1: <?php
  2: /*
  3:  You may not change or alter any portion of this comment or credits
  4:  of supporting developers from this source code or any supporting source code
  5:  which is considered copyrighted (c) material of the original comment or credit authors.
  6: 
  7:  This program is distributed in the hope that it will be useful,
  8:  but WITHOUT ANY WARRANTY; without even the implied warranty of
  9:  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 10:  */
 11: /**
 12:  * @copyright       The XUUPS Project http://sourceforge.net/projects/xuups/
 13:  * @license         GNU GPL V2 or later (http://www.gnu.org/licenses/gpl-2.0.html)
 14:  * @package         Publisher
 15:  * @since           1.0
 16:  * @author          trabis <lusopoemas@gmail.com>
 17:  * @author          The SmartFactory <www.smartfactory.ca>
 18:  * @version         $Id$
 19:  */
 20: 
 21: include_once dirname(__DIR__) . '/include/common.php';
 22: 
 23: class PublisherMetagen
 24: {
 25:     /**
 26:      * @var Publisher
 27:      * @access public
 28:      */
 29:     public $publisher = null;
 30: 
 31:     /**
 32:      * @var \Xoops\Core\Text\Sanitizer
 33:      */
 34:     public $_myts;
 35: 
 36:     /**
 37:      * @var string
 38:      */
 39:     public $_title;
 40: 
 41:     /**
 42:      * @var string
 43:      */
 44:     public $_original_title;
 45: 
 46:     /**
 47:      * @var string
 48:      */
 49:     public $_keywords;
 50: 
 51:     /**
 52:      * @var string
 53:      */
 54:     public $_categoryPath;
 55: 
 56:     /**
 57:      * @var string
 58:      */
 59:     public $_description;
 60: 
 61:     /**
 62:      * @var int
 63:      *
 64:      */
 65:     public $_minChar = 4;
 66: 
 67:     /**
 68:      * @param string       $title
 69:      * @param string       $keywords
 70:      * @param string       $description
 71:      * @param bool         $categoryPath
 72:      */
 73:     public function __construct($title, $keywords = '', $description = '', $categoryPath = false)
 74:     {
 75:         $this->publisher = Publisher::getInstance();
 76:         $this->_myts = \Xoops\Core\Text\Sanitizer::getInstance();
 77:         $this->setCategoryPath($categoryPath);
 78:         $this->setTitle($title);
 79:         $this->setDescription($description);
 80:         if ($keywords == '') {
 81:             $keywords = $this->createMetaKeywords();
 82:         }
 83:         $this->setKeywords($keywords);
 84:     }
 85: 
 86:     /**
 87:      * @param string $title
 88:      */
 89:     public function setTitle($title)
 90:     {
 91:         $this->_title = $this->html2text($title);
 92:         $this->_original_title = $this->_title;
 93:         $titleTag = array();
 94:         $titleTag['module'] = $this->publisher->getModule()->getVar('name');
 95:         if (isset($this->_title) && ($this->_title != '') && (strtoupper($this->_title) != strtoupper($titleTag['module']))) {
 96:             $titleTag['title'] = $this->_title;
 97:         }
 98:         if (isset($this->_categoryPath) && ($this->_categoryPath != '')) {
 99:             $titleTag['category'] = $this->_categoryPath;
100:         }
101:         $ret = isset($titleTag['title']) ? $titleTag['title'] : '';
102:         if (isset($titleTag['category']) && $titleTag['category'] != '') {
103:             if ($ret != '') {
104:                 $ret .= ' - ';
105:             }
106:             $ret .= $titleTag['category'];
107:         }
108:         if (isset($titleTag['module']) && $titleTag['module'] != '') {
109:             if ($ret != '') {
110:                 $ret .= ' - ';
111:             }
112:             $ret .= $titleTag['module'];
113:         }
114:         $this->_title = $ret;
115:     }
116: 
117:     /**
118:      * @param string $keywords
119:      */
120:     public function setKeywords($keywords)
121:     {
122:         $this->_keywords = $keywords;
123:     }
124: 
125:     /**
126:      * @param string $categoryPath
127:      */
128:     public function setCategoryPath($categoryPath)
129:     {
130:         $categoryPath = $this->html2text($categoryPath);
131:         $this->_categoryPath = $categoryPath;
132:     }
133: 
134:     /**
135:      * @param string $description
136:      */
137:     public function setDescription($description)
138:     {
139:         $description = $this->html2text($description);
140:         $description = $this->purifyText($description);
141:         $this->_description = $description;
142:     }
143: 
144:     /**
145:      * Does nothing
146:      */
147:     public function createTitleTag()
148:     {
149:     }
150: 
151:     /**
152:      * @param int $maxWords
153:      *
154:      * @return string
155:      */
156:     public function createMetaDescription($maxWords = 30)
157:     {
158:         $description = $this->purifyText($this->_description);
159:         $description = $this->html2text($description);
160:         $words = explode(" ", $description);
161:         $ret = '';
162:         $i = 1;
163:         $wordCount = count($words);
164:         foreach ($words as $word) {
165:             $ret .= $word;
166:             if ($i < $wordCount) {
167:                 $ret .= ' ';
168:             }
169:             ++$i;
170:         }
171:         return $ret;
172:     }
173: 
174:     /**
175:      * @param string $text
176:      * @param int    $minChar
177:      *
178:      * @return array
179:      */
180:     public function findMetaKeywords($text, $minChar)
181:     {
182:         $keywords = array();
183:         $text = $this->purifyText($text);
184:         $text = $this->html2text($text);
185:         $originalKeywords = explode(" ", $text);
186:         foreach ($originalKeywords as $originalKeyword) {
187:             $secondRoundKeywords = explode("'", $originalKeyword);
188:             foreach ($secondRoundKeywords as $secondRoundKeyword) {
189:                 if (strlen($secondRoundKeyword) >= $minChar) {
190:                     if (!in_array($secondRoundKeyword, $keywords)) {
191:                         $keywords[] = trim($secondRoundKeyword);
192:                     }
193:                 }
194:             }
195:         }
196:         return $keywords;
197:     }
198: 
199:     /**
200:      * @return string
201:      */
202:     public function createMetaKeywords()
203:     {
204:         $keywords = $this->findMetaKeywords($this->_original_title . " " . $this->_description, $this->_minChar);
205:         $moduleKeywords = $this->publisher->getConfig('seo_meta_keywords');
206:         if ($moduleKeywords != '') {
207:             $moduleKeywords = explode(",", $moduleKeywords);
208:             $keywords = array_merge($keywords, array_map('trim', $moduleKeywords));
209:         }
210:         $ret = implode(',', $keywords);
211:         return $ret;
212:     }
213: 
214:     /**
215:      * Does nothing
216:      */
217:     public function autoBuildMeta_keywords()
218:     {
219:     }
220: 
221:     /**
222:      * Build Metatags
223:      */
224:     public function buildAutoMetaTags()
225:     {
226:         $this->_keywords = $this->createMetaKeywords();
227:         $this->_description = $this->createMetaDescription();
228:         //$this->_title = $this->createTitleTag();
229:     }
230: 
231:     /**
232:      * Creates meta tags
233:      */
234:     public function createMetaTags()
235:     {
236:         global $xoopsTpl, $xoTheme;
237:         if ($this->_keywords != '') {
238:             $xoTheme->addMeta('meta', 'keywords', $this->_keywords);
239:         }
240:         if ($this->_description != '') {
241:             $xoTheme->addMeta('meta', 'description', $this->_description);
242:         }
243:         if ($this->_title != '') {
244:             $xoopsTpl->assign('xoops_pagetitle', $this->_title);
245:         }
246:     }
247: 
248:     /**
249:      * Return true if the string is length > 0
250:      *
251:      * @credit psylove
252:      * @var string $string Chaine de caractère
253:      * @return boolean
254:      */
255:     static public function emptyString($var)
256:     {
257:         return (strlen($var) > 0);
258:     }
259: 
260:     /**
261:      * Create a title for the short_url field of an article
262:      *
263:      * @credit psylove
264:      *
265:      * @param string $title    title of the article
266:      * @param bool   $withExt  do we add an html extension or not
267:      *
268:      * @return string short url for article
269:      */
270:     static  function generateSeoTitle($title = '', $withExt = true)
271:     {
272:         // Transformation de la chaine en minuscule
273:         // Codage de la chaine afin d'éviter les erreurs 500 en cas de caractères imprévus
274:         $title = rawurlencode(strtolower($title));
275:         // Transformation des ponctuations
276:         //                 Tab     Space      !        "        #        %        &        '        (        )        ,        /        :        ;        <        =        >        ?        @        [        \        ]        ^        {        |        }        ~       .
277:         $pattern = array("/%09/", "/%20/", "/%21/", "/%22/", "/%23/", "/%25/", "/%26/", "/%27/", "/%28/", "/%29/", "/%2C/", "/%2F/", "/%3A/", "/%3B/", "/%3C/", "/%3D/", "/%3E/", "/%3F/", "/%40/", "/%5B/", "/%5C/", "/%5D/", "/%5E/", "/%7B/", "/%7C/", "/%7D/", "/%7E/", "/\./");
278:         $rep_pat = array("-", "-", "-", "-", "-", "-100", "-", "-", "-", "-", "-", "-", "-", "-", "-", "-", "-", "-", "-at-", "-", "-", "-", "-", "-", "-", "-", "-", "-");
279:         $title = preg_replace($pattern, $rep_pat, $title);
280:         // Transformation des caractères accentués
281:         //                  °        è        é        ê        ë        ç        à        â        ä        î        ï        ù        ü        û        ô        ö
282:         $pattern = array("/%B0/", "/%E8/", "/%E9/", "/%EA/", "/%EB/", "/%E7/", "/%E0/", "/%E2/", "/%E4/", "/%EE/", "/%EF/", "/%F9/", "/%FC/", "/%FB/", "/%F4/", "/%F6/");
283:         $rep_pat = array("-", "e", "e", "e", "e", "c", "a", "a", "a", "i", "i", "u", "u", "u", "o", "o");
284:         $title = preg_replace($pattern, $rep_pat, $title);
285:         $tableau = explode("-", $title); // Transforme la chaine de caractères en tableau
286:         $tableau = array_filter($tableau, array('PublisherMetagen', 'emptyString')); // Supprime les chaines vides du tableau
287:         $title = implode("-", $tableau); // Transforme un tableau en chaine de caractères séparé par un tiret
288:         if (sizeof($title) > 0) {
289:             if ($withExt) {
290:                 $title .= '.html';
291:             }
292:             return $title;
293:         }
294:         return '';
295:     }
296: 
297:     /**
298:      * @param string  $text
299:      * @param boolean $keyword
300:      *
301:      * @return string
302:      */
303:     public function purifyText($text, $keyword = false)
304:     {
305:         $text = str_replace('&nbsp;', ' ', $text);
306:         $text = str_replace('<br />', ' ', $text);
307:         $text = strip_tags($text);
308:         $text = html_entity_decode($text);
309:         $text = $this->_myts->undoHtmlSpecialChars($text);
310:         $text = str_replace(')', ' ', $text);
311:         $text = str_replace('(', ' ', $text);
312:         $text = str_replace(':', ' ', $text);
313:         $text = str_replace('&euro', ' euro ', $text);
314:         $text = str_replace('&hellip', '...', $text);
315:         $text = str_replace('&rsquo', ' ', $text);
316:         $text = str_replace('!', ' ', $text);
317:         $text = str_replace('?', ' ', $text);
318:         $text = str_replace('"', ' ', $text);
319:         $text = str_replace('-', ' ', $text);
320:         $text = str_replace('\n', ' ', $text);
321:         if ($keyword) {
322:             $text = str_replace('.', ' ', $text);
323:             $text = str_replace(',', ' ', $text);
324:             $text = str_replace('\'', ' ', $text);
325:         }
326:         $text = str_replace(';', ' ', $text);
327:         return $text;
328:     }
329: 
330:     /**
331:      * @param string $document
332:      *
333:      * @return mixed
334:      */
335:     public function html2text($document)
336:     {
337:         // PHP Manual:: function preg_replace
338:         // $document should contain an HTML document.
339:         // This will remove HTML tags, javascript sections
340:         // and white space. It will also convert some
341:         // common HTML entities to their text equivalent.
342:         // Credits : newbb2
343:         $search = array(
344:             "'<script[^>]*?>.*?</script>'si", // Strip out javascript<?
345:             "'<img.*?/>'si", // Strip out img tags
346:             "'<[\/\!]*?[^<>]*?>'si", // Strip out HTML tags<?
347:             "'([\r\n])[\s]+'", // Strip out white space
348:             "'&(quot|#34);'i", // Replace HTML entities
349:             "'&(amp|#38);'i",
350:             "'&(lt|#60);'i",
351:             "'&(gt|#62);'i",
352:             "'&(nbsp|#160);'i",
353:             "'&(iexcl|#161);'i",
354:             "'&(cent|#162);'i",
355:             "'&(pound|#163);'i",
356:             "'&(copy|#169);'i",
357:             //"'&#(\d+);'e"
358:         );
359:         // evaluate as php
360:         $replace = array(
361:             "",
362:             "",
363:             "",
364:             "\\1",
365:             "\"",
366:             "&",
367:             "<",
368:             ">",
369:             " ",
370:             chr(161),
371:             chr(162),
372:             chr(163),
373:             chr(169),
374:             //"chr(\\1)"
375:         );
376:         $text = preg_replace($search, $replace, $document);
377:         return $text;
378:     }
379: }
380: