从请求的页面提取关键词

它可以从一个给定的URL检索网页提取出一些关键词

例如从代码珠玑的首页可以提取出类似下面图片中的关键词

2. if(!empty($_REQUEST["url"])){
4. include 'class.keywords.php';
6. $keywords = new keywordsugest();
7. $keywords->_lang = 'es';
8. $keywords->_encoding = 'iso-8859-1';
9. $keywords->_catego = 'telecom';
10. $keywords->_keyCount = 100; // is like the porcent %
11. $keywords->file($_REQUEST['url']);
13. #$keywords->readMetaKeyWords();
15. #$keywords->readHtmlKeyWords();
17. $keywords->readAll();
19. echo 'Keywords found :';
21. $i = 1;
23. foreach($keywords->get() as $word) echo $i++.". $word
    ";
24. }
25. //url例如：http://www.codepearl.com
26. echo "
    ";
29. ?>

复制代码

2. class keywordsugest{
4. var $_html = FALSE;
5. var $_keyCount = 5;
6. var $_keyWords = array();
7. var $_encoding = 'UTF-8';
8. var $_lang = 'es';
9. var $_catego = 'telecom';
10. var $_url = '';
12. /**
13. * # read meta keywords
14. *
15. */
16. public function readMetaKeyWords() {
18. if (! $this->_html) return;
20. preg_match('/"]*)\"?[\s]*[\/]?[\s]*>/is', $this->_html, $match);
22. //$tags = get_meta_tags($this->_url);
23. //echo $tags['keywords'];
26. if (count($match)) {
27. $this->_keyWords = array_unique(explode(',', preg_replace('/\s/i', ' ', mb_strtolower($match[1], $this->_encoding))));
28. }
29. }
31. /**
32. * strip tags
33. *
34. * @param mixed $string
35. */
36. private function rip_tags($string) {
38. // ----- remove HTML TAGs -----
39. $string = preg_replace ('/]*>/', ' ', $string);
41. /* // ----- remove control characters -----
42. $string = str_replace("\r", '', $string); // --- replace with empty space
43. $string = str_replace("\n", ' ', $string); // --- replace with space
44. $string = str_replace("\t", ' ', $string); // --- replace with space
45. */
46. // ----- remove multiple spaces -----
47. $string = trim(preg_replace('/ {2,}/', ' ', $string));
49. return $string;
51. }
53. /**
54. * # read keywords from page body or string
55. *
56. */
57. public function readHtmlKeyWords() {
59. if (! $this->_html) return;
61. if(!empty($this->_keyWords)){
62. $implo = implode(' ',$this->_keyWords);
63. $this->_html = $this->_html." ".$implo;
64. $this->_keyWords = array();
65. }
67. $this->_html = str_replace(' ',' ', $this->_html);
69. # remove unneeded parts
70. $toRemove = array('head', 'script', 'style', 'object', 'embed', 'noembed', 'applet', 'noframes', 'noscript');
72. foreach ($toRemove as $remove) $this->_html = preg_replace("/\.*?\/is", ' ', $this->_html);
74. # remove comments
75. $this->_html = preg_replace("/\/is", ' ', $this->_html);
77. # delete html tags
78. $this->_html = mb_strtolower($this->rip_tags($this->_html), $this->_encoding);
80. $this->_html = htmlspecialchars_decode($this->_html);
82. # decode encoded hmtl entities
83. $this->_html = html_entity_decode ($this->_html, ENT_COMPAT, $this->_encoding);
85. # break into words
86. $words = preg_split("/[\s]+|[\t]+|[\.]+|[\,]+|[\:]+|[\;]+|[\!]+|[\?]+|[\|]+/s", $this->_html, -1, PREG_SPLIT_NO_EMPTY);
88. if (count($words)) {
90. $frequency = array_count_values($words);
91. unset($frequency['']);
93. if (count($frequency)) {
95. # delete stop words and interpunctions
96. include('stopwords_'.$this->_lang.'.php');
97. include('glodic_'.$this->_catego.'_'.$this->_lang.'.php');
99. $punct = '~!@#$%^&*()_+|}{[];:\'\",<.>/?`-=\\';
101. foreach (array_keys($frequency) as $word) {
102. if ( (in_array($word, $stopWords)) or (strspn($word, $punct) == strlen($word)) ){ unset($frequency[$word]); }
103. }
105. $max = max($frequency);
106. $count = count($frequency);
107. $tot = round(($max * 100) / $count);
108. $tot2 = round(($this->_keyCount * 100) / $count);
109. if($tot > $count){$tot = $tot / 2;}
110. if($tot2 > $count){$tot = $tot / 2;}
111. $showmax = round(($tot + $tot2) / 2);
113. foreach (array_keys($frequency) as $word) {
114. if ( in_array($word, $glodic) ){$frequency[$word] = $frequency[$word] + $showmax; }
115. }
117. # sort by frequency
118. arsort($frequency, SORT_NUMERIC);
120. # add them to keyword array
121. $i = 0;
123. foreach ($frequency as $word=>$count) {
125. if ( (! in_array($word, $this->_keyWords)) &&
126. (! is_numeric($word)) &&
127. (! empty($word)) ) {
128. $this->_keyWords[] = (string)$word;
130. $i++;
132. if ($i == $showmax) break;
133. }
134. }
135. }
136. }
137. }
140. /**
141. * change the encoding from default utf-8
142. *
143. * @param mixed $enc
144. */
145. private function encoding($enc = FALSE) {
147. if ($enc) $this->_encoding = $enc;
148. }
151. /**
152. * # reads from file or url
153. *
154. * @param mixed $fileUrl
155. */
156. public function file($fileUrl = FALSE) {
158. if ($fileUrl){ $this->_html = @file_get_contents($fileUrl);
159. $this->_url = $fileUrl;
160. }
161. }
165. /**
166. * # define html as string
167. *
168. * @param mixed $page
169. */
170. public function html($page = FALSE) {
172. if ($page) $this->_html = $page;
173. }
176. /**
177. * # reads both meta keywords and from body
178. *
179. */
180. public function readAll() {
182. if ($this->_html !== FALSE) {
184. $this->readMetaKeyWords();
186. $this->readHtmlKeyWords();
187. }
189. $this->_keyWords = array_unique($this->_keyWords);
190. }
193. /**
194. * # returns keywords as array
195. *
196. */
197. public function get() {
199. return $this->_keyWords;
200. }
201. }
203. ?>

复制代码