利用php实现中文分词类实例详解

这篇文章主要介绍了php实现的中文分词类,结合完整实例形式分析了php基于字符串的遍历、转换、运算等技巧实现中文分词功能的具体方法,需要的朋友可以参考下

本文实例讲述了php实现的中文分词类。分享给大家供大家参考，具体如下：

PHP5 和 MySQL 圣经

本书是全面讲述PHP与MySQL的经典之作，书中不但全面介绍了两种技术的核心特性，还讲解了如何高效地结合这两种技术构建健壮的数据驱动的应用程序。本书涵盖了两种技术新版本中出现的最新特性，书中大量实际的示例和深入的分析均来自于作者在这方面多年的专业经验，可用于解决开发者在实际中所面临的各种挑战。

下载

该中文分词类源码使用http://tools.jb51.net/code/jb51_php_format进行了格式化处理，便于阅读。具体代码如下：

class Segmentation {
  var $options = array('lowercase' => TRUE, 
  'segment_english' => FALSE);
  var $dict_name = 'Unknown';
  var $dict_words = array();
  function setLowercase($value) {
    if ($value) {
      $this->options['lowercase'] = TRUE;
    } else {
      $this->options['lowercase'] = FALSE;
    }
    return TRUE;
  }
  function setSegmentEnglish($value) {
    if ($value) {
      $this->options['segment_english'] = TRUE;
    } else {
      $this->options['segment_english'] = FALSE;
    }
    return TRUE;
  }
  function load($dict_file) {
    if (!file_exists($dict_file)) {
      return FALSE;
    }
    $fp = fopen($dict_file, 'r');
    $temp = fgets($fp, 1024);
    if ($temp === FALSE) {
      return FALSE;
    } else {
      if (strpos($temp, "	") !== FALSE) {
        list ($dict_type, $dict_name) = explode("	", trim($temp));
      } else {
        $dict_type = trim($temp);
        $dict_name = 'Unknown';
      }
      $this->dict_name = $dict_name;
      if ($dict_type !== 'DICT_WORD_W') {
        return FALSE;
      }
    }
    while (!feof($fp)) {
      $this->dict_words[rtrim(fgets($fp, 32))] = 1;
    }
    fclose($fp);
    return TRUE;
  }
  function getDictName() {
    return $this->dict_name;
  }
  function segmentString($str) {
    if (count($this->dict_words) === 0) {
      return FALSE;
    }
    $lines = explode("
", $str);
    return $this->_segmentLines($lines);
  }
  function segmentFile($filename) {
    if (count($this->dict_words) === 0) {
      return FALSE;
    }
    $lines = file($filename);
    return $this->_segmentLines($lines);
  }
  function _segmentLines($lines) {
    $contents_segmented = '';
    foreach ($lines as $line) {
      $contents_segmented .= $this->_segmentLine(rtrim($line)) . " 
";
    }
    do {
      $contents_segmented = str_replace(' ', ' ', $contents_segmented);
    }
    while (strpos($contents_segmented, ' ') !== FALSE);
    return $contents_segmented;
  }
  function _segmentLine($str) {
    $str_final = '';
    $str_array = array();
    $str_length = strlen($str);
    if ($str_length > 0) {
      if (ord($str{$str_length-1}) >= 129) {
        $str .= ' ';
      }
    }
    for ($i=0; $i<$str_length; $i++) {
      if (ord($str{$i}) >= 129) {
        $str_array[] = $str{$i} . $str{$i+1};
        $i++;
      } else {
        $str_tmp = $str{$i};
        for ($j=$i+1; $j<$str_length; $j++) {
          if (ord($str{$j}) < 129) {
            $str_tmp .= $str{$j};
          } else {
            break;
          }
        }
        $str_array[] = array($str_tmp);
        $i = $j - 1;
      }
    }
    $pos = count($str_array);
    while ($pos > 0) {
      $char = $str_array[$pos-1];
      if (is_array($char)) {
        $str_final_tmp = $char[0];
        if ($this->options['segment_english']) {
          $str_final_tmp = preg_replace("/([!"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~	]+)/", " $1 ", $str_final_tmp); 
$str_final_tmp = preg_replace("/([!"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~	])([!"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~	])/", " $1 $2 ", $str_final_tmp);
        }
        if ($this->options['lowercase']) {
          $str_final_tmp = strtolower($str_final_tmp);
        }
        $str_final = " $str_final_tmp$str_final";
        $pos--;
      } else {
        $word_found = 0;
        $word_array = array(0 => '');
        if ($pos < 4) {
          $word_temp = $pos + 1;
        } else {
          $word_temp = 5;
        }
        for ($i=1; $i<$word_temp; $i++) {
          $word_array[$i] = $str_array[$pos-$i] . $word_array[$i-1];
        }
        for ($i=($word_temp-1); $i>1; $i--) {
          if (array_key_exists($word_array[$i], $this->dict_words)) {
            $word_found = $i;
            break;
          }
        }
        if ($word_found) {
          $str_final = " $word_array[$word_found]$str_final";
          $pos = $pos - $word_found;
        } else {
          $str_final = " $char$str_final";
          $pos--;
        }
      }
    }
    return $str_final;
  }
}
?>