Last active
January 29, 2019 12:54
-
-
Save msaari/f9e8d74cfff6e2780368c600caf4c178 to your computer and use it in GitHub Desktop.
Suomenkielinen Porter stemmer
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| <?php | |
| add_filter( 'relevanssi_stemmer', 'suomi_stemmer' ); | |
| function suomi_stemmer( $merkkijono ) { | |
| $stemmer = new FinnishStemmer(); | |
| return $stemmer->process( $merkkijono ); | |
| } | |
| class FinnishStemmer { | |
| private static $vowels = array("a", "e", "i", "o", "u", "y", "å", "ä", "ö"); | |
| private static $long_vowels = array("aa", "ee", "ii", "oo", "uu", "yy", "åå", "ää", "öö"); | |
| private static $consonants = array("b", "c", "d", "f", "g", "h", "j", "k", "l", "m", "n", "p", "q", "r", "s", "t", "v", "w", "x", "z"); | |
| protected $word; | |
| protected $r1; | |
| protected $r2; | |
| protected $removed_in_step_3; | |
| private function find_r($in) { | |
| $word_vowels = array(); | |
| for ($i = 0; $i < mb_strlen($in); $i++) { | |
| $letter = mb_substr($in, $i, 1); | |
| if (in_array($letter, static::$vowels)) $word_vowels[] = $i; | |
| } | |
| $index = mb_strlen($in); | |
| $value = ""; | |
| foreach ($word_vowels as $pos) { | |
| $after = $pos + 1; | |
| $letter = mb_substr($in, $after, 1); | |
| if (!in_array($letter, static::$vowels)) { | |
| $index = $after + 1; | |
| $value = mb_substr($in, $after + 1); | |
| break; | |
| } | |
| } | |
| return array($index, $value); | |
| } | |
| private function is_suffix($suffix, $in, $preceeded = "", $negative_preceeded = "") { | |
| $length = mb_strlen($suffix) * -1; | |
| if (mb_substr($in, $length) == $suffix) { | |
| if (is_array($preceeded)) { | |
| foreach($preceeded as $preceed_candidate) { | |
| $candidate_length = mb_strlen($preceed_candidate); | |
| $preceeding_letter = mb_substr($this->word, $length - $candidate_length, $candidate_length); | |
| if ($preceeding_letter == $preceed_candidate) { | |
| return true; | |
| } | |
| } | |
| } else if (is_array($negative_preceeded)) { | |
| foreach($negative_preceeded as $preceed_candidate) { | |
| $candidate_length = mb_strlen($preceed_candidate); | |
| $preceeding_letter = mb_substr($this->word, $length - $candidate_length, $candidate_length); | |
| if ($preceeding_letter != $preceed_candidate) { | |
| return true; | |
| } | |
| } | |
| } else { | |
| return true; | |
| } | |
| } | |
| return false; | |
| } | |
| private function is_suffix_cv($suffix, $in) { | |
| $length = mb_strlen($suffix) * -1; | |
| if (mb_substr($in, $length) == $suffix) { | |
| $preceeding_letter = mb_substr($this->word, $length, 1); | |
| if (in_array($preceeding_letter, static::$vowels)) { | |
| $preceeding_letter = mb_substr($this->word, $length - 1, 1); | |
| if (in_array($preceeding_letter, static::$consonants)) { | |
| return true; | |
| } | |
| } | |
| } | |
| return false; | |
| } | |
| private function remove_suffix($suffix) { | |
| $this->word = mb_substr($this->word, 0, mb_strlen($suffix) * -1); | |
| $this->r1 = mb_substr($this->r1, 0, mb_strlen($suffix) * -1); | |
| $this->r2 = mb_substr($this->r2, 0, mb_strlen($suffix) * -1); | |
| } | |
| function step_1() { | |
| $r1 = $this->r1; | |
| $r2 = $this->r2; | |
| $preceeded = array_merge(static::$vowels, array('n', 't')); | |
| if ($this->is_suffix("kin", $r1, $preceeded)) { | |
| $this->remove_suffix("kin"); | |
| return true; | |
| } | |
| if ($this->is_suffix("kaan", $r1, $preceeded)) { | |
| $this->remove_suffix("kaan"); | |
| return true; | |
| } | |
| if ($this->is_suffix("kään", $r1, $preceeded)) { | |
| $this->remove_suffix("kään"); | |
| return true; | |
| } | |
| if ($this->is_suffix("han", $r1, $preceeded)) { | |
| $this->remove_suffix("han"); | |
| return true; | |
| } | |
| if ($this->is_suffix("hän", $r1, $preceeded)) { | |
| $this->remove_suffix("hän"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ko", $r1, $preceeded)) { | |
| $this->remove_suffix("ko"); | |
| return true; | |
| } | |
| if ($this->is_suffix("kö", $r1, $preceeded)) { | |
| $this->remove_suffix("kö"); | |
| return true; | |
| } | |
| if ($this->is_suffix("pa", $r1, $preceeded)) { | |
| $this->remove_suffix("pa"); | |
| return true; | |
| } | |
| if ($this->is_suffix("pä", $r1, $preceeded)) { | |
| $this->remove_suffix("pä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("sti", $r2)) { | |
| $this->remove_suffix("sti"); | |
| return true; | |
| } | |
| return false; | |
| } | |
| function step_2() { | |
| $r1 = $this->r1; | |
| $negative_preceeded = array('k'); | |
| if ($this->is_suffix("si", $r1, false, $negative_preceeded)) { | |
| $this->remove_suffix("si"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ni", $r1)) { | |
| $this->remove_suffix("ni"); | |
| return true; | |
| } | |
| if ($this->is_suffix("nsa", $r1)) { | |
| $this->remove_suffix("nsa"); | |
| return true; | |
| } | |
| if ($this->is_suffix("nsä", $r1)) { | |
| $this->remove_suffix("nsä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("mme", $r1)) { | |
| $this->remove_suffix("mme"); | |
| return true; | |
| } | |
| if ($this->is_suffix("nne", $r1)) { | |
| $this->remove_suffix("nne"); | |
| return true; | |
| } | |
| return false; | |
| } | |
| private function step_3() { | |
| $r1 = $this->r1; | |
| $r2 = $this->r2; | |
| $preceeded = array('a'); | |
| if ($this->is_suffix("han", $r1, $preceeded)) { | |
| $this->remove_suffix("han"); | |
| return true; | |
| } | |
| $preceeded = array('e'); | |
| if ($this->is_suffix("hen", $r1, $preceeded)) { | |
| $this->remove_suffix("hen"); | |
| return true; | |
| } | |
| $preceeded = array('i'); | |
| if ($this->is_suffix("hin", $r1, $preceeded)) { | |
| $this->remove_suffix("hin"); | |
| return true; | |
| } | |
| $preceeded = array('o'); | |
| if ($this->is_suffix("hon", $r1, $preceeded)) { | |
| $this->remove_suffix("hon"); | |
| return true; | |
| } | |
| $preceeded = array('y'); | |
| if ($this->is_suffix("hyn", $r1, $preceeded)) { | |
| $this->remove_suffix("hyn"); | |
| return true; | |
| } | |
| $preceeded = array('ä'); | |
| if ($this->is_suffix("hän", $r1, $preceeded)) { | |
| $this->remove_suffix("hän"); | |
| return true; | |
| } | |
| $preceeded = array('ö'); | |
| if ($this->is_suffix("hön", $r1, $preceeded)) { | |
| $this->remove_suffix("hön"); | |
| return true; | |
| } | |
| $preceeded = array('ai', 'ei', 'ii', 'oi', 'ui', 'yi', 'äi', 'öi'); | |
| if ($this->is_suffix("siin", $r1, $preceeded)) { | |
| $this->remove_suffix("siin"); | |
| return true; | |
| } | |
| $preceeded = array('ai', 'ei', 'ii', 'oi', 'ui', 'yi', 'äi', 'öi'); | |
| if ($this->is_suffix("den", $r1, $preceeded)) { | |
| $this->remove_suffix("den"); | |
| return true; | |
| } | |
| $preceeded = array('ai', 'ei', 'ii', 'oi', 'ui', 'yi', 'äi', 'öi'); | |
| if ($this->is_suffix("tten", $r1, $preceeded)) { | |
| $this->remove_suffix("tten"); | |
| return true; | |
