WebSVN – eFlore/Applications.cel – Diff – /branches/v1.5-cisaille/jrest/services/NameParser.php


<?php
 
/**
 * Taxamatch-Webservice PHP v1.0.0
 * @author Michael Giddens
 * @link http://www.silverbiology.com
 */
 
 /* Adapation par David Delon Decembre 2010 : gestion sous espece
 */
 
 
	/**
	 * Class NameParser
	 * Used to convert a string to a standarized format.
	 */
	class NameParser {
 
		/**
		 * Whether to debug or nor
		 * @var bool|integer
		 */
		public $debug_flag;
 
 
 
		/**
		 * Constructor 
		 */
		public function __construct( ) {
		}
 
		/**
		 * Sets value to the method property
		 * @param mixed $name class property name
		 * @param mixed $value class property value
		 */
		public function set($name,$value) {
			$this->$name = $value;
		}
 
 
		/**
		 * Reduce Spaces
		 * This will reduce the string to only allow once space between characters
		 * @param string $str : string to reduce space
		 * @return string : string with only once space between characters
		 */
		private function reduce_spaces( $str ) {
		
			$str = preg_replace("/ {2,}/", ' ', $str );
			$str = trim( $str );
			
			return( $str );
		}
 
 
		/**
		 * Function: parse_auth
		 * Purpose: Produce a parsed version of authority of a taxon name
		 * @author Tony Rees (Tony.Rees@csiro.au)
		 * Date created: March 2008
		 * Inputs: authority string as str
		 * Remarks:
		 *  (1) Performs authority expension of known abbreviated authornames using
		 *   table "auth_abbrev_test1" (must be available and populated with relevant content)
		 *  (2) Recognises "and", "et", "&" as equivalents (special case for "et al.") - all parsed to ampersand
		 *  (3) Recognises (e.g.) "Smith 1980" and "Smith, 1980" as equivalents - comma is removed in these cases
		 *  (4) Recognises (e.g.) "F. J. R. Taylor, 1980" and "F.J.R. Taylor, 1980" as equivalents -
		 *      extra space after full stops is ignored in these cases
		 *  (5) Returns uppercase string, diacritical marks intact
		 *
		 * @param string $str : authority string
		 * @param integer $upcase : convert to uppercase if $upcase = 1
		 * @return string : parsed author string
		 */
		public function parse_auth( $str, $upcase=1 ) {
 
			$this->debug['parse_auth'][] = "1";
			$temp = $str = trim($str);
			
  		if ( ($str == NULL) || ($str == '') ) {
				$this->debug['parse_auth'][] = "1a";
		    return '';
			}
 
			if ( ( $temp == null ) || ( $temp == '') ) {
				$this->debug['parse_auth'][] = "2a";
				return('');
			} else {
 
				$this->debug['parse_auth'][] = "2b";
			
				// add space after full stops, except at end (NB, will also add spece before some close brackets)
				$temp = rtrim( str_replace('.', '. ', $temp) );
				$this->debug['parse_auth'][] = "4 (temp:$temp)";
				
				//normalise "et", "and" to ampersand (et al. is a special case)
//				if ( $temp like '% et al%' ) {
				if ( ereg(' et al', $temp) ) {
					$temp = str_replace(' et al','zzzzz', $temp);
					$this->debug['parse_auth'][] = "4a (temp:$temp)";
				}
				
//				$temp = str_replace(temp,' et ',' '||'&'||' ');
//				$temp = replace(temp,' and ',' '||'&'||' ');
				$temp = str_replace(' et ',' & ', $temp );
				$temp = str_replace(' and ',' & ', $temp );
//				if ( $temp like '%zzzzz%' ) {
//				if ( ereg('zzzzz', $temp) ) {
				$temp = str_replace('zzzzz',' et al', $temp);
//				}
 
				$this->debug['parse_auth'][] = "5 (temp:$temp)";
				
				//remove commas before dates (only)
				//	like '%, 17%'
				if ( ereg(', 17', $temp) ) {
					$temp = str_replace(', 17',' 17', $temp);
					$this->debug['parse_auth'][] = "5a (temp:$temp)";
				}
				
				//	like '%, 18%'
				if ( ereg(', 18', $temp) ) {
					$temp = str_replace(', 18',' 18', $temp);
					$this->debug['parse_auth'][] = "5b (temp:$temp)";
				}
				
				//	like '%, 19%'
				if ( ereg(', 19', $temp) ) {
					$temp = str_replace(', 19',' 19', $temp);
					$this->debug['parse_auth'][] = "5c (temp:$temp)";
				}
				
				//	like '%, 20%'
				if ( ereg(', 20', $temp) ) {
					$temp = str_replace(', 20',' 20', $temp);
					$this->debug['parse_auth'][] = "5d (temp:$temp)";
				}
				
				// reduce multiple internal spaces to single space
				$temp = $this->reduce_spaces( $temp );
				
				//	like '% -%'
				$temp = str_replace(' -', '-', $temp);
 
				$this->debug['parse_auth'][] = "6 (temp:$temp)";
				
				$elapsed_chars = '';
				foreach( explode(' ', $temp) as $this_word ) {
				
					$this->debug['parse_auth'][] = "7 (this_word:$this_word)";
					
					//	like '(%'
					if ( ereg('^\(', $this_word) ) {
						$elapsed_chars .= '(';
						$this_word = substr( $this_word, 1 );
						$this->debug['parse_auth'][] = "7a (this_word:$this_word) (elapsed_chars:$elapsed_chars)";
					}
 
					// Add back the word to the final translation
					$elapsed_chars .= $this_word . ' ';
					$this->debug['parse_auth'][] = "7c (this_word:$this_word) (elapsed_chars:$elapsed_chars)";
				}
				
				$elapsed_chars = $this->reduce_spaces( str_replace(' )', ')', $elapsed_chars) );
				
				return trim( $elapsed_chars ) ;
			}
 
		}
		
		/**
		 * Function: parse
		 * Purpose: Produces parsed version of an input string (scientific name components)
		 * @author Tony Rees (Tony.Rees@csiro.au)
		 * Date created: June 2007-November 2008
		 * Inputs: input string as str (this version presumes genus, genus+species, or
		 * genus+species+authority)
		 * Outputs: parsed version of input string, for match purposes
		 * Remarks:
		 *    (1) Removes known text elements e.g.
		 *      'aff.', 'cf.', 'subsp.', subgenera if enclosed in brackets, etc. as desired
		 *    (2) Removes accented and non A-Z characters other than full stops 
		 *       (in scientific name portions)
		 *    (3) Returns uppercase scientific name (genus + species only) 
		 *       plus unaltered (presumed) authority
		 *     examples;
		 *       Anabaena cf. flos-aquae Ralfs ex Born. et Flah. => ANABAENA FLOSAQUAE Ralfs 
		 *       ex Born. et Flah.
		 *       Abisara lemÈe-pauli => ABISARA LEMEEPAULI
		 *       Fuc/us Vesiculos2us => FUCUS VESICULOSUS
		 *       Buffo ignicolor LacÈpËde, 1788 => BUFFO IGNICOLOR LacÈpËde, 1788
		 *       Barbatia (Mesocibota) bistrigata (Dunker, 1866) => BARBATIA BISTRIGATA (Dunker, 1866)
		 *    (4) Thus version does not handle genus+author, or genus+species+infraspecies
		 *       (second" good" term is presumed to be species epithet, anything after is 
		 *       considered to be start of the authority), however could be adapted further as required
         *         and actually it was done in this version for Tela Botanica
		 *    (5) There is a separate function "parse_auth" for normalizing authorities when required
		 *      (e.g. for authority comparisons)
		 *
		 * @param string $str : input string ( genus, genus+species, or genus+species+authority )
		 * @return string : parsed string
		 */
		public function parse( $str = NULL ) {
			
			unset($this->debug['parse']);
 
