lib/wiki_to_markdown.php

   1 <?php
   2
   3 // This file is part of Moodle - http://moodle.org/
   4 //
   5 // Moodle is free software: you can redistribute it and/or modify
   6 // it under the terms of the GNU General Public License as published by
   7 // the Free Software Foundation, either version 3 of the License, or
   8 // (at your option) any later version.
   9 //
  10 // Moodle is distributed in the hope that it will be useful,
  11 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13 // GNU General Public License for more details.
  14 //
  15 // You should have received a copy of the GNU General Public License
  16 // along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
  17
  18 /**
  19  * Utility function to convert wiki-like to Markdown format
  20  *
  21  * @package    core
  22  * @subpackage lib
  23  * @copyright  Howard Miller, 2005
  24  * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
  25  */
  26
  27 defined('MOODLE_INTERNAL') || die();
  28
  29 /**#@+
  30  *  state defines
  31  */
  32 define( "STATE_NONE",1 ); // blank line has been detected, so looking for first line on next para
  33 define( "STATE_PARAGRAPH",2 ); // currently processing vanilla paragraph
  34 define( "STATE_BLOCKQUOTE",3 ); // currently processing blockquote section
  35 define( "STATE_PREFORM",4 ); // currently processing preformatted text
  36 define( "STATE_NOTIKI",5 ); // currently processing preformatted / no formatting
  37 /**#@-*/
  38 /**#@+
  39  * list defines
  40  */
  41 define( "LIST_NONE", 1 ); // no lists active
  42 define( "LIST_UNORDERED", 2 ); // unordered list active
  43 define( "LIST_ORDERED", 3 ); // ordered list active
  44 define( "LIST_DEFINITION", 4 ); // definition list active
  45 /**#@-*/
  46
  47 /**
  48  * @package   moodlecore
  49  * @copyright Howard Miller, 2005
  50  * @license   http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
  51  */
  52 class WikiToMarkdown {
  53
  54   var $block_state;
  55   var $list_state;
  56   var $list_depth;
  57   var $list_backtrack;
  58   var $output; // output buffer
  59   var $courseid;
  60
  61   function close_block( $state ) {
  62     // provide appropriate closure for block according to state
  63
  64     // if in list close this first
  65     $lclose = "";
  66     if ($this->list_state != LIST_NONE) {
  67       $lclose = $this->do_list( " ",true );
  68     }
  69
  70     $sclose = "";
  71     switch ($state) {
  72       case STATE_PARAGRAPH:
  73         $sclose =  "\n";
  74         break;
  75       case STATE_BLOCKQUOTE:
  76         $sclose =  "\n";
  77         break;
  78       case STATE_PREFORM:
  79         $sclose =  "</pre>\n";
  80         break;
  81       case STATE_NOTIKI:
  82         $sclose =  "\n";
  83         break;
  84     }
  85
  86     return $lclose . $sclose;
  87   }
  88
  89   function do_replace( $line, $mark, $tag ) {
  90     // do the regex thingy for things like bold, italic etc
  91     // $mark is the magic character, and $tag the HTML tag to insert
  92
  93     // BODGE: replace inline $mark characters in places where we want them ignored
  94     // they will be put back after main substitutue, stops problems with eg, and/or
  95     $bodge = chr(1);
  96     $line = preg_replace( '/([[:alnum:]])'.$mark.'([[:alnum:]])/i', '\\1'.$bodge.'\\2',$line );
  97
  98     $regex = '/(^| |[(.,])'.$mark.'([^'.$mark.']*)'.$mark.'([^[:alnum:]]|$)/i';
  99     $replace = '\\1<'.$tag.'>\\2</'.$tag.'>\\3';
 100     $line = preg_replace( $regex, $replace, $line );
 101
 102     // BODGE: back we go
 103     $line = preg_replace( '/'.$bodge.'/i', $mark, $line );
 104
 105     return $line;
 106   }
 107
 108
 109   function do_replace_markdown( $line, $mark, $tag ) {
 110     // do the regex thingy for things like bold, italic etc
 111     // $mark is the magic character, and $tag the HTML tag to insert
 112     // MARKDOWN version does not generate HTML tags, just straigt replace
 113
 114     // BODGE: replace inline $mark characters in places where we want them ignored
 115     // they will be put back after main substitutue, stops problems with eg, and/or
 116     $bodge = chr(1);
 117     $line = preg_replace( '/([[:alnum:]])'.$mark.'([[:alnum:]])/i', '\\1'.$bodge.'\\2',$line );
 118
 119     $regex = '/(^| |[(.,])'.$mark.'([^'.$mark.']*)'.$mark.'([^[:alnum:]]|$)/i';
 120     $replace = '\\1'.$tag.'\\2'.$tag.'\\3';
 121     $line = preg_replace( $regex, $replace, $line );
 122
 123     // BODGE: back we go
 124     $line = preg_replace( '/'.$bodge.'/i', $mark, $line );
 125
 126     return $line;
 127   }
 128
 129
 130   function do_replace_sub( $line, $mark, $tag ) {
 131     // do regex for subscript and superscript (slightly different)
 132     // $mark is the magic character and $tag the HTML tag to insert
 133
 134     $regex = '/'.$mark.'([^'.$mark.']*)'.$mark.'/i';
 135     $replace = '<'.$tag.'>\\1</'.$tag.'>';
 136
 137     return preg_replace( $regex, $replace, $line );
 138   }
 139
 140   function do_list( $line, $blank=false ) {
 141     // handle line with list character on it
 142     // if blank line implies drop to level 0
 143
 144     // get magic character and then delete it from the line if not blank
 145     if ($blank) {
 146       $listchar="";
 147       $count = 0;
 148     }
 149     else {
 150       $listchar = $line{0};
 151       $count = strspn( $line, $listchar );
 152       $line = preg_replace( "/^[".$listchar."]+ /i", "", $line );
 153     }
 154
 155     // find what sort of list this character represents
 156     $list_tag = "";
 157     $list_close_tag = "";
 158     $item_tag = "";
 159     $item_close_tag = "";
 160     $list_style = LIST_NONE;
 161     switch ($listchar) {
 162       case '*':
 163         $list_tag = "";
 164         $list_close_tag = "";
 165         $item_tag = "*";
 166         $item_close_tag = "";
 167         $list_style = LIST_UNORDERED;
 168         break;
 169       case '#':
 170         $list_tag = "";
 171         $list_close_tag = "";
 172         $item_tag = "1.";
 173         $item_close_tag = "";
 174         $list_style = LIST_ORDERED;
 175         break;
 176       case ';':
 177         $list_tag = "<dl>";
 178         $list_close_tag = "</dl>";
 179         $item_tag = "<dd>";
 180         $item_close_tag = "</dd>";
 181         $list_style = LIST_DEFINITION;
 182         break;
 183       case ':':
 184         $list_tag = "<dl>";
 185         $list_close_tag = "</dl>";
 186         $item_tag = "<dt>";
 187         $item_close_tag = "</dt>";
 188         $list_style = LIST_DEFINITION;
 189         break;
 190       }
 191
 192     // tag opening/closing regime now - fun bit :-)
 193     $tags = "";
 194
 195     // if depth has reduced do number of closes to restore level
 196     for ($i=$this->list_depth; $i>$count; $i-- ) {
 197       $close_tag = array_pop( $this->list_backtrack );
 198       $tags = $tags . $close_tag;
 199       }
 200
 201     // if depth has increased do number of opens to balance
 202     for ($i=$this->list_depth; $i<$count; $i++ ) {
 203       array_push( $this->list_backtrack, "$list_close_tag" );
 204       $tags = $tags . "$list_tag";
 205     }
 206
 207     // ok, so list state is now same as style and depth same as count
 208     $this->list_state = $list_style;
 209     $this->list_depth = $count;
 210
 211     // get indent
 212     $indent = substr( "                      ",1,$count-1 );
 213
 214     if ($blank) {
 215       $newline = $tags;
 216     }
 217     else {
 218       $newline = $tags . $indent . "$item_tag " . $line . "$item_close_tag";
 219     }
 220
 221     return $newline;
 222   }
 223
 224
 225   function line_replace( $line ) {
 226     // return line after various formatting replacements
 227     // have been made - order is vital to stop them interfering with each other
 228
 229     global $CFG;
 230
 231     // ---- (at least) means a <hr />
 232     // MARKDOWN: no change so leave
 233
 234     // is this a list line (starts with * # ; :)
 235     if (preg_match( "/^([*]+|[#]+|[;]+|[:]+) /i", $line )) {
 236       $line = $this->do_list( $line );
 237     }
 238
 239    // typographic conventions
 240    // MARKDOWN: no equiv. so convert to entity as before
 241     // $line = str_replace( "--", "&#8212;", $line );
 242     // $line = str_replace( " - ", " &#8211; ", $line );
 243     $line = str_replace( "...", " &#8230; ", $line );
 244     $line = str_replace( "(R)", "&#174;", $line );
 245     $line = str_replace( "(r)", "&#174;", $line );
 246     $line = str_replace( "(TM)", "&#8482;", $line );
 247     $line = str_replace( "(tm)", "&#8482;", $line );
 248     $line = str_replace( "(C)", "&#169;", $line );
 249     $line = str_replace( "1/4", "&#188;", $line );
 250     $line = str_replace( "1/2", "&#189;", $line );
 251     $line = str_replace( "3/4", "&#190;", $line );
 252     $line = preg_replace( "/([[:digit:]]+[[:space:]]*)x([[:space:]]*[[:digit:]]+)/i", "\\1&#215;\\2", $line ); // (digits) x (digits) - multiply
 253     // do formatting tags
 254     // NOTE: The / replacement  *has* to be first, or it will screw the
 255     //    HTML tags that are added by the other ones
 256     // MARKDOWN: only bold and italic change, rest are just HTML
 257     $line = $this->do_replace_markdown( $line, "\*", "**" );
 258     $line = $this->do_replace_markdown( $line, "/", "*" );
 259     $line = $this->do_replace( $line, "\+", "ins" );
 260     // $line = $this->do_replace( $line, "-", "del" );
 261     $line = $this->do_replace_sub( $line, "~", "sub" );
 262     $line = $this->do_replace_sub( $line, "\^", "sup" );
 263     $line = $this->do_replace( $line, "%", "code" );
 264     $line = $this->do_replace( $line, "@", "cite" );
 265
 266     // convert urls into proper link with optional link text URL(text)
 267     // MARDOWN: HTML conversion should work fine
 268     $line = preg_replace("/([[:space:]]|^)([[:alnum:]]+)://([^[:space:]]*)([[:alnum:]#?/&=])\(([^)]+)\)/i",
 269       "\\1[\\5](\\2://\\3\\4)", $line);
 270     $line = preg_replace("/([[:space:]])www\.([^[:space:]]*)([[:alnum:]#?/&=])\(([^)]+)\)/i",
 271       "\\1[\\5](http://www.\\2\\3)", $line);
 272
 273     // make urls (with and without httpd) into proper links
 274     $line = preg_replace("/([[:space:]]|^)([[:alnum:]]+)://([^[:space:]]*)([[:alnum:]#?/&=])/i",
 275       "\\1<\\2://\\3\\4>", $line);
 276     $line = preg_replace("/([[:space:]])www\.([^[:space:]]*)([[:alnum:]#?/&=])/i",
 277       "\\1<http://www.\\2\\3\>", $line);
 278
 279     // make email addresses into mailtos....
 280     // MARKDOWN doesn't quite support this, so do as html
 281     $line = preg_replace("/([[:space:]]|^)([[:alnum:]._-]+@[[:alnum:]._-]+)\(([^)]+)\)/i",
 282        "\\1<a href=\"mailto:\\2\">\\3</a>", $line);
 283
 284     // !# at the beginning of any lines means a heading
 285     // MARKDOWN: value (1-6) becomes number of hashes
 286     if (preg_match( "/^!([1-6]) (.*)$/i", $line, $regs )) {
 287       $depth = substr( $line, 1, 1 );
 288       $out = substr( '##########', 0, $depth);
 289       $line = preg_replace( "/^!([1-6]) (.*)$/i", "$out \\2", $line );
 290     }
 291
 292     // acronym handing, example HTML(Hypertext Markyp Language)
 293     // MARKDOWN: no equiv. so just leave as HTML
 294     $line = preg_replace( "/([A-Z]+)\(([^)]+)\)/", "<acronym title=\"\\2\">\\1</acronym>", $line );
 295
 296     // Replace resource link >>##(Description Text)
 297     // MARKDOWN: change to MD web link style
 298     $line = preg_replace("/ ([a-zA-Z]+):([0-9]+)\(([^)]+)\)/i",
 299        " [\\3](".$CFG->wwwroot."/mod/\\1/view.php?id=\\2) ", $line );
 300
 301     $coursefileurl = array(moodle_url::make_legacyfile_url($this->courseid, null));
 302
 303     // Replace picture resource link
 304     $line = preg_replace("#/([a-zA-Z0-9./_-]+)(png|gif|jpg)\(([^)]+)\)#i",
 305             "![\\3](".$coursefileurl."/\\1\\2)", $line );
 306
 307     // Replace file resource link
 308     $line = preg_replace("#file:/([[:alnum:]/._-]+)\(([^)]+)\)#i",
 309             "[\\2](".$coursefileurl."/\\1)", $line );
 310
 311     return $line;
 312   }
 313
 314   function convert( $content,$courseid ) {
 315
 316     // main entry point for processing Wiki-like text
 317     // $content is string containing text with Wiki-Like formatting
 318     // return: string containing Markdown formatting
 319
 320     // initialisation stuff
 321     $this->output = "";
 322     $this->block_state = STATE_NONE;
 323     $this->list_state = LIST_NONE;
 324     $this->list_depth = 0;
 325     $this->list_backtrack = array();
 326     $this->spelling_on = false;
 327     $this->courseid = $courseid;
 328
 329     // split content into array of single lines
 330     $lines = explode( "\n",$content );
 331     $buffer = "";
 332
 333     // run through lines
 334     foreach( $lines as $line ) {
 335       // is this a blank line?
 336       $blank_line = preg_match( "/^[[:blank:]\r]*$/i", $line );
 337       if ($blank_line) {
 338         // first end current block according to state
 339         $buffer = $buffer . $this->close_block( $this->block_state );
 340         $this->block_state = STATE_NONE;
 341         continue;
 342       }
 343
 344       // act now depending on current block state
 345       if ($this->block_state == STATE_NONE) {
 346         // first character of line defines block type
 347         if (preg_match( "/^> /i",$line )) {
 348           // blockquote
 349           $buffer = $buffer . $this->line_replace( $line ). "\n";
 350           $this->block_state = STATE_BLOCKQUOTE;
 351         }
 352         else
 353         if (preg_match( "/^  /i",$line) ) {
 354           // preformatted text
 355           // MARKDOWN: no real equiv. so just use <pre>
 356           $buffer = $buffer . "<pre>\n";
 357           $buffer = $buffer . $this->line_replace($line) . "\n";
 358           $this->block_state = STATE_PREFORM;
 359         }
 360         else
 361         if (preg_match("/^\% /i",$line) ) {
 362                 // preformatted text - no processing
 363                 // MARKDOWN: this is MD code form of a paragraph
 364                 $buffer = $buffer . "    " . preg_replace( "/^\%/i","",$line) . "\n";
 365                 $this->block_state = STATE_NOTIKI;
 366         }
 367         else {
 368           // ordinary paragraph
 369           $buffer = $buffer . $this->line_replace($line) . "\n";
 370           $this->block_state = STATE_PARAGRAPH;
 371         }
 372         continue;
 373       }
 374
 375       if (($this->block_state == STATE_PARAGRAPH) |
 376           ($this->block_state == STATE_BLOCKQUOTE) |
 377           ($this->block_state == STATE_PREFORM) ) {
 378         $buffer = $buffer . $this->line_replace($line) . "\n";
 379         continue;
 380       }
 381       elseif ($this->block_state == STATE_NOTIKI) {
 382         $buffer = $buffer . "    " .$line . "\n";
 383       }
 384     }
 385
 386     // close off any block level tags
 387     $buffer = $buffer . $this->close_block( $this->block_state );
 388
 389     //return $buffer;
 390     return $buffer;
 391   }
 392 }