| } | |
| $preceeded = static::$long_vowels; | |
| if ($this->is_suffix("seen", $r1, $preceeded)) { | |
| $this->remove_suffix("seen"); | |
| return true; | |
| } | |
| $preceeded = array('e'); | |
| if ($this->is_suffix("tta", $r1, $preceeded)) { | |
| $this->remove_suffix("tta"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ttä", $r1, $preceeded)) { | |
| $this->remove_suffix("ttä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("sta", $r1)) { | |
| $this->remove_suffix("sta"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ssa", $r1)) { | |
| $this->remove_suffix("ssa"); | |
| return true; | |
| } | |
| if ($this->is_suffix("stä", $r1)) { | |
| $this->remove_suffix("stä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("lta", $r1)) { | |
| $this->remove_suffix("lta"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ltä", $r1)) { | |
| $this->remove_suffix("ltä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("lla", $r1)) { | |
| $this->remove_suffix("lla"); | |
| return true; | |
| } | |
| if ($this->is_suffix("llä", $r1)) { | |
| $this->remove_suffix("llä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("lle", $r1)) { | |
| $this->remove_suffix("lle"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ksi", $r1)) { | |
| $this->remove_suffix("ksi"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ine", $r1)) { | |
| $this->remove_suffix("ine"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ta", $r1)) { | |
| $this->remove_suffix("ta"); | |
| return true; | |
| } | |
| if ($this->is_suffix("tä", $r1)) { | |
| $this->remove_suffix("tä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("na", $r1)) { | |
| $this->remove_suffix("na"); | |
| return true; | |
| } | |
| if ($this->is_suffix("nä", $r1)) { | |
| $this->remove_suffix("nä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("n", $r1)) { | |
| $this->remove_suffix("n"); | |
| $preceeding = array_merge(static::$long_vowels, array('ie')); | |
| if (in_array(mb_substr($this->word, -2), $preceeding)) { | |
| $this->remove_suffix("n"); | |
| } | |
| return true; | |
| } | |
| if ($this->is_suffix_cv("a", $r1)) { | |
| $this->remove_suffix("a"); | |
| return true; | |
| } | |
| return false; | |
| } | |
| private function step_4() { | |
| $r2 = $this->r2; | |
| $negative_preceeded = array('po'); | |
| if ($this->is_suffix("mpi", $r2, false, $negative_preceeded)) { | |
| $this->remove_suffix("mpi"); | |
| return true; | |
| } | |
| if ($this->is_suffix("mpa", $r2, false, $negative_preceeded)) { | |
| $this->remove_suffix("mpa"); | |
| return true; | |
| } | |
| if ($this->is_suffix("mpä", $r2, false, $negative_preceeded)) { | |
| $this->remove_suffix("mpä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("mmi", $r2, false, $negative_preceeded)) { | |
| $this->remove_suffix("mmi"); | |
| return true; | |
| } | |
| if ($this->is_suffix("mma", $r2, false, $negative_preceeded)) { | |
| $this->remove_suffix("mma"); | |
| return true; | |
| } | |
| if ($this->is_suffix("mmä", $r2, false, $negative_preceeded)) { | |
| $this->remove_suffix("mmä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("impi", $r2)) { | |
| $this->remove_suffix("impi"); | |
| return true; | |
| } | |
| if ($this->is_suffix("impa", $r2)) { | |
| $this->remove_suffix("impa"); | |
| return true; | |
| } | |
| if ($this->is_suffix("impä", $r2)) { | |
| $this->remove_suffix("impä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("immi", $r2)) { | |
