QSIMPLE":
|
|
if(is_array($tree["VALUE"]))
|
|
$s = implode("</font></b> , <b><font color='green'>", $tree["VALUE"]);
|
|
else
|
|
$s = $tree["VALUE"];
|
|
print(""<b><font color='green'>" . $s . "</font></b>"");
|
|
break;
|
|
case "PHRASEA_KW_ALL":
|
|
printf("<b><font color='red'>%s</font></b>", $tree["VALUE"][0]);
|
|
break;
|
|
case "PHRASEA_KW_LAST":
|
|
if($tree["PNUM"]!==null)
|
|
printf("<b><font color='blue'>%s <i>%s</i></font></b>", $tree["VALUE"][0], $tree["PNUM"]);
|
|
else
|
|
printf("<b><font color='blue'>%s</font></b>", $tree["VALUE"][0]);
|
|
break;
|
|
// case PHRASEA_KW_FIRST:
|
|
// if($tree["PNUM"]!==null)
|
|
// printf("<b><font color='blue'>%s <i>%s</i></font></b>", $tree["VALUE"], $tree["PNUM"]);
|
|
// else
|
|
// printf("<b><font color='blue'>%s</font></b>", $tree["VALUE"]);
|
|
// break;
|
|
case "OPS":
|
|
case "OPK":
|
|
print("<table border='1'>");
|
|
print("<tr>");
|
|
print("<td colspan='2' align='center'>");
|
|
if(isset($tree["PNUM"]))
|
|
printf(" %s[%s] ", $tree["VALUE"], $tree["PNUM"]);
|
|
else
|
|
printf(" %s ", $tree["VALUE"]);
|
|
print("</td>");
|
|
print("</tr>");
|
|
print("<tr>");
|
|
print("<td width='50%' align='center' valign='top'>");
|
|
print($this->dump($tree["LB"]));
|
|
print("</td>");
|
|
print("<td width='50%' align='center' valign='top'>");
|
|
print($this->dump($tree["RB"]));
|
|
print("</td>");
|
|
print("</tr>");
|
|
print("</table>");
|
|
break;
|
|
}
|
|
}
|
|
|
|
function priority_opk(&$tree, $depth=0)
|
|
{
|
|
if(!$tree)
|
|
return;
|
|
if($tree["CLASS"]=="OPK" && ($tree["LB"]["CLASS"]=="OPS" || $tree["LB"]["CLASS"]=="OPK"))
|
|
{
|
|
// on a un truc du genre ((a ou b) < 5), on le transforme en (a ou (b < 5))
|
|
$t = $tree["LB"];
|
|
$tree["LB"] = $t["RB"];
|
|
$t["RB"] = $tree;
|
|
$tree = $t;
|
|
}
|
|
if(isset($tree["LB"]))
|
|
$this->priority_opk($tree["LB"], $depth+1);
|
|
if(isset($tree["RB"]))
|
|
$this->priority_opk($tree["RB"], $depth+1);
|
|
}
|
|
|
|
function distrib_opk(&$tree, $depth=0)
|
|
{
|
|
if(!$tree)
|
|
return;
|
|
if($tree["CLASS"]=="OPK" && ($tree["RB"]["CLASS"]=="OPS"))
|
|
{
|
|
// on a un truc du genre (a = (5 ou 6)), on le transforme en ((a = 5) ou (a = 6))
|
|
$tmp = array("CLASS"=>$tree["CLASS"],
|
|
"NODETYPE"=>$tree["NODETYPE"],
|
|
"VALUE"=>$tree["VALUE"],
|
|
"PNUM"=>$tree["PNUM"],
|
|
"LB"=>$tree["LB"],
|
|
"RB"=>$tree["RB"]["RB"],
|
|
"DEPTH"=>$tree["LB"]["DEPTH"] );
|
|
$t = $tree["RB"];
|
|
$tree["RB"] = $t["LB"];
|
|
$t["LB"] = $tree;
|
|
$t["RB"] = $tmp;
|
|
$tree = $t;
|
|
}
|
|
if(isset($tree["LB"]))
|
|
$this->distrib_opk($tree["LB"], $depth+1);
|
|
if(isset($tree["RB"]))
|
|
$this->distrib_opk($tree["RB"], $depth+1);
|
|
}
|
|
|
|
|
|
function thesaurus2_apply(&$tree, $bid)
|
|
{
|
|
if(!$tree)
|
|
return;
|
|
if( ($tree["CLASS"]=="SIMPLE" || $tree["CLASS"]=="QSIMPLE") && isset($tree["SREF"]) && isset($tree["SREF"]["TIDS"]) )
|
|
{
|
|
$tids = array();
|
|
foreach($tree["SREF"]["TIDS"] as $tid)
|
|
{
|
|
if($tid["bid"]==$bid)
|
|
$tids[] = $tid["pid"];
|
|
}
|
|
if(count($tids)>=1)
|
|
{
|
|
/*
|
|
if(count($tids)==1)
|
|
{
|
|
// on cherche un id simple, on utilisera la syntaxe sql 'like' (l'extension repérera elle méme la syntaxe car la value finit par '%')
|
|
$val = str_replace(".", "d", $tids[0]) . "d%";
|
|
$tree["VALUE"] = array($val);
|
|
}
|
|
else
|
|
{
|
|
// on cherche plusieurs id's, on utilisera la syntaxe 'regexp' (l'extension repérera elle méme la syntaxe car la value finit par '$'
|
|
$val = "";
|
|
foreach($tids as $tid)
|
|
$val .= ($val?"|":"") . "(" . str_replace(".", "d", $tid) . "d.*)";
|
|
$tree["VALUE"] = array("^" . $val);
|
|
}
|
|
*/
|
|
$tree["VALUE"] = array();
|
|
foreach($tids as $tid)
|
|
$tree["VALUE"][] = str_replace(".", "d", $tid) . "d%";;
|
|
}
|
|
else
|
|
{
|
|
// le mot n'est pas dans le thesaurus
|
|
}
|
|
/*
|
|
*/
|
|
}
|
|
if($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
{
|
|
$this->thesaurus2_apply($tree["LB"], $bid);
|
|
$this->thesaurus2_apply($tree["RB"], $bid);
|
|
}
|
|
}
|
|
|
|
// étend (ou remplace) la recherche sur les termes simples en recherche sur thesaurus
|
|
// ex: (a et b)
|
|
// full-text only : ==> (a et b)
|
|
// thesaurus only : ==> ((th:a) et (th:b))
|
|
// ft et thesaurus : ==> ((a ou (th:a)) et (b ou (th:b)))
|
|
// RETOURNE l'arbre résultat sans modifier l'arbre d'origine
|
|
function extendThesaurusOnTerms(&$tree, $useFullText, $useThesaurus, $keepfuzzy)
|
|
{
|
|
$copy = $tree;
|
|
$this->_extendThesaurusOnTerms($tree, $copy, $useFullText, $useThesaurus, $keepfuzzy, 0, "");
|
|
|
|
// var_dump($tree);
|
|
$this->proposals["QRY"] = "<span id=\"thprop_q\">" . $this->_queryAsHTML($tree) . "</span>";
|
|
|
|
return($copy);
|
|
}
|
|
function _extendThesaurusOnTerms(&$tree, &$copy, $useFullText, $useThesaurus, $keepfuzzy, $depth, $path)
|
|
{
|
|
if($depth == 0)
|
|
$ret = $tree;
|
|
if(!$useThesaurus)
|
|
return; // full-text only : inchangé
|
|
if(($tree["CLASS"]=="SIMPLE" || $tree["CLASS"]=="QSIMPLE"))
|
|
{
|
|
if(isset($tree["CONTEXT"]))
|
|
$copy = $this->_extendToThesaurus_Simple($tree, false, $keepfuzzy, $path);
|
|
else
|
|
$copy = $this->_extendToThesaurus_Simple($tree, $useFullText, $keepfuzzy, $path);
|
|
}
|
|
else
|
|
{
|
|
if($tree["CLASS"]=="OPK" && $tree["NODETYPE"]==PHRASEA_OP_COLON)
|
|
{
|
|
// on a 'field:value' , on traite 'value'
|
|
$tree["RB"]["PATH"] = $copy["RB"]["PATH"] = $path."R";
|
|
if(isset($tree["RB"]["CONTEXT"]))
|
|
$copy["CONTEXT"] = $tree["CONTEXT"] = $tree["RB"]["CONTEXT"];
|
|
else
|
|
if(!$keepfuzzy)
|
|
$copy["CONTEXT"] = $tree["CONTEXT"] = "*";
|
|
|
|
$copy["RB"]["SREF"] = &$tree["RB"];
|
|
}
|
|
else
|
|
{
|
|
$recursL = $recursR = false;
|
|
if($tree["CLASS"]=="OPS" && ($tree["NODETYPE"]==PHRASEA_OP_AND || $tree["NODETYPE"]==PHRASEA_OP_OR || $tree["NODETYPE"]==PHRASEA_OP_EXCEPT))
|
|
{
|
|
// on a une branche à gauche de 'ET', 'OU', 'SAUF'
|
|
$recursL = true;
|
|
}
|
|
if($tree["CLASS"]=="OPS" && ($tree["NODETYPE"]==PHRASEA_OP_AND || $tree["NODETYPE"]==PHRASEA_OP_OR || $tree["NODETYPE"]==PHRASEA_OP_EXCEPT))
|
|
{
|
|
// on a une branche à droite de 'ET', 'OU', 'SAUF'
|
|
$recursR = true;
|
|
}
|
|
if($recursL)
|
|
$this->_extendThesaurusOnTerms($tree["LB"], $copy["LB"], $useFullText, $useThesaurus, $keepfuzzy, $depth+1, $path."L");
|
|
if($recursR)
|
|
$this->_extendThesaurusOnTerms($tree["RB"], $copy["RB"], $useFullText, $useThesaurus, $keepfuzzy, $depth+1, $path."R");
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
// étend (ou remplace) un terme cherché en 'full-text' à une recherche thesaurus (champ non spécifié, tout le thésaurus = '*')
|
|
// le contexte éventuel est rapporté à l'opérateur ':'
|
|
// ex : a[k] ==> (a ou (TH :[k] a))
|
|
function _extendToThesaurus_Simple(&$simple, $keepFullText, $keepfuzzy, $path)
|
|
{
|
|
$simple["PATH"] = $path;
|
|
$context = null;
|
|
if(isset($simple["CONTEXT"]))
|
|
{
|
|
$context = $simple["CONTEXT"];
|
|
// unset($simple["CONTEXT"]);
|
|
}
|
|
if($keepFullText)
|
|
{
|
|
// on fait un OU entre la recherche ft et une recherche th
|
|
$tmp = array("CLASS"=>"OPS",
|
|
"NODETYPE"=>PHRASEA_OP_OR,
|
|
"VALUE"=>"OR",
|
|
"PNUM"=>null,
|
|
"DEPTH"=>$simple["DEPTH"],
|
|
"LB"=>$simple,
|
|
"RB"=>array("CLASS"=>"OPK",
|
|
"NODETYPE"=>PHRASEA_OP_COLON,
|
|
"VALUE"=>":",
|
|
// "CONTEXT"=>$context,
|
|
"PNUM"=>null,
|
|
"DEPTH"=>$simple["DEPTH"]+1,
|
|
"LB"=>array("CLASS"=>"SIMPLE",
|
|
"NODETYPE"=>PHRASEA_KEYLIST,
|
|
"VALUE"=>array("*"),
|
|
"DEPTH"=>$simple["DEPTH"]+2
|
|
),
|
|
"RB"=>$simple
|
|
)
|
|
);
|
|
// on vire le contexte du coté fulltext
|
|
unset($tmp["LB"]["CONTEXT"]);
|
|
// ajoute le contexte si nécéssaire
|
|
if($context !== null)
|
|
$tmp["RB"]["CONTEXT"] = $context;
|
|
else
|
|
if(!$keepfuzzy)
|
|
$tmp["RB"]["CONTEXT"] = "*";
|
|
// corrige les profondeurs des 2 copies du 'simple' d'origine
|
|
$tmp["LB"]["DEPTH"] += 1;
|
|
$tmp["RB"]["RB"]["DEPTH"] += 2;
|
|
// note une référence vers le terme d'origine
|
|
$tmp["RB"]["RB"]["SREF"] = &$simple;
|
|
$tmp["RB"]["RB"]["PATH"] = $path;
|
|
}
|
|
else
|
|
{
|
|
// on remplace le ft par du th
|
|
$tmp = array("CLASS"=>"OPK",
|
|
"NODETYPE"=>PHRASEA_OP_COLON,
|
|
"VALUE"=>":",
|
|
// "CONTEXT"=>$context,
|
|
"PNUM"=>null,
|
|
"DEPTH"=>$simple["DEPTH"]+1,
|
|
"LB"=>array("CLASS"=>"SIMPLE",
|
|
"NODETYPE"=>PHRASEA_KEYLIST,
|
|
"VALUE"=>array("*"),
|
|
"DEPTH"=>$simple["DEPTH"]+1
|
|
),
|
|
"RB"=>$simple
|
|
);
|
|
// ajoute le contexte si nécéssaire
|
|
if($context !== null)
|
|
$tmp["CONTEXT"] = $context;
|
|
else
|
|
if(!$keepfuzzy)
|
|
$tmp["CONTEXT"] = "*";
|
|
// corrige la profondeur de la copie du 'simple' d'origine
|
|
$tmp["RB"]["DEPTH"] += 1;
|
|
// note une référence vers le terme d'origine
|
|
$tmp["RB"]["SREF"] = &$simple;
|
|
$tmp["RB"]["PATH"] = $path;
|
|
}
|
|
return($tmp);
|
|
}
|
|
|
|
function thesaurus2(&$tree, $bid, $name, &$domthe, $searchsynonyms=true, $depth=0)
|
|
{
|
|
if($this->debug)
|
|
print("thesaurus2:\n\$tree=" . var_export($tree, true) . "\n");
|
|
|
|
if($depth==0)
|
|
$this->proposals["BASES"]["b$bid"] = array("BID"=>$bid, "NAME"=>$name, "TERMS"=>array());
|
|
|
|
if(!$tree)
|
|
return(0);
|
|
|
|
$ambigus = 0;
|
|
if($tree["CLASS"]=="OPK" && $tree["NODETYPE"]==PHRASEA_OP_COLON)
|
|
{
|
|
// $ambigus = $this->setTids($tree, $tree["RB"], $bid, $domthe, $searchsynonyms);
|
|
$ambigus = $this->setTids($tree, $bid, $domthe, $searchsynonyms);
|
|
}
|
|
elseif($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
{
|
|
$ambigus += $this->thesaurus2($tree["LB"], $bid, $name, $domthe, $searchsynonyms, $depth+1);
|
|
$ambigus += $this->thesaurus2($tree["RB"], $bid, $name, $domthe, $searchsynonyms, $depth+1);
|
|
}
|
|
return($ambigus);
|
|
}
|
|
|
|
function propAsHTML(&$node, &$html, $path, $depth=0)
|
|
{
|
|
global $parm;
|
|
if($depth > 0)
|
|
{
|
|
$tsy = array();
|
|
$lngfound = "?";
|
|
for($n=$node->firstChild; $n; $n=$n->nextSibling)
|
|
{
|
|
if($n->nodeName=="sy")
|
|
{
|
|
$lng = $n->getAttribute("lng");
|
|
if(!array_key_exists($lng, $tsy))
|
|
$tsy[$lng] = array();
|
|
$zsy = array("v"=>$n->getAttribute("v"), "w"=>$n->getAttribute("w"), "k"=>$n->getAttribute("k"));
|
|
|
|
if($lngfound=="?" || ($lng==$this->lng && $lngfound!=$lng))
|
|
{
|
|
$lngfound = $lng;
|
|
$syfound = $zsy;
|
|
}
|
|
else
|
|
{
|
|
}
|
|
$tsy[$lng][] = $zsy;
|
|
}
|
|
}
|
|
$alt = "";
|
|
foreach($tsy as $lng=>$tsy2)
|
|
{
|
|
foreach($tsy2 as $sy)
|
|
{
|
|
$alt .= $alt ? "\n" : "";
|
|
$alt .= "" . $lng . ": " . p4string::MakeString($sy["v"], "js");
|
|
}
|
|
}
|
|
|
|
$thtml = $syfound["v"];
|
|
$kjs = $syfound["k"] ? ("'" . p4string::MakeString($syfound["k"], "js") . "'") : "null";
|
|
$wjs = "'" . p4string::MakeString($syfound["w"], "js") . "'";
|
|
|
|
if($node->getAttribute("term"))
|
|
{
|
|
$thtml = "<b>" . $thtml . "</b>";
|
|
$node->removeAttribute("term");
|
|
}
|
|
|
|
$tab = str_repeat("\t", $depth);
|
|
$html .= $tab."<div style=\"position:relative; left:10px;\">\n";
|
|
$html .= $tab."\t<a title=\"".$alt."\" href=\"javascript:void();\" onclick=\"return(chgProp('".$path."', ".$wjs.", ".$kjs."));\">" . $thtml . "</a>\n";
|
|
}
|
|
|
|
$tsort = array();
|
|
for($n=$node->firstChild; $n; $n=$n->nextSibling)
|
|
{
|
|
if($n->nodeType==XML_ELEMENT_NODE && $n->getAttribute("marked")) // only 'te' marked
|
|
{
|
|
$lngfound = '?';
|
|
$syfound = '?';
|
|
for($n2=$n->firstChild; $n2; $n2=$n2->nextSibling)
|
|
{
|
|
if($n2->nodeName=='sy')
|
|
{
|
|
$lng = $n2->getAttribute('lng');
|
|
if($lngfound=="?" || ($lng==$this->lng && $lngfound!=$lng))
|
|
{
|
|
$lngfound = $lng;
|
|
$syfound = $n2->getAttribute('w');
|
|
}
|
|
}
|
|
}
|
|
$n->removeAttribute("marked");
|
|
for($i=0; array_key_exists($syfound.$i, $tsort) && $i<9999; $i++)
|
|
;
|
|
$tsort[$syfound.$i] = $n;
|
|
}
|
|
}
|
|
ksort($tsort);
|
|
|
|
// var_dump($tsort);
|
|
|
|
foreach($tsort as $n)
|
|
{
|
|
$this->propAsHTML($n, $html, $path, $depth+1);
|
|
}
|
|
|
|
if($depth > 0)
|
|
$html .= $tab."</div>\n";
|
|
}
|
|
|
|
function _queryAsHTML($tree, $depth=0)
|
|
{
|
|
// printf("astext : ");
|
|
// var_dump($tree);
|
|
if($depth==0)
|
|
{
|
|
$ambiguites = array("n"=>0, "refs"=>array());
|
|
}
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "SIMPLE":
|
|
case "QSIMPLE":
|
|
$w = is_array($tree["VALUE"]) ? implode(' ', $tree["VALUE"]) : $tree["VALUE"];
|
|
if(isset($tree["PATH"]))
|
|
{
|
|
$path = $tree["PATH"];
|
|
if(isset($tree["CONTEXT"]))
|
|
$w .= ' [' . $tree["CONTEXT"] . ']';
|
|
$txt = '<span id="thprop_a_'.$path.'">"'.$w.'"</span>';
|
|
}
|
|
else
|
|
{
|
|
if(isset($tree["CONTEXT"]))
|
|
$w .= '[' . $tree["CONTEXT"] . ']';
|
|
if($tree["CLASS"] == "QSIMPLE")
|
|
$txt = '"' . $w . '"';
|
|
else
|
|
$txt = $w;
|
|
}
|
|
return($txt);
|
|
break;
|
|
case "PHRASEA_KW_ALL":
|
|
return($tree["VALUE"][0]);
|
|
break;
|
|
case "PHRASEA_KW_LAST":
|
|
if($tree["PNUM"]!==null)
|
|
return("" . $tree["VALUE"][0] . "[" . $tree["PNUM"] . "]");
|
|
else
|
|
return($tree["VALUE"][0]);
|
|
break;
|
|
case "OPS":
|
|
case "OPK":
|
|
if(isset($tree["PNUM"]))
|
|
return('(' . $this->_queryAsHTML($tree["LB"], $depth+1) . ' ' . $tree["VALUE"] . '[' . $tree["PNUM"] . '] ' . $this->_queryAsHTML($tree["RB"], $depth+1) . ')');
|
|
else
|
|
return('(' . $this->_queryAsHTML($tree["LB"], $depth+1) . ' ' . $tree["VALUE"] . ' ' . $this->_queryAsHTML($tree["RB"], $depth+1) . ')');
|
|
break;
|
|
}
|
|
}
|
|
|
|
/*
|
|
function _queryAsHTML($tree, $mouseCallback="void", $depth=0)
|
|
{
|
|
// printf("astext : ");
|
|
// var_dump($tree);
|
|
if($depth==0)
|
|
{
|
|
$ambiguites = array("n"=>0, "refs"=>array());
|
|
}
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "SIMPLE":
|
|
case "QSIMPLE":
|
|
$w = is_array($tree["VALUE"]) ? implode(" ", $tree["VALUE"]) : $tree["VALUE"];
|
|
$tab = "\n" . str_repeat("\t", $depth);
|
|
if(isset($tree["PATH"]))
|
|
{
|
|
$path = $tree["PATH"];
|
|
if(isset($tree["CONTEXT"]))
|
|
$w .= " [" . $tree["CONTEXT"] . "]";
|
|
$txt = $tab . "<b><span onmouseover=\"return(".$mouseCallback."(event, '$path'));\" onmouseout=\"return(".$mouseCallback."(event, '$path'));\" id=\"thprop_a_".$path."\">";
|
|
$txt .= $tab . "\t\"" . $w . "";
|
|
// $txt .= $tab . "\t<span id='thprop_w_".$path."'></span>\"";
|
|
$txt .= "\"";
|
|
$txt .= $tab . "</span></b>\n";
|
|
}
|
|
else
|
|
{
|
|
if(isset($tree["CONTEXT"]))
|
|
$w .= "[" . $tree["CONTEXT"] . "]";
|
|
if($tree["CLASS"] == "QSIMPLE")
|
|
$txt = $tab . "\"" . $w . "\"\n";
|
|
else
|
|
$txt = $tab . "" . $w . "\n";
|
|
}
|
|
return($txt);
|
|
break;
|
|
case "PHRASEA_KW_ALL":
|
|
return($tree["VALUE"][0]);
|
|
break;
|
|
case "PHRASEA_KW_LAST":
|
|
if($tree["PNUM"]!==null)
|
|
return("" . $tree["VALUE"][0] . "[" . $tree["PNUM"] . "]");
|
|
else
|
|
return($tree["VALUE"][0]);
|
|
break;
|
|
case "OPS":
|
|
case "OPK":
|
|
if(isset($tree["PNUM"]))
|
|
return("(" . $this->_queryAsHTML($tree["LB"], $mouseCallback, $depth+1) . " " . $tree["VALUE"] . "[" . $tree["PNUM"] . "] " . $this->_queryAsHTML($tree["RB"], $mouseCallback, $depth+1) . ")");
|
|
else
|
|
return("(" . $this->_queryAsHTML($tree["LB"], $mouseCallback, $depth+1) . " " . $tree["VALUE"] . " " . $this->_queryAsHTML($tree["RB"], $mouseCallback, $depth+1) . ")");
|
|
break;
|
|
}
|
|
}
|
|
*/
|
|
function setTids(&$tree, $bid, &$domthe, $searchsynonyms)
|
|
{
|
|
if($this->debug)
|
|
print("============================ setTids:\n\$tree=" . var_export($tree, true) . "\n");
|
|
|
|
// $this->proposals["BASES"]["b$bid"] = array("BID"=>$bid, "TERMS"=>array());
|
|
|
|
$ambigus = 0;
|
|
if(is_array($w = $tree["RB"]["VALUE"]))
|
|
$t = $w = implode(" ", $w);
|
|
|
|
if(isset($tree["CONTEXT"]))
|
|
{
|
|
if(!$tree["CONTEXT"])
|
|
{
|
|
$x0 = "@w=\"" . $w ."\" and not(@k)";
|
|
}
|
|
else
|
|
{
|
|
if($tree["CONTEXT"]=="*")
|
|
{
|
|
$x0 = "@w=\"" . $w ."\"";
|
|
}
|
|
else
|
|
{
|
|
$x0 = "@w=\"" . $w ."\" and @k=\"" . $tree["CONTEXT"] . "\"";
|
|
$t .= " (" . $tree["CONTEXT"] . ")";
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
$x0 = "@w=\"" . $w ."\"";
|
|
}
|
|
|
|
$x = "/thesaurus//sy[" . $x0 ."]";
|
|
|
|
if($this->debug)
|
|
printf("searching thesaurus with xpath='%s'<br/>\n", $x);
|
|
|
|
$dxp = new DOMXPath($domthe);
|
|
$nodes = $dxp->query($x);
|
|
|
|
if(!isset($tree["RB"]["SREF"]["TIDS"]))
|
|
$tree["RB"]["SREF"]["TIDS"] = array();
|
|
if($nodes->length >= 1)
|
|
{
|
|
if($nodes->length == 1)
|
|
{
|
|
// on cherche un id simple, on utilisera la syntaxe sql 'like' (l'extension repérera elle méme la syntaxe car la value finira par '%')
|
|
$this->addtoTIDS($tree["RB"], $bid, $nodes->item(0));
|
|
// $this->thesaurusDOMNodes[] = $nodes->item(0);
|
|
}
|
|
else
|
|
{
|
|
// on cherche plusieurs id's, on utilisera la syntaxe 'regexp' (l'extension repérera elle meme la syntaxe car la value finira par '$')
|
|
$val = "";
|
|
foreach($nodes as $node)
|
|
{
|
|
if(!isset($tree["CONTEXT"]))
|
|
$ambigus++;
|
|
$this->addtoTIDS($tree["RB"], $bid, $node);
|
|
}
|
|
}
|
|
$path = $tree["RB"]["SREF"]["PATH"];
|
|
$prophtml = "";
|
|
$this->propAsHTML($domthe->documentElement, $prophtml, $path);
|
|
$this->proposals["BASES"]["b$bid"]["TERMS"][$path]["HTML"] = $prophtml;
|
|
}
|
|
else
|
|
{
|
|
// le mot n'est pas dans le thesaurus
|
|
}
|
|
return($ambigus);
|
|
}
|
|
/*
|
|
function dead_setTids(&$tree, &$simple, $bid, &$domthe, $searchsynonyms)
|
|
{
|
|
// if($this->debug)
|
|
print("setTids:\n\$tree=" . var_export($tree, true) . "\n");
|
|
|
|
$ambigus = 0;
|
|
if(is_array($w = $simple["VALUE"]))
|
|
$t = $w = implode(" ", $w);
|
|
|
|
if(isset($tree["CONTEXT"]))
|
|
{
|
|
if(!$tree["CONTEXT"])
|
|
{
|
|
$x0 = "@w=\"" . $w ."\" and not(@k)";
|
|
}
|
|
else
|
|
{
|
|
if($tree["CONTEXT"]=="*")
|
|
{
|
|
$x0 = "@w=\"" . $w ."\"";
|
|
}
|
|
else
|
|
{
|
|
$x0 = "@w=\"" . $w ."\" and @k=\"" . $tree["CONTEXT"] . "\"";
|
|
$t .= " (" . $tree["CONTEXT"] . ")";
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
$x0 = "@w=\"" . $w ."\"";
|
|
}
|
|
|
|
$x = "/thesaurus//sy[" . $x0 ."]";
|
|
|
|
if($this->debug)
|
|
printf("searching thesaurus with xpath='%s'<br/>\n", $x);
|
|
|
|
$dxp = new DOMXPath($domthe);
|
|
$nodes = $dxp->query($x);
|
|
|
|
if(!isset($tree["RB"]["SREF"]["TIDS"]))
|
|
$tree["RB"]["SREF"]["TIDS"] = array();
|
|
if($nodes->length >= 1)
|
|
{
|
|
if($nodes->length == 1)
|
|
{
|
|
// on cherche un id simple, on utilisera la syntaxe sql 'like' (l'extension repérera elle méme la syntaxe car la value finira par '%')
|
|
$this->addtoTIDS($tree["RB"], $bid, $nodes->item(0));
|
|
// $this->thesaurusDOMNodes[] = $nodes->item(0);
|
|
}
|
|
else
|
|
{
|
|
// on cherche plusieurs id's, on utilisera la syntaxe 'regexp' (l'extension repérera elle meme la syntaxe car la value finira par '$')
|
|
$val = "";
|
|
foreach($nodes as $node)
|
|
{
|
|
if(!isset($tree["CONTEXT"]))
|
|
$ambigus++;
|
|
$this->addtoTIDS($tree["RB"], $bid, $node);
|
|
}
|
|
}
|
|
$path = $tree["RB"]["SREF"]["PATH"];
|
|
$prophtml = "";
|
|
$this->propAsHTML($domthe->documentElement, $prophtml, $path);
|
|
$this->proposals["TERMS"][$path]["HTML"] = $prophtml;
|
|
}
|
|
else
|
|
{
|
|
// le mot n'est pas dans le thesaurus
|
|
}
|
|
return($ambigus);
|
|
}
|
|
*/
|
|
function containsColonOperator(&$tree)
|
|
{
|
|
if(!$tree)
|
|
return(false);
|
|
if($tree["CLASS"]=="OPK" && $tree["NODETYPE"]==PHRASEA_OP_COLON && ($tree["RB"]["CLASS"]=="SIMPLE" || $tree["RB"]["CLASS"]=="QSIMPLE"))
|
|
return(true);
|
|
$ret = false;
|
|
if($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
{
|
|
$ret |= $this->containsColonOperator($tree["LB"]);
|
|
$ret |= $this->containsColonOperator($tree["RB"]);
|
|
}
|
|
return($ret);
|
|
}
|
|
|
|
function addtoTIDS(&$extendednode, $bid, $DOMnode) // ajoute un tid en évitant les doublons
|
|
{
|
|
$id = $DOMnode->getAttribute("id");
|
|
$pid = $DOMnode->parentNode->getAttribute("id");
|
|
$lng = $DOMnode->getAttribute("lng");
|
|
$w = $DOMnode->getAttribute("w");
|
|
$k = $DOMnode->getAttribute("k");
|
|
$p = $DOMnode->parentNode->getAttribute("v"); // le terme général (pére) du terme recherché : utile pour la levée d'ambiguité
|
|
|
|
$path = $extendednode["SREF"]["PATH"];
|
|
if($this->debug)
|
|
printf("found node id='%s', v='%s' w='%s', k='%s', p='%s' for node-path=%s \n", $id, $DOMnode->getAttribute("v"), $w, $k, $p, $path);
|
|
|
|
if(!$k)
|
|
$k = null;
|
|
|
|
$found = false;
|
|
foreach($extendednode["SREF"]["TIDS"] as $ztid)
|
|
{
|
|
if($ztid["bid"] != $bid)
|
|
continue;
|
|
if($ztid["pid"] == $pid)
|
|
{
|
|
$found = true;
|
|
}
|
|
else
|
|
{
|
|
// if($ztid["w"]==$w && $ztid["k"]==$k && $ztid["lng"]==$lng)
|
|
// {
|
|
// // FATAL : il y a un doublon réel dans le thesaurus de cette base (méme terme, méme contexte)
|
|
// // printf("<font color='red'>FATAL doublon on base %d (%s[%s])</font>\n", $bid, $w, $k);
|
|
// $found = true;
|
|
// break;
|
|
// }
|
|
}
|
|
}
|
|
if(!$found)
|
|
$extendednode["SREF"]["TIDS"][] = array("bid"=>$bid, "pid"=>$pid, "id"=>$id, "w"=>$w, "k"=>$k, "lng"=>$lng, "p"=>$p);
|
|
|
|
// on liste les propositions de thésaurus pour ce node (dans l'arbre simple)
|
|
if(!isset($this->proposals["BASES"]["b$bid"]["TERMS"][$path]))
|
|
{
|
|
// $this->proposals["TERMS"][$path] = array("TERM"=>implode(" ", $extendednode["VALUE"]), "PROPOSALS"=>array());
|
|
$term = implode(" ", $extendednode["VALUE"]);
|
|
if(isset($extendednode["CONTEXT"]) && $extendednode["CONTEXT"])
|
|
{
|
|
$term .= " (" . $extendednode["CONTEXT"] . ")";
|
|
}
|
|
$this->proposals["BASES"]["b$bid"]["TERMS"][$path] = array("TERM"=>$term); // , "PROPOSALS"=>array() ); //, "PROPOSALS_TREE"=>new DOMDocument("1.0", "UTF-8"));
|
|
}
|
|
// printf("<%s id='%s'><br/>\n", $DOMnode->tagName, $DOMnode->getAttribute("id"));
|
|
// printf("<b>found node <%s id='%s' w='%s' k='%s'></b><br/>\n", $DOMnode->nodeName, $DOMnode->getAttribute('id'), $DOMnode->getAttribute('w'), $DOMnode->getAttribute('k'));
|
|
|
|
// on marque le terme principal
|
|
$DOMnode->parentNode->setAttribute("term", "1");
|
|
// on commence par marquer les fils directs. rappel:$DOMnode pointe sur un sy
|
|
for($node=$DOMnode->parentNode->firstChild; $node; $node=$node->nextSibling)
|
|
{
|
|
if($node->nodeName=="te")
|
|
{
|
|
$node->setAttribute("marked", "1");
|
|
}
|
|
}
|
|
// puis par remonter au père
|
|
for($node=$DOMnode->parentNode; $node && $node->nodeType==XML_ELEMENT_NODE && $node->parentNode; $node=$node->parentNode)
|
|
{
|
|
$id = $node->getAttribute("id");
|
|
if(!$id)
|
|
break; // on a dépassé la racine du thésaurus
|
|
$node->setAttribute("marked", "1");
|
|
// printf("<%s id='%s'<br/>\n", $node->nodeName, $node->getAttribute("id"));
|
|
}
|
|
}
|
|
|
|
function astext_ambigu($tree, &$ambiguites, $mouseCallback="void", $depth=0)
|
|
{
|
|
// printf("astext : ");
|
|
// var_dump($tree);
|
|
if($depth==0)
|
|
{
|
|
$ambiguites = array("n"=>0, "refs"=>array());
|
|
}
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "SIMPLE":
|
|
case "QSIMPLE":
|
|
$prelink = $postlink = "";
|
|
$w = is_array($tree["VALUE"]) ? implode(" ", $tree["VALUE"]) : $tree["VALUE"];
|
|
$tab = "\n" . str_repeat("\t", $depth);
|
|
if(isset($tree["TIDS"]) && count($tree["TIDS"])>1)
|
|
{
|
|
$ambiguites["refs"][$n=$ambiguites["n"]] = &$tree;
|
|
$txt = $tab . "<b><span onmouseover=\"return(".$mouseCallback."(event, $n));\" onmouseout=\"return(".$mouseCallback."(event, $n));\" id=\"thamb_a_".$ambiguites["n"]."\">";
|
|
$txt .= $tab . "\t\"" . $w . "";
|
|
$txt .= $tab . "\t<span id='thamb_w_".$ambiguites["n"]."'></span>\"";
|
|
$txt .= $tab . "</span></b>\n";
|
|
$ambiguites["n"]++;
|
|
}
|
|
else
|
|
{
|
|
if(isset($tree["CONTEXT"]))
|
|
$w .= "[" . $tree["CONTEXT"] . "]";
|
|
if($tree["CLASS"] == "QSIMPLE")
|
|
$txt = $tab . "\"" . $w . "\"\n";
|
|
else
|
|
$txt = $tab . "" . $w . "\n";
|
|
}
|
|
return($txt);
|
|
break;
|
|
case "PHRASEA_KW_ALL":
|
|
return($tree["VALUE"][0]);
|
|
break;
|
|
case "PHRASEA_KW_LAST":
|
|
if($tree["PNUM"]!==null)
|
|
return("" . $tree["VALUE"][0] . "[" . $tree["PNUM"] . "]");
|
|
else
|
|
return($tree["VALUE"][0]);
|
|
break;
|
|
case "OPS":
|
|
case "OPK":
|
|
if(isset($tree["PNUM"]))
|
|
return("(" . $this->astext_ambigu($tree["LB"], $ambiguites, $mouseCallback, $depth+1) . " " . $tree["VALUE"] . "[" . $tree["PNUM"] . "] " . $this->astext_ambigu($tree["RB"], $ambiguites, $mouseCallback, $depth+1) . ")");
|
|
else
|
|
return("(" . $this->astext_ambigu($tree["LB"], $ambiguites, $mouseCallback, $depth+1) . " " . $tree["VALUE"] . " " . $this->astext_ambigu($tree["RB"], $ambiguites, $mouseCallback, $depth+1) . ")");
|
|
break;
|
|
}
|
|
}
|
|
|
|
function get_ambigu(&$tree, $mouseCallback="void", $depth=0)
|
|
{
|
|
if(!$tree)
|
|
return("");
|
|
unset($tree["DEPTH"]);
|
|
if($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
{
|
|
$this->get_ambigu($tree["LB"], $mouseCallback, $depth+1);
|
|
$this->get_ambigu($tree["RB"], $mouseCallback, $depth+1);
|
|
}
|
|
else
|
|
{
|
|
}
|
|
if($depth==0)
|
|
{
|
|
$t_ambiguites = array();
|
|
$r = ($this->astext_ambigu($tree, $t_ambiguites, $mouseCallback));
|
|
$t_ambiguites["query"] = $r;
|
|
return($t_ambiguites);
|
|
}
|
|
}
|
|
|
|
function set_default(&$tree, &$emptyw, $depth=0)
|
|
{
|
|
if(!$tree)
|
|
return(true);
|
|
if($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
{
|
|
if($tree["CLASS"]=="OPS")
|
|
{
|
|
if(!$this->set_default($tree["LB"], $emptyw, $depth+1))
|
|
return(false);
|
|
if(!$this->set_default($tree["RB"], $emptyw, $depth+1))
|
|
return(false);
|
|
}
|
|
else // OPK !
|
|
{
|
|
// jy 20041223 : ne pas appliquer d'op. par def. derriere un op arith.
|
|
// ex : "d < 1/2/2003" : grouper la liste "1","2","2004" en "mot" unique
|
|
if(!$tree["LB"] || ($tree["LB"]["CLASS"]!="SIMPLE" && $tree["LB"]["CLASS"]!="QSIMPLE") || (is_array($tree["LB"]["VALUE"]) && count($tree["LB"]["VALUE"])!=1) )
|
|
{
|
|
// un op. arith. doit étre précédé d'un seul nom de champ
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= sprintf("\\n");
|
|
$this->errmsg .= sprintf(_('qparser::Formulation incorrecte, un nom de champs est attendu avant l operateur %s'),$tree["VALUE"]);
|
|
return(false);
|
|
}
|
|
if(!$tree["RB"] || ($tree["RB"]["CLASS"]!="SIMPLE" && $tree["RB"]["CLASS"]!="QSIMPLE"))
|
|
{
|
|
// un op. arith. doit étre suivi d'une valeur
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= sprintf("\\n");
|
|
$this->errmsg .= sprintf(_('qparser::Formulation incorrecte, une valeur est attendue apres l operateur %s'),$tree["VALUE"]);
|
|
return(false);
|
|
}
|
|
if(is_array($tree["RB"]["VALUE"]))
|
|
{
|
|
$lw = "";
|
|
foreach($tree["RB"]["VALUE"] as $w)
|
|
$lw .= ($lw==""?"":" ") . $w;
|
|
$tree["RB"]["VALUE"] = $lw;
|
|
}
|
|
}
|
|
|
|
/** gestion des branches null
|
|
* a revoir car ca ppete pas d'erreur mais corrige automatiquement
|
|
****/
|
|
if(!isset($tree["RB"]))
|
|
$tree=$tree["LB"];
|
|
else
|
|
if(!isset($tree["LB"]))
|
|
$tree=$tree["RB"];
|
|
|
|
}
|
|
else
|
|
{
|
|
if(($tree["CLASS"]=="SIMPLE" || $tree["CLASS"]=="QSIMPLE"))
|
|
{
|
|
if(is_array($tree["VALUE"]))
|
|
{
|
|
$treetmp = null;
|
|
$pnum = 0;
|
|
for( $i=0; $i<count($tree["VALUE"]); $i++ )
|
|
{
|
|
// gestion mot vide
|
|
if( isset($emptyw[$tree["VALUE"][$i]]) || $tree["VALUE"][$i]=="?" || $tree["VALUE"][$i]=="*")
|
|
{
|
|
// on a forcé les '?' ou '*' isolés comme des mots vides
|
|
$pnum++;
|
|
}
|
|
else
|
|
{
|
|
if($treetmp == null)
|
|
{
|
|
$treetmp = array("CLASS"=>$tree["CLASS"],
|
|
"NODETYPE"=>$tree["NODETYPE"],
|
|
"VALUE"=>$tree["VALUE"][$i],
|
|
"PNUM"=>$tree["PNUM"],
|
|
"DEPTH"=>$tree["DEPTH"] );
|
|
$pnum = 0;
|
|
}
|
|
else
|
|
{
|
|
$dop = $tree["CLASS"]=="QSIMPLE" ? $this->quoted_defaultop : $this->defaultop ;
|
|
$treetmp = array("CLASS"=>"OPS",
|
|
"VALUE"=>$dop["VALUE"],
|
|
"NODETYPE"=>$dop["NODETYPE"],
|
|
"PNUM"=>$pnum, // peut-être écrasé par defaultop
|
|
"DEPTH"=>$depth,
|
|
"LB"=>$treetmp,
|
|
"RB"=>array("CLASS"=>$tree["CLASS"],
|
|
"NODETYPE"=>$tree["NODETYPE"],
|
|
"VALUE"=>$tree["VALUE"][$i],
|
|
"PNUM"=>$tree["PNUM"],
|
|
"DEPTH"=>$tree["DEPTH"] )
|
|
);
|
|
if(array_key_exists("PNUM", $dop))
|
|
$treetmp["PNUM"] = $dop["PNUM"];
|
|
$pnum = 0;
|
|
}
|
|
}
|
|
}
|
|
$tree = $treetmp;
|
|
}
|
|
}
|
|
}
|
|
return(true);
|
|
}
|
|
|
|
function factor_or(&$tree)
|
|
{
|
|
do
|
|
$n = $this->factor_or2($tree);
|
|
while($n>0);
|
|
}
|
|
|
|
function factor_or2(&$tree, $depth=0)
|
|
{
|
|
// printf("<hr><b>factor_or depth=%s sur</b><br>", $depth);
|
|
// var_dump($tree);
|
|
$nmodif=0;
|
|
if($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
{
|
|
if($tree["NODETYPE"]==PHRASEA_OP_OR && ($tree["LB"]["CLASS"]=="SIMPLE" || $tree["LB"]["CLASS"]=="QSIMPLE") && ($tree["RB"]["CLASS"]=="SIMPLE" || $tree["RB"]["CLASS"]=="QSIMPLE"))
|
|
{
|
|
$tree["CLASS"]="SIMPLE";
|
|
$tree["NODETYPE"]=PHRASEA_KEYLIST;
|
|
$tree["VALUE"] = is_array($tree["LB"]["VALUE"]) ? $tree["LB"]["VALUE"] : array($tree["LB"]["VALUE"]);
|
|
if(is_array($tree["RB"]["VALUE"]))
|
|
{
|
|
foreach($tree["RB"]["VALUE"] as $v)
|
|
$tree["VALUE"][] = $v;
|
|
}
|
|
else
|
|
$tree["VALUE"][] = $tree["RB"]["VALUE"];
|
|
unset($tree["LB"]);
|
|
unset($tree["RB"]);
|
|
// unset($tree["NODETYPE"]);
|
|
unset($tree["PNUM"]);
|
|
$nmodif++;
|
|
// printf("<hr><b>donne</b><br>");
|
|
// var_dump($tree);
|
|
}
|
|
else
|
|
{
|
|
$nmodif += $this->factor_or2($tree["LB"], $depth+1);
|
|
$nmodif += $this->factor_or2($tree["RB"], $depth+1);
|
|
}
|
|
}
|
|
// printf("<br>return %s<br>", $nmodif);
|
|
return($nmodif);
|
|
}
|
|
|
|
function setNumValue(&$tree,$xmlstruct, $depth=0)
|
|
{
|
|
// var_dump($tree);
|
|
if($tree["CLASS"]=="OPK")
|
|
{
|
|
if(isset($tree["RB"]) && ($tree["RB"]["CLASS"]=="SIMPLE" || $tree["RB"]["CLASS"]=="QSIMPLE") && ($tree["LB"]["CLASS"]=="SIMPLE" || $tree["LB"]["CLASS"]=="QSIMPLE") )
|
|
{
|
|
if($sxe = simplexml_load_string($xmlstruct))
|
|
{
|
|
$z = $sxe->xpath('/record/description');
|
|
if($z && is_array($z))
|
|
{
|
|
foreach($z[0] as $ki => $vi)
|
|
{
|
|
$champ = null ;
|
|
if(is_array($tree["LB"]["VALUE"]))
|
|
$champ = $tree["LB"]["VALUE"][0];
|
|
else
|
|
$champ = $tree["LB"]["VALUE"];
|
|
if($champ && strtoupper($ki)==strtoupper($champ))
|
|
{
|
|
foreach( $vi->attributes() as $propname => $val )
|
|
{
|
|
if(strtoupper($propname)==strtoupper("type"))
|
|
{
|
|
if( $tree["NODETYPE"]==PHRASEA_OP_EQUAL ) // cas particulier du "=" sur une date
|
|
$this->changeNodeEquals($tree,$val);
|
|
else
|
|
$this->setNumValue2($tree["RB"],$val);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
if(isset($tree["LB"]))
|
|
$this->setNumValue($tree["LB"],$xmlstruct, $depth+1);
|
|
if(isset($tree["RB"]))
|
|
$this->setNumValue($tree["RB"],$xmlstruct, $depth+1);
|
|
}
|
|
|
|
function changeNodeEquals(&$branch,$type)
|
|
{
|
|
if( strtoupper($type)==strtoupper("Date") )
|
|
{
|
|
$branch = $this->changeNodeEquals2($branch);
|
|
}
|
|
}
|
|
|
|
function changeNodeEquals2($oneBranch)
|
|
{
|
|
## creation branche gauche avec ">="
|
|
// print("changeNodeEquals2\n");
|
|
// print("creation branche gauche ( '>=' ) \n");
|
|
$newTreeLB = array("CLASS"=>"OPK",
|
|
"VALUE"=>">=",
|
|
"NODETYPE"=>PHRASEA_OP_GEQT,
|
|
"PNUM"=>NULL,
|
|
"DEPTH"=>0,
|
|
"LB"=>$oneBranch["LB"],
|
|
"RB"=>array("CLASS"=>"SIMPLE",
|
|
"VALUE"=>$this->isoDate($oneBranch["RB"]["VALUE"],false),
|
|
"NODETYPE"=>PHRASEA_KEYLIST,
|
|
"PNUM"=>NULL,
|
|
"DEPTH"=>0)
|
|
);
|
|
// var_dump($newTreeLB);
|
|
// print("fin creation branche gauche ( '>=' ) \n");
|
|
## fin creation branche gauche ( ">=" )
|
|
|
|
|
|
## creation branche droite avec "<="
|
|
// print("creation branche droite avec '<=' \n");
|
|
$newTreeRB = array("CLASS"=>"OPK",
|
|
"VALUE"=>"<=",
|
|
"NODETYPE"=>PHRASEA_OP_LEQT,
|
|
"PNUM"=>NULL,
|
|
"DEPTH"=>0,
|
|
"LB"=>$oneBranch["LB"],
|
|
"RB"=>array("CLASS"=>"SIMPLE",
|
|
"VALUE"=>$this->isoDate($oneBranch["RB"]["VALUE"],true),
|
|
"NODETYPE"=>PHRASEA_KEYLIST,
|
|
"PNUM"=>NULL,
|
|
"DEPTH"=>0)
|
|
);
|
|
// var_dump($newTreeRB);
|
|
// print("fin creation branche droite avec '<=' \n");
|
|
## fin creation branche droite ( "<=" )
|
|
|
|
$tree = array("CLASS"=>"OPS",
|
|
"VALUE"=>"et",
|
|
"NODETYPE"=>PHRASEA_OP_AND,
|
|
"PNUM"=>NULL,
|
|
"DEPTH"=>0,
|
|
"LB"=>$newTreeLB,
|
|
"RB"=>$newTreeRB );
|
|
|
|
|
|
// et on le retourne
|
|
// var_dump($tree);
|
|
return $tree;
|
|
}
|
|
|
|
|
|
function setNumValue2(&$branch,$type)
|
|
{
|
|
if( strtoupper($type)==strtoupper("Date") )
|
|
{
|
|
$dateEnIso = $this->isoDate($branch["VALUE"]);
|
|
$branch["VALUE"] = $dateEnIso;
|
|
}
|
|
}
|
|
function isoDate($onedate,$max=false)
|
|
{
|
|
$v_y = "1900";
|
|
$v_m = "01";
|
|
$v_d = "01";
|
|
|
|
$v_h = $v_minutes = $v_s = "00";
|
|
if($max)
|
|
{
|
|
$v_h = $v_minutes = $v_s = "99";
|
|
}
|
|
$tmp = $onedate;
|
|
|
|
if(!is_array($tmp))
|
|
$tmp = explode(" ", $tmp );
|
|
|
|
switch(sizeof($tmp))
|
|
{
|
|
// on a une date complete séparé avec des espaces, slash ou tiret
|
|
case 3 :
|
|
if(strlen($tmp[0])==4)
|
|
{
|
|
$v_y = $tmp[0];
|
|
$v_m = $tmp[1];
|
|
$v_d = $tmp[2];
|
|
// on a l'année en premier, on suppose alors que c'est de la forme YYYY MM DD
|
|
}
|
|
elseif(strlen($tmp[2])==4)
|
|
{
|
|
// on a l'année en dernier, on suppose alors que c'est de la forme DD MM YYYY
|
|
$v_y = $tmp[2];
|
|
$v_m = $tmp[1];
|
|
$v_d = $tmp[0];
|
|
}
|
|
else
|
|
{
|
|
// l'année est sur un 2 chiffre et pas 4
|
|
// ca fou la zone
|
|
|
|
$v_d = $tmp[0];
|
|
$v_m = $tmp[1];
|
|
if($tmp[2]<20)
|
|
$v_y = "20".$tmp[2];
|
|
else
|
|
$v_y = "19".$tmp[2];
|
|
}
|
|
break;
|
|
|
|
case 2 :
|
|
// On supposerait n'avoir que le mois et l'année
|
|
if(strlen($tmp[0])==4)
|
|
{
|
|
$v_y = $tmp[0];
|
|
$v_m = $tmp[1];
|
|
// on a l'année en premier, on suppose alors que c'est de la forme YYYY MM DD
|
|
if($max)
|
|
$v_d = "99" ;
|
|
else
|
|
$v_d = "00" ;
|
|
|
|
}
|
|
elseif(strlen($tmp[1])==4)
|
|
{
|
|
// on a l'année en premier, on suppose alors que c'est de la forme DD MM YYYY
|
|
$v_y = $tmp[1];
|
|
$v_m = $tmp[0];
|
|
if($max)
|
|
$v_d = "99" ;
|
|
else
|
|
$v_d = "00" ;
|
|
}
|
|
else
|
|
{
|
|
// on a l'anné sur 2 chiffres
|
|
if($tmp[1]<20)
|
|
$v_y = "20".$tmp[1];
|
|
else
|
|
$v_y = "19".$tmp[1];
|
|
$v_m = $tmp[0];
|
|
if($max)
|
|
$v_d = "99" ;
|
|
else
|
|
$v_d = "00" ;
|
|
}
|
|
break;
|
|
|
|
|
|
// lé ca devient la zone pour savoir si on a que l'année ou si c'est une date sans espaces,slash ou tiret
|
|
case 1 :
|
|
switch(strlen($tmp[0]))
|
|
{
|
|
case 14 :
|
|
// date iso YYYYMMDDHHMMSS
|
|
$v_y = substr($tmp[0],0,4);
|
|
$v_m = substr($tmp[0],4,2);
|
|
$v_d = substr($tmp[0],6,2);
|
|
$v_h = substr($tmp[0],8,2);
|
|
$v_minutes = substr($tmp[0],10,2);
|
|
$v_s = substr($tmp[0],12,2);
|
|
break;
|
|
case 8 :
|
|
// date iso YYYYMMDD
|
|
$v_y = substr($tmp[0],0,4);
|
|
$v_m = substr($tmp[0],4,2);
|
|
$v_d = substr($tmp[0],6,2);
|
|
break;
|
|
case 6 :
|
|
// date iso YYYYMM
|
|
$v_y = substr($tmp[0],0,4);
|
|
$v_m = substr($tmp[0],4,2);
|
|
if($max)
|
|
$v_d = "99" ;
|
|
else
|
|
$v_d = "00" ;
|
|
break;
|
|
case 4 :
|
|
// date iso YYYY
|
|
$v_y = $tmp[0];
|
|
|
|
if($max)
|
|
$v_m = "99" ;
|
|
else
|
|
$v_m = "00" ;
|
|
|
|
if($max)
|
|
$v_d = "99" ;
|
|
else
|
|
$v_d = "00" ;
|
|
break;
|
|
case 2 :
|
|
// date iso YY
|
|
if($tmp[0]<20)
|
|
$v_y = "20".$tmp[0];
|
|
else
|
|
$v_y = "19".$tmp[0];
|
|
|
|
if($max)
|
|
$v_m = "99" ;
|
|
else
|
|
$v_m = "00" ;
|
|
|
|
if($max)
|
|
$v_d = "99" ;
|
|
else
|
|
$v_d = "00" ;
|
|
break;
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
}
|
|
|
|
return("" . $v_y . $v_m . $v_d . $v_h . $v_minutes . $v_s);
|
|
|
|
}
|
|
|
|
|
|
function distrib_in(&$tree, $depth=0)
|
|
{
|
|
$opdistrib = array(PHRASEA_OP_AND, PHRASEA_OP_OR, PHRASEA_OP_EXCEPT, PHRASEA_OP_NEAR, PHRASEA_OP_BEFORE, PHRASEA_OP_AFTER); // ces opérateurs sont 'distribuables' autour d'un 'IN'
|
|
// printf("<hr><b>distrib_in depth=%s sur</b><br>", $depth);
|
|
// var_dump($tree);
|
|
|
|
if($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
{
|
|
if($tree["NODETYPE"]==PHRASEA_OP_IN || $tree["CLASS"]=="OPK")
|
|
{
|
|
if($tree["LB"]["CLASS"]=="OPK")
|
|
{
|
|
// on a un truc du genre '(t1 = t2) dans t3'
|
|
// ... on ne fait rien
|
|
}
|
|
if($tree["LB"]["CLASS"]=="OPS" && in_array($tree["LB"]["NODETYPE"], $opdistrib))
|
|
{
|
|
// on a un truc du genre '(t1 op t2) {dans|=} t3', on distribue le dans é t1 et t2
|
|
// ==> ((t1 dans t3) op (t2 dans t3))
|
|
$m_v = $tree["VALUE"];
|
|
$m_t = $tree["CLASS"];
|
|
$m_o = $tree["NODETYPE"];
|
|
$m_n = $tree["PNUM"];
|
|
|
|
$tree["CLASS"] = $tree["LB"]["CLASS"];
|
|
$tree["NODETYPE"] = $tree["LB"]["NODETYPE"];
|
|
$tree["VALUE"] = $tree["LB"]["VALUE"];
|
|
$tree["PNUM"] = $tree["LB"]["PNUM"];
|
|
|
|
$tree["LB"]["CLASS"] = $m_t;
|
|
$tree["LB"]["NODETYPE"] = $m_o;
|
|
$tree["LB"]["VALUE"] = $m_v;
|
|
$tree["LB"]["PNUM"] = $m_n;
|
|
|
|
$tree["RB"] = array("CLASS"=>$m_t,
|
|
"NODETYPE"=>$m_o,
|
|
"VALUE"=>$m_v,
|
|
"PNUM"=>$m_n,
|
|
"LB"=>$tree["LB"]["RB"],
|
|
"RB"=>$tree["RB"]);
|
|
|
|
$tree["LB"]["RB"] = $tree["RB"]["RB"];
|
|
// var_dump($tree);
|
|
// return;
|
|
}
|
|
|
|
|
|
if($tree["RB"]["CLASS"]=="OPS" && in_array($tree["RB"]["NODETYPE"], $opdistrib))
|
|
{
|
|
|
|
// on a un truc du genre 't1 {dans|=} (t2 op t3)', on distribue le dans é t2 et t3
|
|
// ==> ((t1 dans t2) ou (t1 dans t3))
|
|
$m_v = $tree["VALUE"];
|
|
$m_t = $tree["CLASS"];
|
|
$m_o = $tree["NODETYPE"];
|
|
$m_n = $tree["PNUM"];
|
|
|
|
$tree["CLASS"] = $tree["RB"]["CLASS"];
|
|
$tree["NODETYPE"] = $tree["RB"]["NODETYPE"];
|
|
$tree["VALUE"] = $tree["RB"]["VALUE"];
|
|
$tree["PNUM"] = $tree["RB"]["PNUM"];
|
|
|
|
$tree["RB"]["CLASS"] = $m_t;
|
|
$tree["RB"]["NODETYPE"] = $m_o;
|
|
$tree["RB"]["VALUE"] = $m_v;
|
|
$tree["RB"]["PNUM"] = $m_n;
|
|
|
|
$tree["LB"] = array("CLASS"=>$m_t,
|
|
"NODETYPE"=>$m_o,
|
|
"VALUE"=>$m_v,
|
|
"PNUM"=>$m_n,
|
|
"LB"=>$tree["LB"],
|
|
"RB"=>$tree["RB"]["LB"]);
|
|
|
|
$tree["RB"]["LB"] = $tree["LB"]["LB"];
|
|
}
|
|
}
|
|
$this->distrib_in($tree["LB"], $depth+1);
|
|
$this->distrib_in($tree["RB"], $depth+1);
|
|
}
|
|
}
|
|
|
|
function makequery($tree)
|
|
{
|
|
$a = array($tree["NODETYPE"]);
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "PHRASEA_KW_LAST":
|
|
if($tree["PNUM"] !== NULL)
|
|
$a[] = $tree["PNUM"];
|
|
break;
|
|
case "PHRASEA_KW_ALL":
|
|
break;
|
|
case "SIMPLE":
|
|
case "QSIMPLE":
|
|
// pas de tid, c'est un terme normal
|
|
if(is_array($tree["VALUE"]))
|
|
{
|
|
foreach($tree["VALUE"] as $k=>$v)
|
|
$a[] = $v;
|
|
}
|
|
else
|
|
{
|
|
$a[] = $tree["VALUE"];
|
|
}
|
|
break;
|
|
case "OPK":
|
|
if($tree["LB"] !== NULL)
|
|
$a[] = $this->makequery($tree["LB"]);
|
|
if($tree["RB"] !== NULL)
|
|
$a[] = $this->makequery($tree["RB"]);
|
|
break;
|
|
case "OPS":
|
|
if($tree["PNUM"] !== NULL)
|
|
$a[] = intval($tree["PNUM"]);
|
|
if($tree["LB"] !== NULL)
|
|
$a[] = $this->makequery($tree["LB"]);
|
|
if($tree["RB"] !== NULL)
|
|
$a[] = $this->makequery($tree["RB"]);
|
|
break;
|
|
}
|
|
return($a);
|
|
}
|
|
|
|
function maketree($depth, $inquote = false)
|
|
{
|
|
// printf("<!-- PARSING $depth -->\n\n");
|
|
$tree = null;
|
|
while($t = $this->nexttoken($inquote))
|
|
{
|
|
if($this->debug)
|
|
printf("got token %s of class %s\n", $t["VALUE"], $t["CLASS"]);
|
|
switch($t["CLASS"])
|
|
{
|
|
case "TOK_RP":
|
|
if($inquote)
|
|
{
|
|
// quand on est entre guillements les tokens perdent leur signification
|
|
$tree = $this->addtotree($tree, $t, $depth, $inquote);
|
|
if(!$tree)
|
|
return(null);
|
|
}
|
|
else
|
|
{
|
|
if($depth<=0) // ')' : retour de récursivité
|
|
{
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= sprintf("\\n");
|
|
$this->errmsg .= _('qparser:: erreur : trop de parentheses fermantes');
|
|
return(null);
|
|
}
|
|
return($tree);
|
|
}
|
|
break;
|
|
case "TOK_LP":
|
|
if($inquote)
|
|
{
|
|
// quand on est entre guillements les tokens perdent leur signification
|
|
$tree = $this->addtotree($tree, $t, $depth, $inquote);
|
|
if(!$tree)
|
|
return(null);
|
|
}
|
|
else // '(' : appel récursif
|
|
{
|
|
if(!$tree)
|
|
$tree = $this->maketree($depth+1);
|
|
else
|
|
{
|
|
if(($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK") && $tree["RB"]==null)
|
|
{
|
|
$tree["RB"] = $this->maketree($depth+1);
|
|
if(!$tree["RB"])
|
|
$tree = null;
|
|
}
|
|
else
|
|
{
|
|
// ici on applique l'opérateur par défaut
|
|
$tree = array("CLASS"=>"OPS",
|
|
"VALUE"=>$this->defaultop["VALUE"],
|
|
"NODETYPE"=>$this->defaultop["NODETYPE"],
|
|
"PNUM"=>$this->defaultop["PNUM"],
|
|
"DEPTH"=>$depth,
|
|
"LB"=>$tree,
|
|
"RB"=>$this->maketree($depth+1) );
|
|
}
|
|
}
|
|
if(!$tree)
|
|
return(null);
|
|
}
|
|
break;
|
|
case "TOK_VOID":
|
|
// ce token est entre guillemets : on le saute
|
|
break;
|
|
case "TOK_QUOTE":
|
|
// une expr entre guillemets est 'comme entre parenthéses',
|
|
// sinon "a b" OU "x y" -> (((a B0 b) OU x) B0 y) au lieu de
|
|
// "a b" OU "x y" -> ((a B0 b) OU (x B0 y))
|
|
if($inquote)
|
|
{
|
|
if($this->debug)
|
|
{
|
|
print("CLOSING QUOTE!\n");
|
|
}
|
|
// fermeture des guillemets -> retour de récursivité
|
|
if($depth<=0) // ')' : retour de récursivité
|
|
{
|
|
print("\nguillemets fermants en trop<br>");
|
|
return(null);
|
|
}
|
|
return($tree);
|
|
}
|
|
else
|
|
{
|
|
if($this->debug)
|
|
{
|
|
print("OPENING QUOTE!<br>");
|
|
}
|
|
// ouverture des guillemets -> récursivité
|
|
if(!$tree)
|
|
$tree = $this->maketree($depth+1, true);
|
|
else
|
|
{
|
|
if(($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK") && $tree["RB"]==null)
|
|
{
|
|
$tree["RB"] = $this->maketree($depth+1, true);
|
|
if(!$tree["RB"])
|
|
$tree = null;
|
|
}
|
|
else
|
|
{
|
|
// ici on applique l'opérateur par défaut
|
|
$tree = array("CLASS"=>"OPS",
|
|
"VALUE"=>$this->defaultop["VALUE"],
|
|
"NODETYPE"=>$this->defaultop["NODETYPE"],
|
|
"PNUM"=>$this->defaultop["PNUM"],
|
|
"DEPTH"=>$depth,
|
|
"LB"=>$tree,
|
|
"RB"=>$this->maketree($depth+1, true) );
|
|
}
|
|
}
|
|
if(!$tree)
|
|
return(null);
|
|
}
|
|
break;
|
|
default:
|
|
$tree = $this->addtotree($tree, $t, $depth, $inquote);
|
|
if($this->debug)
|
|
{
|
|
print("---- après addtotree ----\n");
|
|
var_dump($tree);
|
|
print("-------------------------\n");
|
|
}
|
|
if(!$tree)
|
|
return(null);
|
|
break;
|
|
}
|
|
}
|
|
if(($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK") && $tree["RB"]==null)
|
|
{
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= sprintf("\\n");
|
|
$this->errmsg .= sprintf(_('qparser::Formulation incorrecte, une valeur est attendu apres %s'),$tree["VALUE"]);
|
|
$tree = $tree["LB"];
|
|
}
|
|
return($tree);
|
|
}
|
|
|
|
function addtotree($tree, $t, $depth, $inquote)
|
|
{
|
|
if($this->debug)
|
|
{
|
|
printf("addtotree({tree}, \$t[CLASS]='%s', \$t[VALUE]='%s', \$depth=%d, inquote=%s)\n", $t["CLASS"], $t["VALUE"], $depth, $inquote?"true":"false");
|
|
print("---- avant addtotree ----\n");
|
|
var_dump($tree);
|
|
print("-------------------------\n");
|
|
}
|
|
|
|
if(!$t)
|
|
return($tree);
|
|
switch($t["CLASS"])
|
|
{
|
|
case "TOK_CONTEXT":
|
|
// if($this->debug)
|
|
// {
|
|
// printf("addtotree({tree}, \$t='%s', \$depth=%d, inquote=%s)\n", $t["VALUE"], $depth, $inquote?"true":"false");
|
|
// var_dump($tree);
|
|
// }
|
|
if($tree["CLASS"]=="SIMPLE" || $tree["CLASS"]=="QSIMPLE")
|
|
{
|
|
// un [xxx] suit un terme : il introduit un contexte
|
|
$tree["CONTEXT"] = $t["VALUE"];
|
|
}
|
|
elseif($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK")
|
|
{
|
|
if(!isset($tree["RB"]) || !$tree["RB"])
|
|
{
|
|
// un [xxx] peut suivre un opérateur, c'est un paramétre normalement numérique
|
|
$tree["PNUM"] = $t["VALUE"];
|
|
}
|
|
else
|
|
{
|
|
// [xxx] suit un terme déjé en branche droite ? (ex: a ou b[k])
|
|
if($tree["RB"]["CLASS"]=="SIMPLE" || $tree["RB"]["CLASS"]=="QSIMPLE")
|
|
$tree["RB"]["CONTEXT"] = $t["VALUE"];
|
|
else
|
|
{
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= "\\n";
|
|
$this->errmsg .= sprintf("le contexte [%s] ne peut suivre qu'un terme ou un opérateur<br/>", $t["VALUE"] );
|
|
return(null);
|
|
}
|
|
}
|
|
|
|
}
|
|
else
|
|
{
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= "\\n";
|
|
$this->errmsg .= sprintf("le contexte [%s] ne peut suivre qu'un terme ou un opérateur<br/>", $t["VALUE"] );
|
|
return(null);
|
|
}
|
|
return($tree);
|
|
break;
|
|
case "TOK_CMP":
|
|
// < > <= >= <> = : sont des opérateurs de comparaison
|
|
if(!$tree)
|
|
{
|
|
// printf("\nUne question ne peut commencer par '" . $t["VALUE"] . "'<br>");
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= "\\n";
|
|
$this->errmsg .= sprintf(_('qparser::erreur : une question ne peut commencer par %s'),$t["VALUE"]);
|
|
return(null);
|
|
}
|
|
if(($tree["CLASS"]=="OPS"|| $tree["CLASS"]=="OPK") && $tree["RB"]==null)
|
|
{
|
|
// printf("'" . $t["VALUE"] . "' ne peut suivre un opérateur<br>");
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= "\\n";
|
|
$this->errmsg .= sprintf(_('qparser::Formulation incorrecte, ne peut suivre un operateur : %s'),$t["VALUE"]);
|
|
return(null);
|
|
}
|
|
return(array("CLASS"=>"OPK", "VALUE"=>$t["VALUE"], "NODETYPE"=>$this->opk[$t["VALUE"]]["NODETYPE"], "PNUM"=>null, "DEPTH"=>$depth, "LB"=>$tree, "RB"=>null));
|
|
break;
|
|
case "TOK_WORD":
|
|
if($t["CLASS"]=="TOK_WORD" && isset($this->ops[$t["VALUE"]]) && !$inquote)
|
|
{
|
|
// ce mot est un opérateur phrasea
|
|
if(!$tree)
|
|
{
|
|
// printf("\n581 : Une question ne peut commencer par un opérateur<br>");
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= "\\n";
|
|
$this->errmsg .= sprintf(_('qparser::erreur : une question ne peut commencer par %s'),$t["VALUE"]);
|
|
return(null);
|
|
}
|
|
if(($tree["CLASS"]=="OPS" || $tree["CLASS"]=="OPK") && $tree["RB"]==null)
|
|
{
|
|
|
|
// printf("\n586 : Un opérateur ne peut suivre un opérateur<br>");
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= "\\n";
|
|
$this->errmsg .= sprintf(_('qparser::Formulation incorrecte, %s ne peut suivre un operateur'),$t["VALUE"]);
|
|
return(null);
|
|
}
|
|
$pnum = null;
|
|
if($this->ops[$t["VALUE"]]["CANNUM"])
|
|
{
|
|
// cet opérateur peut étre suivi d'un nombre ('near', 'before', 'after')
|
|
if($tn = $this->nexttoken())
|
|
{
|
|
if($tn["CLASS"]=="TOK_WORD" && is_numeric($tn["VALUE"]))
|
|
$pnum = (int)$tn["VALUE"];
|
|
else
|
|
$this->ungettoken($tn["VALUE"]);
|
|
}
|
|
}
|
|
return(array("CLASS"=>"OPS", "VALUE"=>$t["VALUE"], "NODETYPE"=>$this->ops[$t["VALUE"]]["NODETYPE"], "PNUM"=>$pnum, "DEPTH"=>$depth, "LB"=>$tree, "RB"=>null));
|
|
}
|
|
else
|
|
{
|
|
// ce mot n'est pas un opérateur
|
|
$pnum = null;
|
|
$nodetype = PHRASEA_KEYLIST;
|
|
if($t["CLASS"]=="TOK_WORD" && isset($this->spw[$t["VALUE"]]) && !$inquote)
|
|
{
|
|
// mais c'est un mot 'spécial' de phrasea ('last', 'all')
|
|
$type = $this->spw[$t["VALUE"]]["CLASS"];
|
|
$nodetype = $this->spw[$t["VALUE"]]["NODETYPE"];
|
|
if($this->spw[$t["VALUE"]]["CANNUM"])
|
|
{
|
|
// 'last' peut étre suivi d'un nombre
|
|
if($tn = $this->nexttoken())
|
|
{
|
|
if($tn["CLASS"]=="TOK_WORD" && is_numeric($tn["VALUE"]))
|
|
$pnum = (int)$tn["VALUE"];
|
|
else
|
|
$this->ungettoken($tn["VALUE"]);
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
//printf("sdfsdfsdfsd<br>");
|
|
$type = $inquote ? "QSIMPLE" : "SIMPLE";
|
|
}
|
|
return($this->addsimple($t, $type, $nodetype, $pnum, $tree, $depth));
|
|
}
|
|
break;
|
|
}
|
|
}
|
|
|
|
function addsimple($t, $type, $nodetype, $pnum, $tree, $depth)
|
|
{
|
|
$nok = 0;
|
|
$w = $t["VALUE"];
|
|
if($w != "?" && $w != "*") // on laisse passer les 'isolés' pour les traiter plus tard comme des mots vides
|
|
{
|
|
for($i=0; $i<strlen($w); $i++)
|
|
{
|
|
$c = substr($w, $i, 1);
|
|
if($c=="?" || $c=="*")
|
|
{
|
|
if($nok < GV_min_letters_truncation)
|
|
{
|
|
if($this->errmsg!="")
|
|
$this->errmsg .= sprintf("\\n");
|
|
$this->errmsg .= _('qparser:: Formulation incorrecte, necessite plus de caractere : ') ."<br>". GV_min_letters_truncation;
|
|
return(null);
|
|
}
|
|
// $nok = 0;
|
|
}
|
|
else
|
|
$nok++;
|
|
}
|
|
}
|
|
if(!$tree)
|
|
{
|
|
return(array("CLASS"=>$type, "NODETYPE"=>$nodetype, "VALUE"=>array($t["VALUE"]), "PNUM"=>$pnum, "DEPTH"=>$depth));
|
|
}
|
|
switch($tree["CLASS"])
|
|
{
|
|
case "SIMPLE":
|
|
case "QSIMPLE":
|
|
if($type == "SIMPLE" || $type == "QSIMPLE")
|
|
$tree["VALUE"][] = $t["VALUE"];
|
|
else
|
|
{
|
|
$tree = array("CLASS"=>"OPS",
|
|
"VALUE"=>"et",
|
|
"NODETYPE"=>PHRASEA_OP_AND,
|
|
"PNUM"=>null,
|
|
"DEPTH"=>$depth,
|
|
"LB"=>$tree,
|
|
"RB"=>array("CLASS"=>$type,
|
|
"NODETYPE"=>$nodetype,
|
|
"VALUE"=>array($t["VALUE"]),
|
|
"PNUM"=>$pnum,
|
|
"DEPTH"=>$depth));
|
|
}
|
|
return($tree);
|
|
case "OPS":
|
|
case "OPK":
|
|
if($tree["RB"]==null)
|
|
{
|
|
$tree["RB"] = array("CLASS"=>$type, "NODETYPE"=>$nodetype, "VALUE"=>array($t["VALUE"]), "PNUM"=>$pnum, "DEPTH"=>$depth);
|
|
return($tree);
|
|
}
|
|
else
|
|
{
|
|
if(($tree["RB"]["CLASS"] == "SIMPLE" || $tree["RB"]["CLASS"] == "QSIMPLE") && $tree["RB"]["DEPTH"] == $depth)
|
|
{
|
|
$tree["RB"]["VALUE"][] = $t["VALUE"];
|
|
return($tree);
|
|
}
|
|
if(($tree["RB"]["CLASS"] == "PHRASEA_KW_LAST" || $tree["RB"]["CLASS"] == "PHRASEA_KW_ALL") && $tree["RB"]["DEPTH"] == $depth)
|
|
{
|
|
$tree["RB"] = array("CLASS"=>"OPS",
|
|
"VALUE"=>"et",
|
|
"NODETYPE"=>PHRASEA_OP_AND,
|
|
"PNUM"=>null,
|
|
"DEPTH"=>$depth,
|
|
"LB"=>$tree["RB"],
|
|
"RB"=>array("CLASS"=>$type,
|
|
"NODETYPE"=>$nodetype,
|
|
"VALUE"=>array($t["VALUE"]),
|
|
"PNUM"=>$pnum,
|
|
"DEPTH"=>$depth));
|
|
return($tree);
|
|
}
|
|
return(array("CLASS"=>"OPS",
|
|
"VALUE"=>$this->defaultop["VALUE"],
|
|
"NODETYPE"=>$this->defaultop["NODETYPE"],
|
|
"PNUM"=>$this->defaultop["PNUM"],
|
|
"DEPTH"=>$depth,
|
|
"LB"=>$tree,
|
|
"RB"=>array("CLASS"=>$type, "NODETYPE"=>$nodetype, "VALUE"=>array($t["VALUE"]), "PNUM"=>$pnum, "DEPTH"=>$depth)
|
|
));
|
|
}
|
|
case "PHRASEA_KW_LAST":
|
|
case "PHRASEA_KW_ALL":
|
|
return(array("CLASS"=>"OPS",
|
|
"VALUE"=>"et",
|
|
"NODETYPE"=>PHRASEA_OP_AND,
|
|
"PNUM"=>null,
|
|
"DEPTH"=>$depth,
|
|
"LB"=>$tree,
|
|
"RB"=>array("CLASS"=>$type,
|
|
"NODETYPE"=>$nodetype,
|
|
"VALUE"=>array($t["VALUE"]),
|
|
"PNUM"=>$pnum,
|
|
"DEPTH"=>$depth)));
|
|
}
|
|
}
|
|
function ungettoken($s)
|
|
{
|
|
$this->phq = $s . " " . $this->phq;
|
|
}
|
|
function nexttoken($inquote=false)
|
|
{
|
|
global $endCharacters_utf8; // caractères terminaux dans 'lownodiacritics_utf8.php'
|
|
global $t_LowNoDiacritics;
|
|
// printf("<!-- phq=%s -->\n", $this->phq);
|
|
if($this->phq == "")
|
|
return(null);
|
|
switch($c = substr($this->phq, 0, 1))
|
|
{
|
|
case "<":
|
|
case ">":
|
|
if($inquote)
|
|
{
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_VOID", "VALUE"=>$c));
|
|
}
|
|
$c2 = $c . substr($this->phq, 1, 1);
|
|
if($c2=="<=" || $c2==">=" || $c2=="<>")
|
|
{
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 2, 99999, 'UTF-8'), 'UTF-8');
|
|
$c = $c2;
|
|
}
|
|
else
|
|
{
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
}
|
|
return(array("CLASS"=>"TOK_CMP", "VALUE"=>$c));
|
|
break;
|
|
case "=":
|
|
if($inquote)
|
|
{
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_VOID", "VALUE"=>$c));
|
|
}
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_CMP", "VALUE"=>"="));
|
|
break;
|
|
case ":":
|
|
if($inquote)
|
|
{
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_VOID", "VALUE"=>$c));
|
|
}
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_CMP", "VALUE"=>":"));
|
|
break;
|
|
case "(":
|
|
if($inquote)
|
|
{
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_VOID", "VALUE"=>$c));
|
|
}
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_LP", "VALUE"=>"("));
|
|
break;
|
|
case ")":
|
|
if($inquote)
|
|
{
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_VOID", "VALUE"=>$c));
|
|
}
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_RP", "VALUE"=>")"));
|
|
break;
|
|
case "[":
|
|
// if($inquote)
|
|
// {
|
|
// $this->phq = ltrim(substr($this->phq, 1));
|
|
// return(array("CLASS"=>"TOK_VOID", "VALUE"=>$c));
|
|
// }
|
|
// un '[' introduit un contexte qu'on lit jusqu'au ']'
|
|
$closeb = mb_strpos($this->phq, "]", 1, 'UTF-8');
|
|
if($closeb !== false)
|
|
{
|
|
$context = $this->mb_trim(mb_substr($this->phq, 1, $closeb-1, 'UTF-8'), 'UTF-8');
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, $closeb+1, 99999, 'UTF-8'), 'UTF-8');
|
|
}
|
|
else
|
|
{
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
$this->phq = "";
|
|
}
|
|
$context = noaccent_utf8($context, PARSED);
|
|
return(array("CLASS"=>"TOK_CONTEXT", "VALUE"=>$context));
|
|
break;
|
|
/*
|
|
case "]":
|
|
// if($inquote)
|
|
// {
|
|
// $this->phq = ltrim(substr($this->phq, 1));
|
|
// return(array("CLASS"=>"TOK_VOID", "VALUE"=>$c));
|
|
// }
|
|
$this->phq = ltrim(substr($this->phq, 1));
|
|
return(array("CLASS"=>"TOK_RB", "VALUE"=>"]"));
|
|
break;
|
|
*/
|
|
case "\"":
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, 1, 99999, 'UTF-8'), 'UTF-8');
|
|
return(array("CLASS"=>"TOK_QUOTE", "VALUE"=>"\""));
|
|
break;
|
|
default:
|
|
$l = mb_strlen($this->phq, 'UTF-8');
|
|
$t = "";
|
|
$c_utf8 = "";
|
|
for($i=0; $i<$l; $i++)
|
|
{
|
|
if(mb_strpos($endCharacters_utf8, ($c_utf8 = mb_substr($this->phq, $i, 1, 'UTF-8')))===false)
|
|
{
|
|
// $c = mb_strtolower($c);
|
|
// $t .= isset($this->noaccent[$c]) ? $this->noaccent[$c] : $c;
|
|
// printf("// '%s' ", $c);
|
|
$t .= isset($t_LowNoDiacritics["cmap"][$c_utf8]) ? $t_LowNoDiacritics["cmap"][$c_utf8] : $c_utf8;
|
|
// printf(" : '%s' \n", $c);
|
|
}
|
|
else
|
|
break;
|
|
}
|
|
if($c_utf8=="(" || $c_utf8==")" || $c_utf8=="[" || $c_utf8=="]" || $c_utf8=="=" || $c_utf8==":" || $c_utf8=="<" || $c_utf8==">" || $c_utf8=="\"") // ces caractéres sont des délimiteurs avec un sens, il faut les garder
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, $i, 99999, 'UTF-8'), 'UTF-8');
|
|
else
|
|
$this->phq = $this->mb_ltrim(mb_substr($this->phq, $i+1, 99999, 'UTF-8'), 'UTF-8'); // le délimiteur était une simple ponctuation, on le saute
|
|
if($t != "")
|
|
return(array("CLASS"=>"TOK_WORD", "VALUE"=>$t));
|
|
else
|
|
return(array("CLASS"=>"TOK_VOID", "VALUE"=>$t));
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
function txt2sql($t)
|
|
{
|
|
return( str_replace(array('\\', '\''), array('\\\\', '\'\''), $t) );
|
|
}
|