1 <?php
  2  /**
  3  * This file is part of phpMorphy library
  4  *
  5  * Copyright c 2007-2008 Kamaev Vladimir <heromantor@users.sourceforge.net>
  6  *
  7  * This library is free software; you can redistribute it and/or
  8  * modify it under the terms of the GNU Lesser General Public
  9  * License as published by the Free Software Foundation; either
 10  * version 2 of the License, or (at your option) any later version.
 11  *
 12  * This library is distributed in the hope that it will be useful,
 13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
 14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 15  * Lesser General Public License for more details.
 16  *
 17  * You should have received a copy of the GNU Lesser General Public
 18  * License along with this library; if not, write to the
 19  * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 20  * Boston, MA 02111-1307, USA.
 21  */
 22 
 23 /**
 24  * This file is autogenerated at Wed, 14 Oct 2009 01:34:00 +0400, don`t change it!
 25  */
 26 class phpMorphy_Graminfo_File extends phpMorphy_Graminfo {
 27     function getGramInfoHeaderSize() {
 28         return 20;
 29     }
 30 
 31     function readGramInfoHeader($offset) {
 32         $__fh = $this->resource;
 33         fseek($__fh, $offset); 
 34         
 35         $result = unpack(
 36             'vid/vfreq/vforms_count/vpacked_forms_count/vancodes_count/vancodes_offset/vancodes_map_offset/vaffixes_offset/vaffixes_size/vbase_size',
 37             fread($__fh, 20) 
 38         );
 39         
 40         $result['offset'] = $offset;
 41         
 42         return $result;
 43     }
 44 
 45     protected function readAncodesMap($info) {
 46         $__fh = $this->resource;
 47         
 48         // TODO: this can be wrong due to aligning ancodes map section
 49         $offset = $info['offset'] + 20 + $info['forms_count'] * 2;
 50                 
 51         fseek($__fh, $offset); 
 52             
 53         $forms_count = $info['packed_forms_count'];
 54         return unpack("v$forms_count",  fread($__fh, $forms_count * 2));
 55     }
 56 
 57     protected function splitAncodes($ancodes, $map) {
 58         $result = array();
 59         for($i = 1, $c = count($map), $j = 1; $i <= $c; $i++) {
 60             $res = array();
 61             
 62             for($k = 0, $kc = $map[$i]; $k < $kc; $k++, $j++) {
 63                 $res[] = $ancodes[$j];
 64             }
 65             
 66             $result[] = $res;
 67         }
 68 
 69         return $result;
 70     }
 71     
 72     function readAncodes($info) {
 73         $__fh = $this->resource;
 74         
 75         // TODO: this can be wrong due to aligning ancodes section
 76         $offset = $info['offset'] + 20;
 77 
 78         fseek($__fh, $offset); 
 79             
 80         $forms_count = $info['forms_count'];
 81         $ancodes = unpack("v$forms_count", fread($__fh, $forms_count * 2));
 82 
 83         /*
 84         if(!$expand) {
 85             return $ancodes;
 86         }
 87         */
 88 
 89         $map = $this->readAncodesMap($info);
 90         
 91         return $this->splitAncodes($ancodes, $map);
 92     }
 93     
 94     function readFlexiaData($info) {
 95         $__fh = $this->resource;
 96         
 97         $offset = $info['offset'] + 20;
 98 
 99         if(isset($info['affixes_offset'])) {
100             $offset += $info['affixes_offset'];
101         } else {
102             $offset += $info['forms_count'] * 2 + $info['packed_forms_count'] * 2;
103         }
104         
105         fseek($__fh, $offset); 
106         return explode($this->ends, fread($__fh, $info['affixes_size'] - $this->ends_size));
107     }
108     
109     function readAllGramInfoOffsets() {
110         return $this->readSectionIndex($this->header['flex_index_offset'], $this->header['flex_count']);
111     }
112 
113     protected function readSectionIndex($offset, $count) {
114         $__fh = $this->resource;
115         
116         fseek($__fh, $offset); 
117         
118         return array_values(unpack("V$count", fread($__fh, $count * 4)));
119     }
120 
121     function readAllFlexia() {
122         $__fh = $this->resource;
123         
124         $result = array();
125         
126         $offset = $this->header['flex_offset'];
127         
128         foreach($this->readSectionIndexAsSize($this->header['flex_index_offset'], $this->header['flex_count'], $this->header['flex_size']) as $size) {
129             $header = $this->readGramInfoHeader($offset);
130             $affixes = $this->readFlexiaData($header);
131             $ancodes = $this->readAncodes($header, true);
132 
133             $result[$header['id']] = array(
134                 'header' => $header,
135                 'affixes' => $affixes,
136                 'ancodes' => $ancodes
137             );
138 
139             $offset += $size;
140         }
141         
142         return $result;
143     }
144     
145     function readAllPartOfSpeech() {
146         $__fh = $this->resource;
147         
148         $result = array();
149         
150         $offset = $this->header['poses_offset'];
151         
152         foreach($this->readSectionIndexAsSize($this->header['poses_index_offset'], $this->header['poses_count'], $this->header['poses_size']) as $size) {
153             fseek($__fh, $offset); 
154             
155             $res = unpack(
156                 'vid/Cis_predict',
157                 fread($__fh, 3) 
158             );
159             
160             $result[$res['id']] = array(
161                 'is_predict' => (bool)$res['is_predict'],
162                 'name' => $this->cleanupCString(fread($__fh, $size - 3))
163             );
164 
165             $offset += $size;
166         }
167         
168         return $result;
169     }
170     
171     function readAllGrammems() {
172         $__fh = $this->resource;
173         
174         $result = array();
175         
176         $offset = $this->header['grammems_offset'];
177         
178         foreach($this->readSectionIndexAsSize($this->header['grammems_index_offset'], $this->header['grammems_count'], $this->header['grammems_size']) as $size) {
179             fseek($__fh, $offset); 
180             
181             $res = unpack(
182                 'vid/Cshift',
183                 fread($__fh, 3) 
184             );
185             
186             $result[$res['id']] = array(
187                 'shift' => $res['shift'],
188                 
189                 'name' => $this->cleanupCString(fread($__fh, $size - 3))
190             );
191 
192             $offset += $size;
193         }
194         
195         return $result;
196     }
197 
198     function readAllAncodes() {
199         $__fh = $this->resource;
200         
201         $result = array();
202         
203         $offset = $this->header['ancodes_offset'];
204         fseek($__fh, $offset); 
205         
206         for($i = 0; $i < $this->header['ancodes_count']; $i++) {
207             $res = unpack('vid/vpos_id', fread($__fh, 4));
208             $offset += 4;
209             
210             list(, $grammems_count) = unpack('v', fread($__fh, 2));
211             $offset += 2;
212             
213             $result[$res['id']] = array(
214                 'pos_id' => $res['pos_id'],
215                 'grammem_ids' => $grammems_count ? 
216                     array_values(unpack("v$grammems_count", fread($__fh, $grammems_count * 2))) :
217                     array(),
218                 'offset' => $offset,
219             );
220             
221             $offset += $grammems_count * 2;
222         }
223 
224         return $result;
225     }   
226 }
227