| $this->remove_suffix("immi"); | |
| return true; | |
| } | |
| if ($this->is_suffix("imma", $r2)) { | |
| $this->remove_suffix("imma"); | |
| return true; | |
| } | |
| if ($this->is_suffix("immä", $r2)) { | |
| $this->remove_suffix("immä"); | |
| return true; | |
| } | |
| if ($this->is_suffix("eja", $r2)) { | |
| $this->remove_suffix("eja"); | |
| return true; | |
| } | |
| if ($this->is_suffix("ejä", $r2)) { | |
| $this->remove_suffix("ejä"); | |
| return true; | |
| } | |
| } | |
| private function step_5() { | |
| $r1 = $this->r1; | |
| $r2 = $this->r2; | |
| if ($this->removed_in_step_3) { | |
| if ($this->is_suffix("i", $r1)) { | |
| $this->remove_suffix("i"); | |
| return true; | |
| } | |
| if ($this->is_suffix("j", $r1)) { | |
| $this->remove_suffix("j"); | |
| return true; | |
| } | |
| } | |
| else { | |
| $preceeded = static::$vowels; | |
| if ($this->is_suffix("t", $r1, $preceeded)) { | |
| $this->remove_suffix("t"); | |
| if ($this->is_suffix("imma", $r2)) { | |
| $this->remove_suffix("imma"); | |
| } | |
| $negative_preceeded = array('po'); | |
| if ($this->is_suffix("mma", $r2, false, $negative_preceeded)) { | |
| $this->remove_suffix("mma"); | |
| } | |
| return true; | |
| } | |
| } | |
| } | |
| private function step_6() { | |
| $ending = mb_substr($this->r1, -2); | |
| if (in_array($ending, static::$long_vowels)) { | |
| $this->word = mb_substr($this->word, 0, -1); | |
| $this->r1 = mb_substr($this->r1, 0, -1); | |
| $this->r2 = mb_substr($this->r2, 0, -1); | |
| } | |
| $last = mb_substr($this->r1, -1); | |
| if (in_array($last, array('a', 'e', 'i', 'ä'))) { | |
| $second_to_last = mb_substr($this->r1, -2, 1); | |
| if (in_array($second_to_last, static::$consonants)) { | |
| $this->word = mb_substr($this->word, 0, -1); | |
| $this->r1 = mb_substr($this->r1, 0, -1); | |
| $this->r2 = mb_substr($this->r2, 0, -1); | |
| } | |
| } | |
| $ending = mb_substr($this->r1, -2); | |
| if (in_array($ending, array('oj', 'uj'))) { | |
| $this->word = mb_substr($this->word, 0, -1); | |
| $this->r1 = mb_substr($this->r1, 0, -1); | |
| $this->r2 = mb_substr($this->r2, 0, -1); | |
| } | |
| $ending = mb_substr($this->r1, -2); | |
| if (in_array($ending, array('jo'))) { | |
| $this->word = mb_substr($this->word, 0, -1); | |
| $this->r1 = mb_substr($this->r1, 0, -1); | |
| $this->r2 = mb_substr($this->r2, 0, -1); | |
| } | |
| $last_consonant_found = false; | |
| for ($i = mb_strlen($this->word); $i > 0; $i--) { | |
| $letter = mb_substr($this->word, $i, 1); | |
| if (in_array($letter, static::$consonants)) { | |
| $prev_letter = mb_substr($this->word, $i - 1, 1); | |
| if ($prev_letter == $letter) { | |
| $this->word = mb_substr($this->word, 0, $i - 1) . mb_substr($this->word, $i); | |
| break; | |
| } | |
| else { | |
| break; | |
| } | |
| } | |
| } | |
| } | |
| private function r1r2() { | |
| list($r1_index, $this->r1) = $this->find_r($this->word); | |
| if ($this->r1) { | |
| list($r2_index, $this->r2) = $this->find_r($this->r1); | |
| } | |
| } | |
| function process($word) { | |
| $this->word = $word; | |
| $this->r1r2(); | |
| $this->step_1(); | |
| $this->step_2(); | |
| $removed = $this->step_3(); | |
| $this->removed_in_step_3 = $removed; | |
| $this->step_4(); | |
| $this->step_5(); | |
| $this->step_6(); | |
| ; | |
| return $this->word; | |
| } | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment