lib/csvlib.class.php

   1 <?php
   2 // This file is part of Moodle - http://moodle.org/
   3 //
   4 // Moodle is free software: you can redistribute it and/or modify
   5 // it under the terms of the GNU General Public License as published by
   6 // the Free Software Foundation, either version 3 of the License, or
   7 // (at your option) any later version.
   8 //
   9 // Moodle is distributed in the hope that it will be useful,
  10 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 // GNU General Public License for more details.
  13 //
  14 // You should have received a copy of the GNU General Public License
  15 // along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
  16
  17 /**
  18  * This is a one-line short description of the file
  19  *
  20  * You can have a rather longer description of the file as well,
  21  * if you like, and it can span multiple lines.
  22  *
  23  * @package    core
  24  * @subpackage lib
  25  * @copyright  Petr Skoda
  26  * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
  27  */
  28
  29 defined('MOODLE_INTERNAL') || die();
  30
  31 /**
  32  * Utitily class for importing of CSV files.
  33  * @copyright Petr Skoda
  34  * @license   http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
  35  * @package   moodlecore
  36  */
  37 class csv_import_reader {
  38     /**
  39      * @var int import identifier
  40      */
  41     var $_iid;
  42     /**
  43      * @var string which script imports?
  44      */
  45     var $_type;
  46     /**
  47      * @var string|null Null if ok, error msg otherwise
  48      */
  49     var $_error;
  50     /**
  51      * @var array cached columns
  52      */
  53     var $_columns;
  54     /**
  55      * @var object file handle used during import
  56      */
  57     var $_fp;
  58
  59     /**
  60      * Contructor
  61      *
  62      * @param int $iid import identifier
  63      * @param string $type which script imports?
  64      */
  65     function csv_import_reader($iid, $type) {
  66         $this->_iid  = $iid;
  67         $this->_type = $type;
  68     }
  69
  70     /**
  71      * Parse this content
  72      *
  73      * @global object
  74      * @global object
  75      * @param string $content passed by ref for memory reasons, unset after return
  76      * @param string $encoding content encoding
  77      * @param string $delimiter_name separator (comma, semicolon, colon, cfg)
  78      * @param string $column_validation name of function for columns validation, must have one param $columns
  79      * @return bool false if error, count of data lines if ok; use get_error() to get error string
  80      */
  81     function load_csv_content(&$content, $encoding, $delimiter_name, $column_validation=null) {
  82         global $USER, $CFG;
  83
  84         $this->close();
  85         $this->_error = null;
  86
  87         $textlib = textlib_get_instance();
  88
  89         $content = $textlib->convert($content, $encoding, 'utf-8');
  90         // remove Unicode BOM from first line
  91         $content = $textlib->trim_utf8_bom($content);
  92         // Fix mac/dos newlines
  93         $content = preg_replace('!\r\n?!', "\n", $content);
  94         // is there anyting in file?
  95         $columns = strtok($content, "\n");
  96         if ($columns === false) {
  97             $this->_error = get_string('csvemptyfile', 'error');
  98             return false;
  99         }
 100         $csv_delimiter = csv_import_reader::get_delimiter($delimiter_name);
 101         $csv_encode    = csv_import_reader::get_encoded_delimiter($delimiter_name);
 102
 103         // process header - list of columns
 104         $columns   = explode($csv_delimiter, $columns);
 105         $col_count = count($columns);
 106         if ($col_count === 0) {
 107             $this->_error = get_string('csvemptyfile', 'error');
 108             return false;
 109         }
 110
 111         foreach ($columns as $key=>$value) {
 112             $columns[$key] = str_replace($csv_encode, $csv_delimiter, trim($value));
 113         }
 114         if ($column_validation) {
 115             $result = $column_validation($columns);
 116             if ($result !== true) {
 117                 $this->_error = $result;
 118                 return false;
 119             }
 120         }
 121         $this->_columns = $columns; // cached columns
 122
 123         // open file for writing
 124         $filename = $CFG->tempdir.'/csvimport/'.$this->_type.'/'.$USER->id.'/'.$this->_iid;
 125         $fp = fopen($filename, "w");
 126         fwrite($fp, serialize($columns)."\n");
 127
 128         // again - do we have any data for processing?
 129         $line = strtok("\n");
 130         $data_count = 0;
 131         while ($line !== false) {
 132             $line = explode($csv_delimiter, $line);
 133             foreach ($line as $key=>$value) {
 134                 $line[$key] = str_replace($csv_encode, $csv_delimiter, trim($value));
 135             }
 136             if (count($line) !== $col_count) {
 137                 // this is critical!!
 138                 $this->_error = get_string('csvweirdcolumns', 'error');
 139                 fclose($fp);
 140                 $this->cleanup();
 141                 return false;
 142             }
 143             fwrite($fp, serialize($line)."\n");
 144             $data_count++;
 145             $line = strtok("\n");
 146         }
 147
 148         fclose($fp);
 149         return $data_count;
 150     }
 151
 152     /**
 153      * Returns list of columns
 154      *
 155      * @return array
 156      */
 157     function get_columns() {
 158         if (isset($this->_columns)) {
 159             return $this->_columns;
 160         }
 161
 162         global $USER, $CFG;
 163
 164         $filename = $CFG->tempdir.'/csvimport/'.$this->_type.'/'.$USER->id.'/'.$this->_iid;
 165         if (!file_exists($filename)) {
 166             return false;
 167         }
 168         $fp = fopen($filename, "r");
 169         $line = fgets($fp);
 170         fclose($fp);
 171         if ($line === false) {
 172             return false;
 173         }
 174         $this->_columns = unserialize($line);
 175         return $this->_columns;
 176     }
 177
 178     /**
 179      * Init iterator.
 180      *
 181      * @global object
 182      * @global object
 183      * @return bool Success
 184      */
 185     function init() {
 186         global $CFG, $USER;
 187
 188         if (!empty($this->_fp)) {
 189             $this->close();
 190         }
 191         $filename = $CFG->tempdir.'/csvimport/'.$this->_type.'/'.$USER->id.'/'.$this->_iid;
 192         if (!file_exists($filename)) {
 193             return false;
 194         }
 195         if (!$this->_fp = fopen($filename, "r")) {
 196             return false;
 197         }
 198         //skip header
 199         return (fgets($this->_fp) !== false);
 200     }
 201
 202     /**
 203      * Get next line
 204      *
 205      * @return mixed false, or an array of values
 206      */
 207     function next() {
 208         if (empty($this->_fp) or feof($this->_fp)) {
 209             return false;
 210         }
 211         if ($ser = fgets($this->_fp)) {
 212             return unserialize($ser);
 213         } else {
 214             return false;
 215         }
 216     }
 217
 218     /**
 219      * Release iteration related resources
 220      *
 221      * @return void
 222      */
 223     function close() {
 224         if (!empty($this->_fp)) {
 225             fclose($this->_fp);
 226             $this->_fp = null;
 227         }
 228     }
 229
 230     /**
 231      * Get last error
 232      *
 233      * @return string error text of null if none
 234      */
 235     function get_error() {
 236         return $this->_error;
 237     }
 238
 239     /**
 240      * Cleanup temporary data
 241      *
 242      * @global object
 243      * @global object
 244      * @param boolean $full true means do a full cleanup - all sessions for current user, false only the active iid
 245      */
 246     function cleanup($full=false) {
 247         global $USER, $CFG;
 248
 249         if ($full) {
 250             @remove_dir($CFG->tempdir.'/csvimport/'.$this->_type.'/'.$USER->id);
 251         } else {
 252             @unlink($CFG->tempdir.'/csvimport/'.$this->_type.'/'.$USER->id.'/'.$this->_iid);
 253         }
 254     }
 255
 256     /**
 257      * Get list of cvs delimiters
 258      *
 259      * @return array suitable for selection box
 260      */
 261     static function get_delimiter_list() {
 262         global $CFG;
 263         $delimiters = array('comma'=>',', 'semicolon'=>';', 'colon'=>':', 'tab'=>'\\t');
 264         if (isset($CFG->CSV_DELIMITER) and strlen($CFG->CSV_DELIMITER) === 1 and !in_array($CFG->CSV_DELIMITER, $delimiters)) {
 265             $delimiters['cfg'] = $CFG->CSV_DELIMITER;
 266         }
 267         return $delimiters;
 268     }
 269
 270     /**
 271      * Get delimiter character
 272      *
 273      * @param string separator name
 274      * @return string delimiter char
 275      */
 276     static function get_delimiter($delimiter_name) {
 277         global $CFG;
 278         switch ($delimiter_name) {
 279             case 'colon':     return ':';
 280             case 'semicolon': return ';';
 281             case 'tab':       return "\t";
 282             case 'cfg':       if (isset($CFG->CSV_DELIMITER)) { return $CFG->CSV_DELIMITER; } // no break; fall back to comma
 283             case 'comma':     return ',';
 284         }
 285     }
 286
 287     /**
 288      * Get encoded delimiter character
 289      *
 290      * @global object
 291      * @param string separator name
 292      * @return string encoded delimiter char
 293      */
 294     function get_encoded_delimiter($delimiter_name) {
 295         global $CFG;
 296         if ($delimiter_name == 'cfg' and isset($CFG->CSV_ENCODE)) {
 297             return $CFG->CSV_ENCODE;
 298         }
 299         $delimiter = csv_import_reader::get_delimiter($delimiter_name);
 300         return '&#'.ord($delimiter);
 301     }
 302
 303     /**
 304      * Create new import id
 305      *
 306      * @global object
 307      * @param string who imports?
 308      * @return int iid
 309      */
 310     function get_new_iid($type) {
 311         global $USER;
 312
 313         $filename = make_temp_directory('csvimport/'.$type.'/'.$USER->id);
 314
 315         // use current (non-conflicting) time stamp
 316         $iiid = time();
 317         while (file_exists($filename.'/'.$iiid)) {
 318             $iiid--;
 319         }
 320
 321         return $iiid;
 322     }
 323 }