 
			$temp = '';
			$first_str_part = NULL;
			$second_str_part = NULL;
			$temp_genus = '';
			$temp_species = '';
			$temp_genus_species = '';
			$temp_authority = '';
			$temp_infra = '';
			
			$this->debug['parse'][] = "1";
 
			if ( ($str == NULL) || ( trim($str) == '') ) {
				$this->debug[] = "N1a<br>";
				return '';
			} else {
				//	trim any leading, trailing spaces or line feeds
				$temp = trim( $str );
				$this->debug['parse'][] = "1b";
			}
 
			if ( $temp == NULL || $temp == '') {
				$this->debug['parse'][] = "2a";
				return '';
			} else {
				$this->debug['parse'][] = "2b";
 
				// replace any HTML ampersands
				$set = array('%', '&', 'amp;%', 'AMP;%');
				$temp = str_replace( $set, '&', $temp );
 
				$this->debug['parse'][] = "2b1 (temp:$temp)";
 
				// remove any content in angle brackets (e.g. html tags - <i>, </i>, etc.)
				$html_pattern = "(\<(/?[^\>]+)\>)";
//? This should not just handle html tags but all <*>				
				$temp = preg_replace( $html_pattern, '', $temp);
				$this->debug['parse'][] = "2b2 (temp:$temp)";
 
				// if second term (only) is in round brackets, presume it is a subgenus or a comment and remove it
				// examples: Barbatia (Mesocibota) bistrigata (Dunker, 1866) => Barbatia bistrigata (Dunker, 1866)
				// Barbatia (?) bistrigata (Dunker, 1866) => Barbatia bistrigata (Dunker, 1866)
				// (obviously this will not suit genus + author alone, where first part of authorname is in brackets,
				// however this is very rare?? and in any case we are not supporting genus+authority in this version)
//if ( $temp like '% (%)%'
				$temp = preg_replace( "/ \(\w*\W*\)/", '', $temp, 1 );
//? Not sure if this will catch if				
				$this->debug['parse'][] = "2b3 (temp:$temp)";
 
				// if second term (only) is in square brackets, presume it is a comment and remove it
				// example: Aphis [?] ficus Theobald, [1918] => Aphis ficus Theobald, [1918]		
//if ( $temp like '% [%]%'
				$temp = preg_replace( "/ \[\w*\W*\]/", '', $temp, 1 );
//? Not sure if this will catch if				
				$this->debug['parse'][] = "2b4 (temp:$temp)";
 
				// drop indicators of questionable id's - presume all are lowercase for now (could extend as needed)
				$temp = preg_replace( "/ cf /", " ", $temp );
				$temp = preg_replace( "/ cf\. /", " ", $temp );
				$temp = preg_replace( "/ near /", " ", $temp );
				$temp = preg_replace( "/ aff\. /", " ", $temp );
				$temp = preg_replace( "/ sp\. /", " ", $temp );
				$temp = preg_replace( "/ spp\. /", " ", $temp );
				$temp = preg_replace( "/ spp /", " ", $temp );
 
				$this->debug['parse'][] = "2b5 (temp:$temp)";
 
				// eliminate or close up any stray spaces introduced by the above
				$temp = $this->reduce_spaces( $temp );
 
				$this->debug['parse'][] = "2b6 (temp:$temp)";
 
				// now presume first element is genus, second (if present) is species, remainder
				//   (if present) is authority
				// look for genus name
				$ar = explode( " ", $temp, 2);
				if ( count( $ar ) ) {
					$temp_genus = $ar[0];
					$temp = @$ar[1];
				} else {
					$temp_genus = $temp;
					$temp = '';
				}
				
				$this->debug['parse'][] = "2b7 (temp_genus:$temp_genus) (temp:$temp)";
 
				// look for species epithet and authority
				$ar = explode( " ", $temp, 2);
				if ( count( $ar ) ) {
					$temp_species = $ar[0];
					$temp_authority = @$ar[1];
				} else {
					$temp_species = $temp;
					$temp_authority = '';
				}
               	// look for subspecies
 
                $infras =array('subsp.','var.');
                
                $temp_infra_authority = '';
                $temp_infra_type = '';
                
                $temp_authority = preg_replace( "/ssp./", "subsp.", $temp_authority);
                $temp_authority = preg_replace( "/ssp /", "subsp.", $temp_authority);
                $temp_authority = preg_replace( "/subsp /", "subsp.", $temp_authority);
                $temp_authority = preg_replace( "/var /", "var.", $temp_authority);
 
                foreach ($infras as $infra) {
                    $pos = strpos($temp_authority, $infra);
                    if ($pos === false) {
                        continue;
                    }
                    else {
                        $temp_infra=substr($temp_authority,$pos+strlen($infra));
                        $temp_authority=substr($temp_authority,0,$pos);
                        $temp_infra=trim($temp_infra);
                        $temp_infra_type=$infra;
                        // look for infra epithet and authority
                        $ar = explode(" ", $temp_infra, 2);
                        if ( count( $ar ) ) {
                            $temp_infra = $ar[0];
                            $temp_infra_authority = @$ar[1];
                        }
                        break; // on s'arrete au premier trouve
                    }
                }
 
				$this->debug['parse'][] = "2b8 (temp_genus:$temp_genus) (temp_species:$temp_species) (temp_authority:$temp_authority) (temp_infra:$temp_infra) (temp_infra_authority:$temp_infra_authority) (temp:$temp)";
 
 
				// replace selected ligatures here (Genus names can contain √Ü, OE ligature)
				$temp_genus = str_replace( '√Ü', 'AE', $temp_genus);
				$temp_species = str_replace( '√Ü', 'AE', $temp_species);
				$temp_infra = str_replace( '√Ü', 'AE', $temp_infra );
				
 
				$this->debug['parse'][] = "2b9 (temp_genus:$temp_genus) (temp_species:$temp_species) (temp_authority:$temp_authority) (temp_infra:$temp_infra) (temp_infra_authority:$temp_infra_authority) (temp:$temp)";
 
                $temp_genus= trim($temp_genus);
				$temp_species= trim($temp_species);
				$temp_infra= trim($temp_infra );
 
				// reduce any new multiple internal spaces to single space, if present
                $temp_genus= $this->reduce_spaces( $temp_genus );
				$temp_species= $this->reduce_spaces( $temp_species );
				$temp_infra= $this->reduce_spaces( $temp_infra );
 
				$this->debug['parse'][] = "2b10 (temp_genus:$temp_genus) (temp_species:$temp_species) (temp_authority:$temp_authority) (temp_infra:$temp_infra) (temp_infra_authority:$temp_infra_authority) (temp:$temp)";
 
                if (isset($temp_authority) && ($temp_authority!='') ) {
                    $temp_authority=$this->parse_auth($temp_authority);
                }
 
                if (isset($temp_infra_authority) && ($temp_infra_authority!='') ) {
                    $temp_infra_authority=$this->parse_auth($temp_infra_authority);
                }
 
 
				$this->debug['parse'][] = "2b11 (temp_genus:$temp_genus) (temp_species:$temp_species) (temp_authority:$temp_authority) (temp_infra:$temp_infra) (temp_infra_authority:$temp_infra_authority) (temp:$temp)";
 
				return array("genus"=>$temp_genus, "species"=>$temp_species, "authority"=>$temp_authority, "infra"=>$temp_infra, "infra_authority"=>$temp_infra_authority, "infra_type"=>$temp_infra_type);
				
			}
			
		} // End NameParser
 
		
		
	} // End Class
 
?>
Subversion Repositories eFlore/Applications.cel

(root)/branches/v1.5-cisaille/jrest/services/NameParser.php @ 1138 – Rev 1080 → 1126
