You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
94 lines
3.1 KiB
94 lines
3.1 KiB
<?php |
|
|
|
class Lingua_Stem_Ru { |
|
|
|
var $VERSION = "0.02"; |
|
var $Stem_Caching = 1; |
|
var $Stem_Cache = array(); |
|
var $VOWEL = '/аеиоуыэюя/'; |
|
var $PERFECTIVEGROUND = '/((ив|ивши|ившись|ыв|ывши|ывшись)|((?<=[ая])(в|вши|вшись)))$/'; |
|
var $REFLEXIVE = '/(с[яь])$/'; |
|
var $ADJECTIVE = '/(ее|ие|ые|ое|ими|ыми|ей|ий|ый|ой|ем|им|ым|ом|его|ого|ему|ому|их|ых|ую|юю|ая|яя|ою|ею)$/'; |
|
var $PARTICIPLE = '/((ивш|ывш|ующ)|((?<=[ая])(ем|нн|вш|ющ|щ)))$/'; |
|
var $VERB = '/((ила|ыла|ена|ейте|уйте|ите|или|ыли|ей|уй|ил|ыл|им|ым|ен|ило|ыло|ено|ят|ует|уют|ит|ыт|ены|ить|ыть|ишь|ую|ю)|((?<=[ая])(ла|на|ете|йте|ли|й|л|ем|н|ло|но|ет|ют|ны|ть|ешь|нно)))$/'; |
|
var $NOUN = '/(а|ев|ов|ие|ье|е|иями|ями|ами|еи|ии|и|ией|ей|ой|ий|й|иям|ям|ием|ем|ам|ом|о|у|ах|иях|ях|ы|ь|ию|ью|ю|ия|ья|я)$/'; |
|
var $RVRE = '/^(.*?[аеиоуыэюя])(.*)$/'; |
|
var $DERIVATIONAL = '/[^аеиоуыэюя][аеиоуыэюя]+[^аеиоуыэюя]+[аеиоуыэюя].*(?<=о)сть?$/'; |
|
|
|
function s(&$s, $re, $to) |
|
{ |
|
$orig = $s; |
|
$s = preg_replace($re, $to, $s); |
|
return $orig !== $s; |
|
} |
|
|
|
function m($s, $re) |
|
{ |
|
return preg_match($re, $s); |
|
} |
|
|
|
function stem_word($word) |
|
{ |
|
$word = mb_strtolower($word); |
|
//$word = strtr($word, 'ё', 'е'); |
|
# Check against cache of stemmed words |
|
if ($this->Stem_Caching && isset($this->Stem_Cache[$word])) { |
|
return $this->Stem_Cache[$word]; |
|
} |
|
$stem = $word; |
|
do { |
|
if (!preg_match($this->RVRE, $word, $p)) break; |
|
$start = $p[1]; |
|
$RV = $p[2]; |
|
if (!$RV) break; |
|
|
|
# Step 1 |
|
if (!$this->s($RV, $this->PERFECTIVEGROUND, '')) { |
|
$this->s($RV, $this->REFLEXIVE, ''); |
|
|
|
if ($this->s($RV, $this->ADJECTIVE, '')) { |
|
$this->s($RV, $this->PARTICIPLE, ''); |
|
} else { |
|
if (!$this->s($RV, $this->VERB, '')) |
|
$this->s($RV, $this->NOUN, ''); |
|
} |
|
} |
|
|
|
# Step 2 |
|
$this->s($RV, '/и$/', ''); |
|
|
|
# Step 3 |
|
if ($this->m($RV, $this->DERIVATIONAL)) |
|
$this->s($RV, '/ость?$/', ''); |
|
|
|
# Step 4 |
|
if (!$this->s($RV, '/ь$/', '')) { |
|
$this->s($RV, '/ейше?/', ''); |
|
$this->s($RV, '/нн$/', 'н'); |
|
} |
|
|
|
$stem = $start.$RV; |
|
} while(false); |
|
if ($this->Stem_Caching) $this->Stem_Cache[$word] = $stem; |
|
return $stem; |
|
} |
|
|
|
function stem_caching($parm_ref) |
|
{ |
|
$caching_level = @$parm_ref['-level']; |
|
if ($caching_level) { |
|
if (!$this->m($caching_level, '/^[012]$/')) { |
|
die(__CLASS__ . "::stem_caching() - Legal values are '0','1' or '2'. '$caching_level' is not a legal value"); |
|
} |
|
$this->Stem_Caching = $caching_level; |
|
} |
|
return $this->Stem_Caching; |
|
} |
|
|
|
function clear_stem_cache() |
|
{ |
|
$this->Stem_Cache = array(); |
|
} |
|
} |
|
|
|
?>
|