summaryrefslogtreecommitdiffstats
path: root/Bugzilla/Markdown.pm
blob: f73e25014745c843ff8177a868fe827b317594e4 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
# This Source Code Form is subject to the terms of the Mozilla Public
# License, v. 2.0. If a copy of the MPL was not distributed with this
# file, You can obtain one at http://mozilla.org/MPL/2.0/.
#
# This Source Code Form is "Incompatible With Secondary Licenses", as
# defined by the Mozilla Public License, v. 2.0.

package Bugzilla::Markdown;

use 5.10.1;
use strict;
use warnings;

use Bugzilla::Constants;
use Bugzilla::Template;

use Digest::MD5 qw(md5_hex);

use parent qw(Text::MultiMarkdown);

# Regex to match balanced [brackets]. See Friedl's
# "Mastering Regular Expressions", 2nd Ed., pp. 328-331.
our ($g_nested_brackets, $g_nested_parens);
$g_nested_brackets = qr{
    (?>                                 # Atomic matching
       [^\[\]]+                         # Anything other than brackets
     |
       \[
         (??{ $g_nested_brackets })     # Recursive set of nested brackets
       \]
    )*
}x;
# Doesn't allow for whitespace, because we're using it to match URLs:
$g_nested_parens = qr{
    (?>                                 # Atomic matching
       [^()\s]+                            # Anything other than parens or whitespace
     |
       \(
         (??{ $g_nested_parens })        # Recursive set of nested brackets
       \)
    )*
}x;

our %g_escape_table;

foreach my $char (split //, '\\`*_{}[]()>#+-.!~') {
    $g_escape_table{$char} = md5_hex($char);
}
$g_escape_table{'<'} = md5_hex('<');

sub new {
    my $invocant = shift;
    my $class = ref $invocant || $invocant;
    return $class->SUPER::new(tab_width => MARKDOWN_TAB_WIDTH,
                              # Bugzilla uses HTML not XHTML
                              empty_element_suffix => '>');
}

sub markdown {
    my $self = shift;
    my $text = shift;
    my $user = Bugzilla->user;

    if ($user->settings->{use_markdown}->{is_enabled}
        && $user->setting('use_markdown') eq 'on')
    {
        return $self->SUPER::markdown($text, @_);
    }

    return Bugzilla::Template::quoteUrls($text);
}

sub _Markdown {
    my $self = shift;
    my $text = shift;

    $text = Bugzilla::Template::quoteUrls($text, undef, undef, undef, 1);

    return $self->SUPER::_Markdown($text, @_);
}

sub _code_blocks {
    my ($self) = @_;
    $self->{code_blocks} = $self->{params}->{code_blocks} ||= [];
    return $self->{code_blocks};
}

sub _RunSpanGamut {
    # These are all the transformations that occur *within* block-level
    # tags like paragraphs, headers, and list items.

    my ($self, $text) = @_;

    $text = $self->_DoCodeSpans($text);
    $text = $self->_EscapeSpecialCharsWithinTagAttributes($text);
    $text = $self->_EscapeSpecialChars($text);

    $text = $self->_DoAnchors($text);

    # Strikethroughs is Bugzilla's extension
    $text = $self->_DoStrikethroughs($text);

    $text = $self->_DoAutoLinks($text);
    $text = $self->_EncodeAmpsAndAngles($text);
    $text = $self->_DoItalicsAndBold($text);

    $text =~ s/\n/<br$self->{empty_element_suffix}\n/g;

    return $text;
}

# We first replace all fenced code blocks with just their
# surrounding backticks and an empty body to know where
# they are exactly for later processing. The bodies of
# blocks will be in an array. This measure is taken to not
# interpret fenced code blocks contents as possible markdown
# structures. The contents of the body will be processed after
# processing markdown structures.
sub _removeFencedCodeBlocks {
    my ($self, $text) = @_;
    $text =~ s{
        ^ `{3,} [\s\t]* \n
        (                # $1 = the entire code block
          (?: .* \n+)+?
        )
        `{3,} [\s\t]* $
        }{
            push @{$self->_code_blocks}, $1;
            "%%%FENCED_BLOCK%%%";
        }egmx;
    return $text;
}

# Override to check for HTML-escaped <>" chars.
sub _StripLinkDefinitions {
    my ($self, $text) = @_;

    $text = $self->_removeFencedCodeBlocks($text);
    #
    # Strips link definitions from text, stores the URLs and titles in
    # hash references.
    #
    my $less_than_tab = $self->{tab_width} - 1;

    # Link defs are in the form: ^[id]: url "optional title"
    while ($text =~ s{
            ^[ ]{0,$less_than_tab}\[(.+)\]: # id = \$1
              [ \t]*
              \n?               # maybe *one* newline
              [ \t]*
            (?:&lt;)?<a\s+href="(.+?)">\2</a>(?:&gt;)?          # url = \$2
              [ \t]*
              \n?               # maybe one newline
              [ \t]*
            (?:
                (?<=\s)         # lookbehind for whitespace
                (?:&quot;|\()
                (.+?)           # title = \$3
                (?:&quot;|\))
                [ \t]*
            )?  # title is optional
            (?:\n+|\Z)
        }{}omx) {
        $self->{_urls}{lc $1} = $self->_EncodeAmpsAndAngles( $2 );    # Link IDs are case-insensitive
        if ($3) {
            $self->{_titles}{lc $1} = $3;
            $self->{_titles}{lc $1} =~ s/"/&quot;/g;
        }

    }

    return $text;
}

# We need to look for HTML-escaped '<' and '>' (i.e. &lt; and &gt;).
# We also remove Email linkification from the original implementation
# as it is already done in Bugzilla's quoteUrls().
sub _DoAutoLinks {
    my ($self, $text) = @_;

    $text =~ s{(?:<|&lt;)((?:https?|ftp):[^'">\s]+?)(?:>|&gt;)}{<a href="$1">$1</a>}gi;
    return $text;
}

# The main reasons for overriding this method are
# resolving URL conflicts with Bugzilla's quoteUrls()
# and also changing '"' to '&quot;' in regular expressions wherever needed.
sub _DoAnchors {
#
# Turn Markdown link shortcuts into <a> tags.
#
    my ($self, $text) = @_;

    # We revert linkifications of non-email links and only
    # those links whose URL and title are the same because
    # this way we can be sure that link is generated by quoteUrls()
    $text =~ s@<a \s+ href="(?! mailto ) (.+?)">\1</a>@$1@xmg;

    #
    # First, handle reference-style links: [link text] [id]
    #
    $text =~ s{
        (                   # wrap whole match in $1
          \[
            ($g_nested_brackets)    # link text = $2
          \]

          [ ]?              # one optional space
          (?:\n[ ]*)?       # one optional newline followed by spaces

          \[
            (.*?)       # id = $3
          \]
        )
    }{
        my $whole_match = $1;
        my $link_text   = $2;
        my $link_id     = lc $3;

        if ($link_id eq "") {
            $link_id = lc $link_text;   # for shortcut links like [this][].
        }

        $link_id =~ s{[ ]*\n}{ }g; # turn embedded newlines into spaces

        $self->_GenerateAnchor($whole_match, $link_text, $link_id);
    }xsge;

    #
    # