web/public_php/webtt/app/vendors/StringParser.php

   1 <?php
   2 /*
   3         Ryzom Core Web-Based Translation Tool
   4         Copyright (C) 2011 Piotr Kaczmarek <p.kaczmarek@openlink.pl>
   5
   6         This program is free software: you can redistribute it and/or modify
   7         it under the terms of the GNU General Public License as published by
   8         the Free Software Foundation, either version 3 of the License, or
   9         (at your option) any later version.
  10
  11         This program is distributed in the hope that it will be useful,
  12         but WITHOUT ANY WARRANTY; without even the implied warranty of
  13         MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14         GNU General Public License for more details.
  15
  16         You should have received a copy of the GNU General Public License
  17         along with this program.  If not, see <http://www.gnu.org/licenses/>.
  18 */
  19 ?>
  20 <?php
  21 class StringParser
  22 {
  23         var $pipeline_directory = "/home/kaczorek/projects/webtt/distfiles/translation/";
  24         var $debug = false;
  25
  26         function removeComments($str)
  27         {
  28 /*              while (($cstart = mb_strpos($str, "/*", $offset)) !== false)
  29                 {
  30                         $cend = mb_strpos();
  31                 }*/
  32 //              var_dump($str);
  33 //As a pertinent note, there's an issue with this function where parsing any string longer than 94326 characters long will silently return null. So be careful where you use it at.
  34 //http://pl.php.net/manual/en/function.preg-replace.php#98843
  35                 ini_set('pcre.backtrack_limit', 10000000);
  36                 //$returnString = preg_replace('!/\*.*?\*/!s', '', $str); // /* .*? */ s
  37                 // added [^/] because there was //******* in translation file
  38                 $returnString = preg_replace('![^/]/\*.*?\*/!s', '', $str); // /* .*? */ s
  39                 // PHP 5.2.0
  40                 // if (PREG_NO_ERROR !== preg_last_error())
  41                 if ($returnString === null)
  42                 {
  43                         $returnStr = $str;
  44                         var_dump("PREG ERROR");
  45                         // exception
  46                 }
  47                 return $returnString;
  48         }
  49
  50         function removeBOM($str)
  51         {
  52 //              mb_internal_encoding("ISO-8859-2");
  53 //              var_dump(substr($str, 0, 3));
  54                 if(($bom = substr($str, 0,3)) == pack("CCC",0xef,0xbb,0xbf))
  55                 {
  56 //                      var_dump("jest bom");
  57                         $bla = substr($str, 3);
  58 //                      var_dump($bla);
  59                         return $bla;
  60                 }
  61                 else
  62                 {
  63 //                      var_dump($bom);
  64                         return $str;
  65                 }
  66         }
  67
  68         function addBOM($str)
  69         {
  70                 if(($bom = substr($str, 0,3)) != pack("CCC",0xef,0xbb,0xbf))
  71                         return pack("CCC",0xef,0xbb,0xbf) . $str;
  72                 else
  73                         return $str;
  74         }
  75
  76         function parseLine($str)
  77         {
  78                 $arr = array();
  79 //              var_dump(mb_internal_encoding());
  80 //              mb_internal_encoding("ISO-8859-2");
  81                 if (mb_substr($str, 0, 2) == "//")
  82                 {
  83                         if (mb_substr($str, 0, 7) == "// DIFF")
  84                         {
  85                                 list($j, $type, $command, $args) = explode(" ", $str);
  86                                 $command = mb_strtolower($command);
  87                                 if ($command == "add" || $command == "changed")
  88                                 {
  89                                         $index = intval($args);
  90                                         $command = mb_substr($str, 3);
  91                                 }
  92                                 else
  93                                         unset($type);
  94                         }
  95                         else if (mb_substr($str, 0, 8) == "// INDEX")
  96                         {
  97                                 list($j, $type, $index) = explode(" ", $str);
  98                                 $type = "internal_index";
  99 //                              $arr = explode(" ", $str);
 100                         }
 101                         else if (mb_substr($str, 0, 13) == "// HASH_VALUE")
 102                         {
 103                                 list($j, $type, $hash_value) = explode(" ", $str);
 104                                 $type = "hash_value";
 105                         }
 106 /*                      if (!isset($type))
 107                         {
 108                                 var_dump(isset($type));
 109                                 debug_print_backtrace();
 110                         }
 111                         var_dump($type);*/
 112                         if (isset($type))
 113                         {
 114                                 $type = mb_strtolower($type);
 115                                 $arr = compact("type","command","index","hash_value");
 116                         }
 117                 }
 118                 else if (!(mb_substr($str, 0, 2) == "//") && mb_strlen($str))
 119                 {
 120                         //list($ident, $j
 121                         $type = "string";
 122                         $lBracket = mb_strpos($str, "[");
 123                         $rBracket = mb_strrpos($str, "]");
 124                         $sStart = $lBracket + 1;
 125                         $sEnd = $rBracket - ($sStart);
 126                         $identifier = trim(mb_substr($str, 0, $lBracket));
 127                         if (!$rBracket)
 128                                 $sEnd = mb_strlen($str);
 129                         $string = mb_substr($str, $sStart, $sEnd);
 130                         $string = str_replace(
 131                                         array('\\\\', '\[','\]'),
 132                                         array('\\', '[',']'),                           // '
 133                                         $string
 134                                 );
 135                         $arr = compact("type", "identifier", "string");
 136                 }
 137 /*              echo "<pre>################################\n";
 138                 var_dump($str);
 139                 var_dump($arr);
 140                 echo "</pre>\n";*/
 141                 return $arr;
 142         }
 143
 144         function parseFile($file)
 145         {
 146                 $parsedEnt = array();
 147                 $newEnt = false;
 148                 $prevStringLine = false;
 149                 $entities = array();
 150
 151 //              $file = file_get_contents($this->pipeline_directory . $file);
 152 //              var_dump(mb_substr($file, 0,3));
 153 //              var_dump(substr($file, 0,3));
 154 //              var_dump($file);
 155                 $file = $this->removeBOM($file);
 156 //              var_dump($file);
 157                 $file = $this->removeComments($file);
 158 //              var_dump($file);
 159                 $lines = explode("\n", $file);
 160                 if ($this->debug)
 161                 {
 162                         echo "<pre>\n\n";
 163                 }
 164                 $line_no=1;
 165                 foreach ($lines as $line)
 166                 {
 167                         if ($this->debug)
 168                         {
 169                                 echo "\n\t#################### LINE NUMBER " . $line_no++ . "\n\n";
 170                         }
 171
 172 //                      var_dump($line);
 173 //                      $line = rtrim($line);
 174 //                      $line = str_replace(array("\r\n","\n","\r"), '', $line);
 175                         $line = rtrim($line, "\r\n");
 176                         $parsedLine = $this->parseLine($line);
 177
 178                         if ($this->debug)
 179                         {
 180                                 echo "%%%% parsedLine\n";
 181                                 var_dump($parsedLine);
 182                                 echo "\n";
 183
 184                                 echo "%%%% prevStringLine\n";
 185                                 var_dump($prevStringLine);
 186                                 echo "\n";
 187                         }
 188
 189                         if (!$parsedLine)
 190                                 continue;
 191
 192                         // if line start with diff (diff files) or hash_value (translated files) and before was line with translation, then we start new ent
 193
 194                         if ($prevStringLine && (
 195                                         ($parsedLine["type"] == "diff" && $parsedEnt) || ($parsedLine["type"] == "hash_value" && $parsedEnt)
 196                                 ))
 197                         {
 198 /*                              echo "%%%% prevStringLine %%%%%\n";
 199                                 var_dump($parsedEnt);*/
 200                                 $newEnt = true;
 201                         }
 202
 203                         if ($newEnt)
 204                         {
 205                                 if ($this->debug)
 206                                 {
 207                                         echo "\t%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n";
 208                                         echo "\t%%%% newEnt %%%%%%%%% newEnt %%%%%%%%% newEnt %%%%%%%%% newEnt %%%%%\n";
 209                                         echo "\t%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%\n\n";
 210                                         var_dump($parsedEnt);
 211                                 }
 212                                 if (!isset($parsedEnt["diff"]) && !isset($parsedEnt["index"]))
 213                                         $parsedEnt["index"] = $parsedEnt["internal_index"];
 214
 215                                 $entities[] = $parsedEnt;
 216                                 $parsedEnt =array();
 217                                 $newEnt = false;
 218                         }
 219
 220                         if ($parsedLine["type"] == "internal_index")
 221                                         $parsedEnt["internal_index"] = $parsedLine["index"];
 222
 223                         if ($parsedLine["type"] == "string")
 224                         {
 225                                 $prevStringLine = true;
 226
 227                                 if ($this->debug)
 228                                 {
 229                                         echo "%%%% parsedEnt %%%%%\n";
 230                                         var_dump($parsedEnt);
 231
 232 //                                      echo "%%%% parsedLine %%%%%\n";
 233 //                                      var_dump($parsedLine);
 234                                 }
 235
 236                                 if (!$parsedLine['identifier'])
 237                                 {
 238                                         if ($this->debug) echo "ZLACZENIE \n";
 239                                         if ($this->debug && !isset($parsedEnt['string']))
 240                                         {
 241                                                 echo "!isset parsedEnt['string']\n";
 242                                                 var_dump($line);
 243                                                 var_dump($parsedEnt);
 244                                                 var_dump($parsedLine);
 245                                         }
 246 //                                      $parsedEnt['string'] .= $parsedLine['string'] . "\n";
 247                                         $parsedEnt['string'] .= "\n" . $parsedLine['string'];
 248                                 }
 249                                 else
 250                                 {
 251                                         if ($this->debug) echo "DODANIE \n";
 252                                         $parsedEnt += $parsedLine;
 253 //                                      $parsedEnt['string'] .= "\n";
 254                                 }
 255
 256                                 if ($this->debug)
 257                                 {
 258                                         echo "%%%% parsedEnt after %%%%%\n";
 259                                         var_dump($parsedEnt);
 260                                 }
 261                         }
 262                         else
 263                                 $prevStringLine = false;
 264
 265                         if ($parsedLine["type"] == "diff")
 266                         {
 267                                 $parsedEnt["diff"] = $parsedEnt["command"] = $parsedLine["command"];
 268                                 $parsedEnt["index"] = $parsedLine["index"];
 269                         }
 270                 }
 271                 if ($parsedEnt)
 272                 {
 273                         if (!isset($parsedEnt["diff"]) && !isset($parsedEnt["index"]))
 274                                 $parsedEnt["index"] = $parsedEnt["internal_index"];
 275
 276                         $entities[] = $parsedEnt;
 277                 }
 278
 279                 if ($this->debug)
 280                 {
 281                         echo "<pre>";
 282                         var_dump($entities);
 283                         echo "</pre>\n";
 284                 }
 285                 return $entities;
 286         }
 287
 288         function CRLF($s)
 289         {
 290                 $s = str_replace("\r\n", "\n", $s);
 291                 $s = str_replace("\n", "\r\n", $s);
 292                 return $s;
 293         }
 294
 295
 296         function buildFile($entities)
 297         {
 298                 $content = '';
 299                 foreach ($entities as $ent)
 300                 {
 301                         if (isset($ent['command']))
 302                                 $content .= '// ' . $ent['command'] . "\n";
 303                         if (isset($ent['hash_value']))
 304                                 $content .= '// HASH_VALUE ' . $ent['hash_value'];
 305 /*                      if (isset($ent['command']))
 306                                 $content .= '// INDEX ' . $ent['internal_index'] . "\n";
 307                         else*/
 308                         if (!isset($ent['command']))
 309                                 $content .= '// INDEX ' . $ent['index'] . "\n";
 310                         $content .= $ent['identifier'] . "\t" . '[' . $ent['string'] . ']' . "\n";
 311                         $content .= "\n";
 312                 }
 313                 return $this->addBOM($this->CRLF($content));
 314         }
 315 }
 316 ?>