xref: /plugin/siteexport/syntax/toc.php (revision a8c17ab5b37308343f86651acb8c4a1b3f36f0ae)
1<?php
2/**
3 * Search with Scopes
4 *
5 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
6 * @author     i-net software <tools@inetsoftware.de>
7 * @author     Gerry Weissbach <gweissbach@inetsoftware.de>
8 */
9
10// must be run within Dokuwiki
11if (!defined('DOKU_INC')) die();
12if (!defined('DOKU_PLUGIN')) define('DOKU_PLUGIN', DOKU_INC . 'lib/plugins/');
13
14require_once(DOKU_PLUGIN . 'syntax.php');
15
16class syntax_plugin_siteexport_toc extends DokuWiki_Syntax_Plugin {
17
18    private $insideToc = false;
19    private $savedToc = array();
20    private $options = array();
21
22    private $mergedPages = array();
23    private $includedPages = array();
24    private $merghintIds = array();
25
26    public function getType() { return 'protected'; }
27    public function getPType() { return 'block'; }
28    public function getAllowedTypes() { return array('container'); }
29    public function getSort() { return 100; }
30
31    /**
32     * Connect pattern to lexer
33     */
34    public function connectTo($mode) {
35        $this->Lexer->addEntryPattern('<toc>(?=.*?</toc>)', $mode, 'plugin_siteexport_toc');
36        $this->Lexer->addEntryPattern('<toc .+?>(?=.*?</toc>)', $mode, 'plugin_siteexport_toc');
37        $this->Lexer->addSpecialPattern("\[\[.+?\]\]", $mode, 'plugin_siteexport_toc');
38    }
39
40    public function postConnect() {
41        $this->Lexer->addExitPattern('</toc.*?>', 'plugin_siteexport_toc');
42    }
43
44    public function handle($match, $state, $pos, Doku_Handler $handler) {
45        global $ID, $INFO;
46
47        switch ($state) {
48            case DOKU_LEXER_ENTER:
49
50                $this->insideToc = true;
51                $this->options = explode(' ', substr($match, 5, -1)?:"");
52                return array('start' => true, 'pos' => $pos, 'options' => $this->options);
53
54            case DOKU_LEXER_SPECIAL:
55
56                if ($this->insideToc) {
57
58                    $link = preg_replace(array('/^\[\[/', '/\]\]$/u'), '', $match);
59                    // Split title from URL
60                    $link = explode('|', $link, 2);
61                    if (!isset($link[1])) {
62                        $link[1] = NULL;
63                    } else if (preg_match('/^\{\{[^\}]+\}\}$/', $link[1])) {
64                        // If the title is an image, convert it to an array containing the image details
65                        $link[1] = Doku_Handler_Parse_Media($link[1]);
66                    }
67                    $link[0] = trim($link[0]);
68
69                    if (!(preg_match('/^[a-zA-Z0-9\.]+>{1}.*$/u', $link[0]) ||
70                    preg_match('/^\\\\\\\\[\w.:?\-;,]+?\\\\/u', $link[0]) ||
71                    preg_match('#^([a-z0-9\-\.+]+?)://#i', $link[0]) ||
72                    preg_match('<' . PREG_PATTERN_VALID_EMAIL . '>', $link[0]) ||
73                    preg_match('!^#.+!', $link[0]))
74                    ) {
75
76                        // Get current depth from call stack
77                        $depth = 1;
78                        if ($handler->CallWriter instanceof Doku_Handler_List) {
79
80                            $calls = array_reverse($handler->CallWriter->calls);
81                            $call = $calls[0];
82                            foreach ($calls as $item) {
83                                if (in_array($item[0], array('list_item', 'list_open'))) { $call = $item; break; }
84                            }
85
86                            $depth = $handler->CallWriter->interpretSyntax($call[1][0], $listType = null)-1; // Minus one because of plus one inside the interpret function
87                        }
88
89                        if (empty($link[0])) { break; } // No empty elements. This would lead to problems
90                        return array($link[0], $link[1], $depth);
91                    } else {
92                        // use parser! - but with another p
93                        $handler->internallink($match, $state, $pos);
94                    }
95                } else {
96                    // use parser!
97                    $handler->internallink($match, $state, $pos);
98                }
99
100                return false;
101            case DOKU_LEXER_UNMATCHED:
102
103                $handler->_addCall('cdata', array($match), $pos);
104
105                return false;
106            case DOKU_LEXER_EXIT:
107
108                $this->insideToc = false;
109                return 'save__meta';
110        }
111        return false;
112    }
113
114    public function render($mode, Doku_Renderer $renderer, $data) {
115        global $ID, $lang, $INFO;
116
117        list($SID, $NAME, $DEPTH) = $data;
118
119        resolve_pageid(getNS($ID), $SID, $exists = null);
120//        $SID = cleanID($SID); // hier kein cleanID, da sonst moeglicherweise der anker verloren geht
121
122        //    Render XHTML and ODT
123        if ($mode == 'xhtml' || $mode == 'odt') {
124
125            // TOC Title
126            if (is_array($data) && $data['start'] == true) {
127
128                if (is_Array($data['options'])) {
129                    foreach ($data['options'] as $opt) {
130                        switch ($opt) {
131                            case 'description' : $renderer->meta['sitetoc']['showDescription'] = true; break;
132                            case 'notoc' : $renderer->meta['sitetoc']['noTOC'] = true; break;
133                            case 'merge' : $renderer->meta['sitetoc']['mergeDoc'] = true; break;
134                            case 'nohead' : $renderer->meta['sitetoc']['noTocHeader'] = true; break;
135                            case 'mergeheader' : $renderer->meta['sitetoc']['mergeHeader'] = true; break;
136                            case 'pagebreak' : $renderer->meta['sitetoc']['pagebreak'] = true; break;
137                            case 'mergehint' : $renderer->meta['sitetoc']['mergehint'] = true; break;
138                        }
139                    }
140                }
141
142                $renderer->section_open("1 sitetoc");
143                if ($renderer->meta['sitetoc']['noTocHeader'] === false) {
144                    $renderer->header($lang['toc'], 1, $data['pos']);
145                }
146
147                return true;
148            } else
149
150            // All Output has been done
151            if (!is_array($data) && $data == 'save__meta') {
152
153                // Close TOC
154                $renderer->section_close();
155
156                if ($renderer->meta['sitetoc']['noTOC'] === true) {
157                    $renderer->doc = preg_replace("/<div.*?sitetoc.*?$/si", "", $renderer->doc);
158                }
159
160                // If this is not set, we may have it as Metadata
161                if (empty($this->mergedPages) && $renderer->meta['sitetoc']['mergeDoc']) {
162                    $toc = $renderer->meta['sitetoc']['siteexportTOC'];
163
164                    if (is_array($toc)) {
165                        foreach ($toc as $tocItem) {
166                            $this->mergedPages[] = array($tocItem['id'], $tocItem['depth']);
167                        }
168                    }
169
170                }
171
172                // If there is some data to be merged
173                if (count($this->mergedPages) > 0) {
174
175                    $renderer->doc = ''; // Start fresh!
