public function __construct($params)
{
$this->lang = (isset($params["lang"]) && $params["lang"] != '') ? $params["lang"] : 'en';
$this->skip_len = $params["skip_length"];
$this->pspell = (function_exists('pspell_config_create') && ($params["use_pspell"] == "Y"));
//$this->custom_spell = $params["use_custom_spell"] == "Y";
$this->custom_spell = false;
$this->pspell_mode = $params["mode"];
$this->dics_path = $this->checkDicPath();
$this->user_dics_path = $this->dics_path."/custom.pws";
$this->custom_dics_path = $this->dics_path.'/custom_dics/'.$this->lang.'_';
if($this->custom_spell)
{
$this->dic = array();
}
if ($this->pspell)
{
$pspell_config = pspell_config_create ($this->lang, null, null, 'utf-8');
pspell_config_ignore($pspell_config, $this->skip_len);
pspell_config_mode($pspell_config, $params["mode"]);
pspell_config_personal($pspell_config, $this->user_dics_path);
$this->pspell_link = pspell_new_config($pspell_config);
}
}
/**
* Creates a new Pspell spell checker
*
* @param string $language the language used by this spell checker. This
* should be a two-letter ISO 639 language code
* followed by an optional two digit ISO 3166
* country code separated by a dash or underscore.
* For example, 'en', 'en-CA' and 'en_CA' are
* valid languages.
* @param string $personal_wordlist optional. The filename of the personal
* wordlist for this spell checker. If not
* specified, no personal wordlist is
* used. The personal wordlist may contain
* spellings for words that are correct
* but are not in the regular dictionary.
*
* @throws NateGoSearchException if the Pspell extension is not available.
* @throws NateGoSearchtException if a dictionary in the specified language
* could not be loaded.
*/
public function __construct($language, $path_to_data = '', $repl_pairs = '', $personal_wordlist = '')
{
if (!extension_loaded('pspell')) {
throw new NateGoSearchException('The Pspell PHP extension is ' . 'required for NateGoSearchPSpellSpellChecker.');
}
$config = pspell_config_create($language, '', '', 'utf-8');
pspell_config_mode($config, PSPELL_FAST);
if ($path_to_data != '') {
pspell_config_data_dir($config, $path_to_data);
pspell_config_dict_dir($config, $path_to_data);
}
if ($repl_pairs != '') {
pspell_config_repl($config, $repl_pairs);
}
if ($personal_wordlist != '') {
pspell_config_personal($config, $personal_wordlist);
if (file_exists($personal_wordlist) && fileowner($personal_wordlist) == posix_getuid()) {
// update permissions (-rw-rw----)
chmod($personal_wordlist, 0666);
}
$this->personal_wordlist = $personal_wordlist;
}
$this->dictionary = pspell_new_config($config);
if ($this->dictionary === false) {
throw new NateGoSearchException(sprintf("Could not create Pspell dictionary with language '%s'.", $this->language));
}
$this->loadBlacklistedSuggestions();
}
function pspellConfig()
{
$pspell_config = pspell_config_create($this->lang);
pspell_config_ignore($pspell_config, $this->skip_len);
pspell_config_mode($pspell_config, $this->mode);
pspell_config_personal($pspell_config, $this->personal_path);
$this->pspell_link = pspell_new_config($pspell_config);
}
function init()
{
$link = pspell_config_create($this->langcode);
pspell_config_ignore($link, $this->ignore);
if ($this->modus == 0) {
pspell_config_mode($link, PSPELL_FAST);
} elseif ($this->modus == 2) {
pspell_config_mode($link, PSPELL_BAD_SPELLERS);
} else {
pspell_config_mode($link, PSPELL_NORMAL);
}
$this->resid = @pspell_new_config($link);
if (!$this->resid) {
$this->errormsg = 'Could not open dictionary "' . $this->langcode . '"';
}
}
function check($word)
{
$pspell_config = pspell_config_create("ru", "", "", "UTF-8");
pspell_config_mode($pspell_config, PSPELL_FAST);
$pspell_link = pspell_new_config($pspell_config);
if (!pspell_check($pspell_link, $word)) {
$arr = array();
$suggestions = pspell_suggest($pspell_link, $word);
foreach ($suggestions as $suggestion) {
array_push($arr, $suggestion);
}
$json = json_encode($arr);
} else {
$json = true;
}
echo $json;
}
/**
* Set aspell options.
*
*/
function pspell_new($lang = "en", $spelling = "", $jargon = "", $enc = "utf-8", $mode = 0)
{
global $__pspell;
$i = count($__pspell);
$__pspell[$i] = array_merge($__pspell[0], array("lang" => $lang, "variety" => $spelling, "jargon" => $jargon, "encoding" => $enc));
if ($mode) {
pspell_config_mode($i, $mode);
}
return $i;
}
static function getLibrary()
{
global $IP;
$pspell_config = pspell_config_create("en", 'american');
pspell_config_mode($pspell_config, PSPELL_FAST);
//no longer using the custom dictionary
//pspell_config_personal($pspell_config, $IP . wikiHowDictionary::DICTIONARY_LOC);
$pspell_link = pspell_new_config($pspell_config);
return $pspell_link;
}
function init_spell($type, $dict)
{
$pspell_config = pspell_config_create($dict);
pspell_config_mode($pspell_config, $type);
pspell_config_personal($pspell_config, $GLOBALS['FORUM_SETTINGS_PATH'] . "forum.pws");
pspell_config_ignore($pspell_config, 2);
define('__FUD_PSPELL_LINK__', pspell_new_config($pspell_config));
return true;
}
//Set whether users are allowed to enter custom text
// THIS OPTION IS CURRENTLY UNDER TESTING
$allowCustomInserts = false;
// If using a personal dictionary, set the path to it. Default is in the
// personal_dictionary subdirectory of the location of spell_checker.php.
$path_to_personal_dictionary = dirname(__FILE__) . "/personal_dictionary/personal_dictionary.txt";
//If pspell doesn't exist, then include the pspell wrapper for aspell.
if (!function_exists('pspell_suggest')) {
// Set the path to aspell if you need to use it.
define('ASPELL_BIN', '/usr/bin/aspell');
require_once "pspell_comp.php";
}
// Create and configure a link to the pspell module.
$pspell_config = pspell_config_create("es");
//$pspell_config = pspell_new("en");
pspell_config_mode($pspell_config, PSPELL_FAST);
if ($usePersonalDict) {
// Allows the use of a custom dictionary (Thanks to Dylan Thurston for this addition).
//pspell_config_personal($pspell_config, $path_to_personal_dictionary);
}
$pspell_link = pspell_new_config($pspell_config);
require_once "cpaint/cpaint2.inc.php";
//AJAX library file
$cp = new cpaint();
$cp->register('showSuggestions');
$cp->register('spellCheck');
$cp->register('switchText');
$cp->register('addWord');
$cp->start();
$cp->return_data();
/*************************************************************
function parseStringP($txt, $dialects)
{
global $errorstring, $parameters;
$dialect = "british";
if (isset($dialects[1]) && $dialects[1] != "") {
$dialect = $dialects[1];
$lang = $dialects[0];
}
//pspell parse initialisation
//$psp_conf = pspell_config_create ("en", $dialect);
$psp_conf = pspell_config_create($lang);
pspell_config_runtogether($psp_conf, false);
pspell_config_mode($psp_conf, PSPELL_NORMAL);
$psp = pspell_new_config($psp_conf);
//join the string so that it is easier to parse - does not matter when formatting because it is html
$txt = str_replace("\r\n", " ", $txt);
$txt = str_replace("\r", " ", $txt);
$txt = str_replace("\n", " ", $txt);
$jstext = "";
$jsarrindex = 0;
//split string by html tags
$regexp = "#(<[^>]*>)#";
$arr = preg_split($regexp, $txt, -1, PREG_SPLIT_OFFSET_CAPTURE | PREG_SPLIT_DELIM_CAPTURE);
foreach ($arr as $comp) {
$arr2[] = $comp[0];
}
//we now have the $arr2 - containing html tags and plain text portions
for ($i = 0; $i < count($arr2); $i++) {
$portion = $arr2[$i];
//the part
if (strstr($portion, "<a")) {
//we have a link
//$arr2[$i] = str_replace('href', 'href1', $arr2[$i]); //replace so when clicking on the content we do not redirect
}
if (!strstr($portion, "<") && $portion != "") {
//we do not work on html portions and empty text - which should be resolved in the preg_split function
$portion = str_replace(" ", " ", $portion);
//replace for parsing
//get distinct words from string portion
$words = explode(" ", $portion);
$arr2[$i] = $words;
for ($j = 0; $j < count($arr2[$i]); $j++) {
$word = $arr2[$i][$j];
//we check if the "word" is really a word
$cond = true;
//is it a link address ?
if (strstr($word, "http://")) {
$cond = false;
}
/*
//is it a number?
if (preg_match("/^[0-9\.%]*$/", $word)) {
$cond = false;
}
//does it begin with a capital letter ?
if (ucwords($word) == $word) {
//$cond = false;
}
*/
if ($cond) {
$regexp3 = "/([^a-zA-Z]*)([a-zA-Z']*)([^a-zA-Z]*)/";
//get the actual words - we can have strings like ",something.". We separate in 3 portions
preg_match($regexp3, $word, $mt);
$beforeword = $mt[1];
$inword = $mt[2];
if (addedWord($inword)) {
continue;
}
$afterword = $mt[3];
if (!pspell_check($psp, $inword)) {
$sugs = pspell_suggest($psp, $inword);
if (count($sugs) > 0) {
//build the option string
//$arr2[$i][$j] = $beforeword.'<span onclick="fillOptions(this)" options="'.implode('|', $sugs).'" class="havesuggestion">'.$inword.'</span>'.$afterword;
$arr2[$i][$j] = $beforeword . '<span badword=1>' . $inword . '</span>' . $afterword;
$jstext .= "\nwords[" . $jsarrindex . "] = new Object();";
$jstext .= "\nwords[" . $jsarrindex . "].status = 0;";
$jstext .= "\nwords[" . $jsarrindex . "].word = \"" . $word . "\";";
$jstext .= "\nwords[" . $jsarrindex . "].suggestions = new Array();";
$jsarrindex2 = 0;
foreach ($sugs as $suggestion) {
$jstext .= "\nwords[" . $jsarrindex . "].suggestions[" . $jsarrindex2 . "] = \"" . str_replace(array("\r", "\n"), array("", ""), $suggestion) . "\";";
$jsarrindex2++;
}
$jsarrindex++;
} else {
//build the option string with red so that we know there is no replacement
//$arr2[$i][$j] = $beforeword.'<span onclick="fillOptions(this)" options="'.implode('|', $sugs).'" class="nosuggestion">'.$inword.'</span>'.$afterword;
$arr2[$i][$j] = $beforeword . '<span badword=1>' . $inword . '</span>' . $afterword;
$jstext .= "\nwords[" . $jsarrindex . "] = new Object();";
$jstext .= "\nwords[" . $jsarrindex . "].status = 0;";
$jstext .= "\nwords[" . $jsarrindex . "].word = \"" . $word . "\";";
$jstext .= "\nwords[" . $jsarrindex . "].suggestions = new Array();";
$jsarrindex2 = 0;
foreach ($sugs as $suggestion) {
$jstext .= "\nwords[" . $jsarrindex . "].suggestions[" . $jsarrindex2 . "] = \"" . str_replace(array("\r", "\n"), array("", ""), $suggestion) . "\";";
$jsarrindex2++;
}
$jsarrindex++;
}
}
}
}
}
}
//implosion of the parts to reconstitute the string
$toret = "";
foreach ($arr2 as $portion) {
if (is_array($portion)) {
$toret .= implode(" ", $portion);
} else {
$toret .= $portion;
}
}
//$toret = str_replace("\"", """, $toret);
return array($toret, $jstext);
}
/**
* Konstrukor zapewniaj±cy ogóln± inicjalizacjê systemu przygotowywania danych:
* tokenizer, korekta ortograficzna, uzupe³nianie polskich znaków, wulgaryzmy.
*
* @param mysqli $dbconn Obiekt po³±czenia z baz± danych u¿ywany w podklasach.
* @param string $dictdir Folder ze s³ownikami, stoplistami itp.
* @param int $idc Identyfikator wykorzystywanego zestawu komentarzy.
* @param bool $copy_unknown Czy pozostawiaæ nierozpoznane wyrazy?
* @param array $options Parametry konkretnego klasyfikatora jako tab. asocjacyjna.
*/
function __construct($dbconn, $dictdir, $idc, $copy_unknown, $options = null)
{
$this->idc = $idc;
$this->copy_unknown = $copy_unknown;
$this->dbconn = $dbconn;
if (is_null(self::$validation)) {
self::$validation = new Validation($dictdir . '/vulgarism.txt');
}
if (is_null(self::$tokenizer)) {
self::$tokenizer = new Tokenizer($dictdir . '/stoplist.txt');
}
if (is_null(self::$fsaa)) {
self::$fsaa = new Fsaa($dictdir . '/lort_acc_full.fsa');
}
if (is_null(self::$fsal)) {
self::$fsal = new Fsal($dictdir . '/llems_full.fsa');
}
if (is_null(self::$pspell)) {
$pspell_config = pspell_config_create("pl");
// opcje zapewniaj±ce wiêksz± szybko¶æ dzia³ania aspell
pspell_config_ignore($pspell_config, 4);
pspell_config_mode($pspell_config, PSPELL_FAST);
pspell_config_runtogether($pspell_config, false);
self::$pspell = pspell_new_config($pspell_config);
}
}
