plugins/af_sort_bayes/lib/class.naivebayesianstorage.php

   1 <?php
   2         /*
   3          ***** BEGIN LICENSE BLOCK *****
   4          This file is part of PHP Naive Bayesian Filter.
   5
   6          The Initial Developer of the Original Code is
   7          Loic d'Anterroches [loic_at_xhtml.net].
   8          Portions created by the Initial Developer are Copyright (C) 2003
   9          the Initial Developer. All Rights Reserved.
  10
  11          Contributor(s):
  12
  13          PHP Naive Bayesian Filter is free software; you can redistribute it
  14          and/or modify it under the terms of the GNU General Public License as
  15          published by the Free Software Foundation; either version 2 of
  16          the License, or (at your option) any later version.
  17
  18          PHP Naive Bayesian Filter is distributed in the hope that it will
  19          be useful, but WITHOUT ANY WARRANTY; without even the implied
  20          warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
  21          See the GNU General Public License for more details.
  22
  23          You should have received a copy of the GNU General Public License
  24          along with Foobar; if not, write to the Free Software
  25          Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  26
  27          Alternatively, the contents of this file may be used under the terms of
  28          the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
  29          in which case the provisions of the LGPL are applicable instead
  30          of those above.
  31
  32          ***** END LICENSE BLOCK *****
  33          */
  34
  35         /** Access to the storage of the data for the filter.
  36
  37          To avoid dependency with respect to any database, this class handle all the
  38          access to the data storage. You can provide your own class as long as
  39          all the methods are available. The current one rely on a MySQL database.
  40
  41          methods:
  42          - array getCategories()
  43          - bool  wordExists(string $word)
  44          - array getWord(string $word, string $categoryid)
  45
  46          */
  47         class NaiveBayesianStorage {
  48                 var $con = null;
  49                 var $owner_uid = null;
  50
  51                 function NaiveBayesianStorage($owner_uid) {
  52                         $this->con = Db::get();
  53                         $this->owner_uid = $owner_uid;
  54
  55                         return true;
  56                 }
  57
  58                 /** get the list of categories with basic data.
  59
  60                  @return array key = category ids, values = array(keys = 'probability', 'word_count')
  61                  */
  62                 function getCategories() {
  63                         $categories = array();
  64                         $rs = $this->con->query('SELECT * FROM ttrss_plugin_af_sort_bayes_categories');
  65
  66                         while ($this->con->fetch_assoc($rs)) {
  67                                 $categories[$rs['category_id']] = array('probability' => $rs['probability'],
  68                                         'word_count'  => $rs['word_count']
  69                                 );
  70
  71
  72                         }
  73
  74                         return $categories;
  75                 }
  76
  77                 /** see if the word is an already learnt word.
  78                  @return bool
  79                  @param string word
  80                  */
  81                 function wordExists($word) {
  82                         $rs = $this->con->query("SELECT * FROM ttrss_plugin_af_sort_bayes_wordfreqs WHERE word='" . $this->con->escape_string($word) . "'");
  83
  84                         return $this->con->num_rows($rs) != 0;
  85                 }
  86
  87                 /** get details of a word in a category.
  88                  @return array ('count' => count)
  89                  @param  string word
  90                  @param  string category id
  91                  */
  92                 function getWord($word, $category_id) {
  93                         $details = array();
  94
  95                         $rs = $this->con->query("SELECT * FROM ttrss_plugin_af_sort_bayes_wordfreqs WHERE word='" . $this->con->escape_string($word) . "' AND category_id='" . $this->con->escape_string($category_id) . "'");
  96
  97                         if ($this->con->num_rows($rs) == 0 ) {
  98                                 $details['count'] = 0;
  99                         }
 100                         else {
 101                                 $details['count'] = $rs['count'];
 102                         }
 103
 104                         return $details;
 105                 }
 106
 107                 /** update a word in a category.
 108                  If the word is new in this category it is added, else only the count is updated.
 109
 110                  @return bool success
 111                  @param string word
 112                  @param int    count
 113                  @paran string category id
 114                  */
 115                 function updateWord($word, $count, $category_id) {
 116                         $oldword = $this->getWord($word, $category_id);
 117
 118                         if (0 == $oldword['count']) {
 119                                 return $this->con->execute("INSERT INTO ttrss_plugin_af_sort_bayes_wordfreqs (word, category_id, count) VALUES ('" . $this->con->escape_string($word) . "', '" . $this->con->escape_string($category_id) . "', '" . $this->con->escape_string((int) $count) . "')");
 120                         }
 121                         else {
 122                                 return $this->con->execute("UPDATE ttrss_plugin_af_sort_bayes_wordfreqs SET count = count + " . (int) $count . " WHERE category_id = '" . $this->con->escape_string($category_id) . "' AND word = '" . $this->con->escape_string($word) . "'");
 123                         }
 124                 }
 125
 126                 /** remove a word from a category.
 127
 128                  @return bool success
 129                  @param string word
 130                  @param int  count
 131                  @param string category id
 132                  */
 133                 function removeWord($word, $count, $category_id) {
 134                         $oldword = $this->getWord($word, $category_id);
 135
 136                         if (0 != $oldword['count'] && 0 >= ($oldword['count'] - $count)) {
 137                                 return $this->con->execute("DELETE FROM ttrss_plugin_af_sort_bayes_wordfreqs WHERE word='" . $this->con->escape_string($word) . "' AND category_id='" . $this->con->escape_string($category_id) . "'");
 138                         }
 139                         else {
 140                                 return $this->con->execute("UPDATE ttrss_plugin_af_sort_bayes_wordfreqs SET count = count - " . (int) $count . " WHERE category_id = '" . $this->con->escape_string($category_id) . "' AND word = '" . $this->con->escape_string($word) . "'");
 141                         }
 142                 }
 143
 144                 /** update the probabilities of the categories and word count.
 145                  This function must be run after a set of training
 146
 147                  @return bool sucess
 148                  */
 149                 function updateProbabilities() {
 150                         // first update the word count of each category
 151                         $rs = $this->con->query("SELECT category_id, SUM(count) AS total FROM ttrss_plugin_af_sort_bayes_wordfreqs WHERE 1 GROUP BY category_id");
 152                         $total_words = 0;
 153
 154                         while ($this->con->fetch_assoc($rs)) {
 155                                 $total_words += $rs['total'];
 156
 157                         }
 158
 159                         $rs->moveStart();
 160
 161                         if ($total_words == 0) {
 162                                 $this->con->execute("UPDATE ttrss_plugin_af_sort_bayes_categories SET word_count=0, probability=0 WHERE 1");
 163
 164                                 return true;
 165                         }
 166
 167                         while ($this->con->fetch_assoc($rs)) {
 168                                 $proba = $rs['total'] / $total_words;
 169                                 $this->con->execute("UPDATE ttrss_plugin_af_sort_bayes_categories SET word_count=" . (int) $rs['total'] . ", probability=" . $proba . " WHERE category_id = '" . $rs['category_id'] . "'");
 170
 171                         }
 172
 173                         return true;
 174                 }
 175
 176                 /** save a reference in the database.
 177
 178                  @return bool success
 179                  @param  string reference if, must be unique
 180                  @param  string category id
 181                  @param  string content of the reference
 182                  */
 183                 function saveReference($doc_id, $category_id, $content) {
 184
 185                         return $this->con->execute("INSERT INTO ttrss_plugin_af_sort_bayes_references (id, category_id, content) VALUES ('" . $this->con->escape_string($doc_id) . "', '" . $this->con->escape_string($category_id) . "', '" . $this->con->escape_string($content) . "')");
 186                 }
 187
 188                 /** get a reference from the database.
 189
 190                  @return array  reference( category_id => ...., content => ....)
 191                  @param  string id
 192                  */
 193                 function getReference($doc_id) {
 194                         $ref = array();
 195                         $rs = $this->con->query("SELECT * FROM ttrss_plugin_af_sort_bayes_references WHERE id='" . $this->con->escape_string($doc_id) . "'");
 196
 197                         if ($this->con->num_rows($rs) == 0 ) {
 198                                 return $ref;
 199                         }
 200
 201                         $ref['category_id'] = $rs['category_id'];
 202                         $ref['content'] = $rs['content'];
 203                         $ref['id'] = $rs['id'];
 204
 205                         return $ref;
 206                 }
 207
 208                 /** remove a reference from the database
 209
 210                  @return bool sucess
 211                  @param  string reference id
 212                  */
 213                 function removeReference($doc_id) {
 214
 215                         return $this->con->execute("DELETE FROM ttrss_plugin_af_sort_bayes_references WHERE id='" . $this->con->escape_string($doc_id) . "'");
 216                 }
 217
 218         }