权重计算，稍加修改亦可用于分词，词频统计，全文和spam检测等

效率非常客观，你要是改成其他用处那效率我就不保证了

1. /* vim: set expandtab tabstop=4 shiftwidth=4: */
2. // +------------------------------------------------------------------------
3. // Name : 权重计算
4. // Description: 稍加修改，亦可用于分词，词频统计，全文检索和垃圾检测
5. // Date : 2013/12/16 08:51
6. // Authors : latel
7. // +------------------------------------------------------------------------
8. //
10. /*外部调用示例*/
11. /*
12. $aItems = array(
13. 'chinaisbig',
14. 'whichisnot',
15. 'totalyrightforme',
16. );
17. $aTable = array(
18. 'china,is|small',
19. 'china,big|me',
20. 'china,is|big,which|not,me',
21. 'totaly|right,for,me',
22. );
24. $oWeight = new ttrie;
25. $oWeight->newItems($aItems);
26. $aResult = $oWeight->newTable($aTable);
27. */
29. class weight {
30. protected $aDict = array(array());
31. protected $aItems = array();
32. protected $sLastRule;
33. protected $aMatchs = array();
34. protected $aShow = array();
36. private function init() {
37. //清空记录的匹配表和输出结果
38. unset($this->aShow);
39. }
41. public function newItems($mItems) {
42. //导入新的项目
43. $this->aItems = (is_array($mItems))? $mItems: array($mItems);
44. $this->init();
45. }
47. public function newTable(array $aTable) {
48. //导入新的对照表，并生成字典
49. foreach($aTable as $iTableKey=>$sTableLine) {
50. $aTableLine = explode(',', str_replace('|', ',', $sTableLine));
51. $setter = function($v, $k, $paraMeter) {
52. $k1 = $paraMeter[0]; $oWeight = $paraMeter[1];
53. $oWeight->genDict($v, $k1);
54. };
55. array_walk($aTableLine, $setter, array($iTableKey, $this));
56. }
57. $this->init();
58. }
60. public function getShow($sRule = 'max') {
61. //获取最终的显示结果
62. if(empty($this->aItems) || empty($this->aDict))
63. return array();
64. if (empty($this->aShow) || $sRule != $this->sLastRule)
65. return $this->genShow($sRule);
66. return $this->aShow;
67. }
69. public function genShow($sRule) {
70. $aShow = array();
71. $aMatchs = array();
72. $getter = function($v, $k, $oWeight) use(&$aShow, &$aMatchs, $sRule) {
73. $t = array_count_values($oWeight->matchWord($v));
74. $aMatchs[] = $t;
75. switch ($sRule) {
76. case 'max':
77. $aShow[$k] = array_keys($t, max($t));
78. break;
79. }
80. };
81. array_walk($this->aItems, $getter, $this);
82. $this->aShow = $aShow;
83. $this->aMatchs = $aMatchs;
84. return $aShow;
85. }
87. private function genDict($mWord, $iKey = '') {
88. $iInsertPonit = count($this->aDict);
89. $iCur = 0; //当前节点号
90. foreach (str_split($mWord) as $iChar) {
91. if (isset($this->aDict[$iCur][$iChar])) {
92. $iCur = $this->aDict[$iCur][$iChar];
93. continue;
94. }
95. $this->aDict[$iInsertPonit] = array();
96. $this->aDict[$iCur][$iChar] = $iInsertPonit;
97. $iCur = $iInsertPonit;
98. $iInsertPonit++;
99. }
100. $this->aDict[$iCur]['acc'][] = $iKey;
102. }
104. function matchWord($sLine) {
105. $iCur = $iOffset = $iPosition = 0;
106. $sLine .= "\0";
107. $iLen = strlen($sLine);
108. $aReturn = array();
109. while($iOffset $sChar = $sLine{$iOffset};
110. if(isset($this->aDict[$iCur][$sChar])) {
111. $iCur = $this->aDict[$iCur][$sChar];
112. if(isset($this->aDict[$iCur]['acc'])) {
113. $aReturn = array_merge($aReturn, $this->aDict[$iCur]['acc']);
115. $iPosition = $iOffset + 1;
116. $iCur = 0;
117. }
118. } else {
119. $iCur = 0;
120. $iOffset = $iPosition;
121. $iPosition = $iOffset + 1;
122. }
123. ++$iOffset;
124. }
125. return $aReturn;
126. }
127. }
129. ?>

复制代码