176
177                    $renderer->section_open("1 mergedsite" . ($renderer->meta['sitetoc']['mergehint'] ? ' mergehint' : ''));
178
179                    // Prepare lookup Array
180                    foreach ($this->mergedPages as $tocItem) {
181                        list($this->includedPages[]) = explode('#', $tocItem[0]);
182                    }
183
184                    // Load the instructions
185                    $instr = array();
186                    foreach ($this->mergedPages as $tocElement) {
187
188                        list($tocItem, $depth) = $tocElement;
189                        $file = wikiFN($tocItem);
190
191                        if (@file_exists($file)) {
192                            $instructions = p_cached_instructions($file, false, $tocItem);
193                        } else {
194                            $instructions = p_get_instructions(io_readWikiPage($file, $tocItem));
195                        }
196
197                        // Convert Link and header instructions
198                        $instructions = $this->_convertInstructions($instructions, $addID = null, $renderer, $depth);
199
200                        if ($renderer->meta['sitetoc']['mergeHeader'] && count($this->mergedPages) > 1) {
201                            // get a hint for merged pages
202                            if ($renderer->meta['sitetoc']['mergehint']) {
203                                // only if the first section is already there
204                                $mergeHint = p_get_metadata($tocItem, 'mergehint', METADATA_RENDER_USING_SIMPLE_CACHE);
205                                if (empty($mergeHint)) { $mergeHint = p_get_metadata($tocItem, 'thema', METADATA_RENDER_USING_SIMPLE_CACHE); }
206                                if (empty($mergeHint)) { $mergeHint = tpl_pagetitle($tocItem, true); }
207                                $instructions = $this->_mergeWithHeaders($this->_initialHeaderStructure($instructions), $instructions, 1, $mergeHint);
208                            }
209                            // Merge
210                            $instr = $this->_mergeWithHeaders($instr, $instructions, 1);
211                        } else
212                        if ($renderer->meta['sitetoc']['pagebreak']) {
213                            $sitepagebreak = array(array(
214                                'plugin',
215                                array(
216                                    'siteexport_toctools',
217                                    array(
218                                        'pagebreak',
219                                        null,
220                                        null
221                                    )
222                                )
223                            ));
224                            $instr = array_merge($instr, $instructions, $sitepagebreak);
225                        } else {
226                            // Concat
227                            $instr = array_merge($instr, $instructions);
228                        }
229                    }
230
231                    if (!empty($instr)) {
232                        $this->_cleanAllInstructions($instr);
233
234                        //if its the document start, cut off the first element(document information)
235                        if ($instr[count($instr)-1][1][0] == 'siteexport_toctools') {
236                            $instr = array_slice($instr, 0, -1);
237                        }
238
239
240                        // print "<pre>"; print_r($instr); print "</pre>";
241                        $this->_render_output($renderer, $mode, $instr);
242                    }
243
244                    $renderer->section_close();
245                }
246                return true;
247            }
248
249            // Save the current ID
250            $LNID = $SID;
251
252            // Add ID to flags['mergeDoc']
253            if ($renderer->meta['sitetoc']['mergeDoc'] === true) { // || (count($renderer->meta['sitetoc']['siteexportTOC']) > 0 && $renderer->meta['sitetoc']['siteexportMergeDoc'] === true) ) {
254                $this->mergedPages[] = array($SID, $DEPTH);
255                resolve_pageid(getNS($ID), $SID, $exists);
256            } else {
257                // // print normal internal link (XHTML odt)
258                $renderer->internallink($LNID, $NAME, null);
259
260                // Display Description underneath
261                if ($renderer->meta['sitetoc']['showDescription'] === true) {
262                    $renderer->cdata(p_get_metadata($SID, 'description abstract', true));
263                }
264            }
265
266            // Render Metadata
267        } else if ($mode == 'metadata') {
268            if (!is_array($data) && $data == 'save__meta') {
269                $renderer->meta['sitetoc']['siteexportTOC'] = $this->savedToc;
270
271                foreach ($this->savedToc as $page) {
272                    $renderer->meta['relation']['references'][$page['id']] = $page['exists'];
273                }
274
275                $this->savedToc = array();
276            } else if (!isset($data['start']) && !isset($data['pos'])) {
277                $this->savedToc[] = $this->__addTocItem($SID, $NAME, $DEPTH, $renderer);
278            }
279        } else {
280            return false;
281        }
282
283        return true;
284    }
285
286    /*
287     * pull apart the ID and create an Entry for the TOC
288     */
289    private function __addTocItem($id, $name, $depth, $renderer) {
290        global $conf;
291        global $ID;
292
293        // Render Title
294        $default = $renderer->_simpleTitle($id);
295        $exists = false; $isImage = false; $linktype = null;
296        resolve_pageid(getNS($ID), $id, $exists);
297        $name = $renderer->_getLinkTitle($name, $default, $isImage, $id, $linktype);
298
299        //keep hash anchor
300        list($id, $hash) = explode('#', $id, 2);
301        if (!empty($hash)) $hash = $renderer->_headerToLink($hash);
302
303        // Build Sitetoc Item
304        $item = array();
305        $item['id'] = $id;
306        $item['name'] = $name;
307        $item['anchor'] = $hash;
308        $item['depth'] = $depth;
309        $item['exists'] = $exists;
310        if (!$conf['skipacl'] && auth_quickaclcheck($item['id']) < AUTH_READ) {
311            return false;
312        }
313
314        return $item;
315    }
316
317    /*
318     * Render the output of one page
319     */
320    private function _render_output($renderer, $mode, $instr) {
321        global $ID;
322
323        // Section IDs
324        // $addID = sectionID($addID, $check);    //not possible to use a:b:c for id
325
326        if ($mode == 'xhtml') {
327
328            //--------RENDER
329            //renderer information(TOC build / Cache used)
330            $info = array();
331            $content = p_render($mode, $instr, $info);
332
333            //Remove TOC`s, section edit buttons and tags
334            $content = $this->_cleanXHTML($content);
335
336            // embed the included page
337            // $renderer->doc .= '<div class="include">';
338            //add an anchor to find start of a inserted page
339            // $renderer->doc .= "<a name='$addID' id='$addID'>";
340            $renderer->doc .= $content;
341            // $renderer->doc .= '</div>';
342        } else if ($mode == 'odt') {
343
344            // Loop through the instructions
345            foreach ($instr as $instruction) {
346                // Execute the callback against the Renderer
347                call_user_func_array(array($renderer, $instruction[0]), $instruction[1]);
348            }
349        }
350    }
351
352    /*
353     * Corrects relative internal links and media and
354     * converts headers of included pages to subheaders of the current page
355     */
356    private function _convertInstructions($instr, $id, &$renderer, $depth = 1) {
357        global $ID;
358        global $conf;
359
360        $n = count($instr);
361
362        for ($i = 0; $i < $n; $i++) {
363            //internal links(links inside this wiki) an relative links
364            if ((substr($instr[$i][0], 0, 12) == 'internallink')) {
365                $this->_convert_link($renderer, $instr[$i], $id);
366            }
367            else if ((substr($instr[$i][0], 0, 13) == 'internalmedia')) {
368                $this->_convert_media($renderer, $instr[$i], $id);
369            }
370            else if ((substr($instr[$i][0], 0, 6) == 'header')) {
371                $this->_convert_header($renderer, $instr[$i], $depth-1); // -1 because the depth starts at 1
372            }
373            else if ((substr($instr[$i][0], 0, 12) == 'section_open')) {
374                $this->_convert_section($renderer, $instr[$i], $depth-1); // -1 because the depth starts at 1
375            }
376        }
377
378        //if its the document start, cut off the first element(document information)
379        if ($instr[0][0] == 'document_start')
380        return array_slice($instr, 1, -1);
381        else
382        return $instr;
383    }
384
385    /*
386     * Convert link of given instruction
387     */
388    private function _convert_link(&$renderer, &$instr, $id) {
389        global $ID;
390
391        $exists = false;
392
393        resolve_pageid(getNS($id), $instr[1][0], $exists);
394        list($pageID, $pageReference) = explode("#", $instr[1][0], 2);
395
396        if (in_array($pageID, $this->includedPages)) {
397            // Crate new internal Links
398            $check = null;
399
400            // Either get existing reference or create from first heading. If still not there take the alternate ID
401            $pageNameLink = empty($pageReference) ? sectionID($pageID, $check) : $pageReference;
402
403            $instr[1][0] = $ID . "#" . $pageNameLink;
404
405        } else {
406            // Convert external Links to plain Text
407
408            $instr = array(
409                        "cdata",
410            array($instr[1][1]),
411            $instr[2]
412            );
413        }
414    }
415
416    /*
417     * Convert internalmedia of given instruction
418     */
419    private function _convert_media(&$renderer, &$instr, $id) {
420        global $ID;
421
422        // Resolvemedia returns the absolute path to media by reference
423        $exists = false;
424        resolve_mediaid(getNS($id), $instr[1][0], $exists);
425    }
426
427    /**
428     * @param integer $depth
429     */
430    private function _convert_header(&$renderer, &$instr, $depth) {
431        // More Depth!
432        $instr[1][1] += $depth;
433    }
434
435    /**
436     * @param integer $depth
437     */
438    private function _convert_section(&$renderer, &$instr, $depth) {
439        // More Depth!
440        $instr[1][0] += $depth;
441    }
442
443    private function _mergeWithHeaders($existing, $newInstructions, $level = 1, $mergeHint = array()) {
444
445        $returnInstructions = array();
446        $preparedInstructions = array();
447        $existingStart = $existingEnd = 0;
448        $firstRun = true;
449
450        while ($this->_findNextHeaderSection($existing, $level, $existingStart, $existingEnd)) {
451
452            if ($firstRun) {
453                $returnInstructions = array_merge($returnInstructions, array_slice($existing, 0, $existingStart));
454                $firstRun = false;
455            }
456
457            $currentSlice = array_slice($existing, $existingStart, $existingEnd-$existingStart);
458
459            // Find matching part with headername
460            $newStart = $newEnd = 0;
461            if ($this->_findNextHeaderSection($newInstructions, $level, $newStart, $newEnd, $currentSlice[0][1][0])) {
462
463                $newSlice = array_slice($newInstructions, $newStart, $newEnd-$newStart);
464                if ($newSlice[0][0] == 'header')
465                    array_shift($newSlice); // Remove Heading
466
467                // merge found parts on next level.
468                $returnedInstructions = $this->_mergeWithHeaders($currentSlice, $newSlice, $level+1, $mergeHint);
469
470                // Put them at the end!
471                $preparedInstructions = array_merge($preparedInstructions, $returnedInstructions);
472
473                // Remove from input
474                array_splice($newInstructions, $newStart, $newEnd-$newStart);
475            } else {
476                // Nothing else found
477                $preparedInstructions = array_merge($preparedInstructions, $currentSlice);
478            }
479
480            $existingStart = $existingEnd;
481        }
482
483        // Append the rest
484        $returnInstructions = array_merge($returnInstructions, array_slice($existing, $existingStart));
485
486        // Check for section close inconsistencies and put one at the very end ...
487        $section_postpend = array();
488        if (
489            (
490                ($tmp1 = array_slice($newInstructions, -1))
491                && ($tmp1[0][0] == 'section_close')
492            )
493            &&
494            (
495                ($tmp2 = array_slice($newInstructions, -2))
496                && ($tmp2[0][0] == 'section_close')
497            )
498        ) {
499            $section_postpend = array_splice($newInstructions, -1);
500        }
501        if (
502            (
503                ($tmp3 = array_slice($returnInstructions, -1))
504                && ($tmp3[0][0] == 'section_close')
505            )
506            &&
507            (
508                ($tmp4 = array_slice($returnInstructions, -2))
509                && ($tmp4[0][0] == 'section_close')
510            )
511        ) {
512            $section_postpend = array_merge($section_postpend, array_splice($returnInstructions, -1));
513        }
514
515        // What if there are headings left inside the $newInstructions?????
516        // Find matching part with headername
517        $newStart = $newEnd = 0;
518        $section_prepend = array();
519        if ($this->_findNextHeaderSection($newInstructions, $level, $newStart, $newEnd)) {
520            // If there are header in here, build a prepend and have the rest at the end
521            $section_prepend = array_splice($newInstructions, 0, $newStart);
522        } else {
523            // If not, prepend all of it.
524            $section_prepend = $newInstructions;
525            $newInstructions = array();
526        }
527
528        $this->_insertMergeHint($section_prepend, $mergeHint);
529
530        $returnInstructions = array_merge($returnInstructions, $section_prepend, $preparedInstructions, $newInstructions, $section_postpend);
531
532        return $returnInstructions;
533    }
534
535    /**
536     * @param integer $level
537     */
538    private function _findNextHeaderSection($section, $level, &$start, &$end, $headerName = null) {
539
540        $inCount = count($section);
541        $currentSlice = -1;
542
543        // Find Level 1 Header that matches.
544        for ($i = $start; $i < $inCount; $i++) {
545
546            $instruction = $section[$i];
547            $end = $i; // Or it will be lost and a section close will be missing.
548
549            // First Level Header
550            if ($instruction[0] == 'header' && $instruction[1][1] == $level) {
551
552                if ($currentSlice > 0) {
553                    return true;
554                }
555
556                if ($headerName == null || ($headerName == $instruction[1][0])) {
557                    // Begin of new slice ...
558                    $start = $currentSlice = $i;
559                }
560            }
561        }
562
563        // Nothing found
564        $end = $i; // Or it will be lost and a section close will be missing.
565        return $currentSlice > 0;
566    }
567
568    private function _cleanAllInstructions(&$instr) {
569        $this->_cleanInstructions($instr, '/section_(close|open)/');
570        $this->_cleanInstructions($instr, '/listu_(close|open)/');
571        $this->_cleanInstructions($instr, '/listo_(close|open)/');
572    }
573
574    /**
575     * @param string $tag
576     */
577    private function _cleanInstructions(&$instructions, $tag) {
578
579        $inCount = count($instructions);
580        for ($i = 0; $i < $inCount; $i++) {
581
582            // Last instruction
583            if ($i == $inCount-1) {
584                break;
585            }
586
587            if (preg_match($tag, $instructions[$i][0]) && preg_match($tag, $instructions[$i+1][0]) && $instructions[$i][0] != $instructions[$i+1][0]) {
588
589                // found different tags, but both match the expression and follow each other - so they can be elliminated
590                array_splice($instructions, $i, 2);
591                $inCount -= 2;
592                $i--;
593            }
594
595        }
596    }
597
598    /**
599     * Strip everything except for the headers
600     */
601    private function _initialHeaderStructure($instructions) {
602        $inCount = count($instructions);
603        for ($i = 0; $i < $inCount; $i++) {
604
605            // Last instruction
606            if ($i == $inCount-1) {
607                break;
608            }
609
610            if (!in_array($instructions[$i][0], array('header', 'section_open', 'section_close', 'p_open', 'p_close'))) {
611                // found non-matching
612                array_splice($instructions, $i, 1);
613                $inCount--;
614                $i--;
615            }
616        }
617        return $instructions;
618    }
619
620    private function _insertMergeHint(&$instructions, $mergeHint) {
621
622        // Surround new slice with a mergehint
623        if (empty($mergeHint)) { return; }
624
625        // No emtpy insruction sets.
626        $this->_cleanAllInstructions($instructions);
627        if (empty($instructions)) { return; }
628
629        // only section content should be surrounded.
630        if ($instructions[0][0] != 'section_open') { return; }
631
632        // save for later use
633        $mergeHints = array();
634        $mergeHintId = sectionid($mergeHint, $mergeHints);
635        $this->merghintIds[$mergeHintId] = $mergeHint;
636
637        $mergeHintPrepend = array(array(
638            'plugin',
639            array(
640                'siteexport_toctools',
641                array(
642                    'mergehint',
643                    'start',
644                    $mergeHint,
645                    $mergeHintId
646                )
647            )
648        ));
649
650        $mergeHintPostpend = array(array(
651            'plugin',
652            array(
653                'siteexport_toctools',
654                array(
655                    'mergehint',
656                    'end',
657                    $mergeHint
658                )
659            )
660        ));
661
662/*
663        print "\n\n#########\n";
664        print_r($instructions);
665        print "\nn#########\n\n";
666*/
667        $instructions = array_merge($mergeHintPrepend, $instructions, $mergeHintPostpend);
668    }
669
670    /**
671     * Remove TOC, section edit buttons and tags
672     */
673    private function _cleanXHTML($xhtml) {
674        $replace = array(
675            '!<div class="toc">.*?(</div>\n</div>)!s' => '', // remove TOCs
676            '#<!-- SECTION \[(\d*-\d*)\] -->#s'       => '', // remove section edit buttons
677            '!<div id="tags">.*?(</div>)!s'           => ''  // remove category tags
678        );
679        $xhtml = preg_replace(array_keys($replace), array_values($replace), $xhtml);
680        return $xhtml;
681    }
682
683    /**
684     * Allow the plugin to prevent DokuWiki creating a second instance of itself
685     *
686     * @return bool   true if the plugin can not be instantiated more than once
687     */
688    public function isSingleton() {
689        return true;
690    }
691}
692// vim:ts=4:sw=4:et:enc=utf-8:
693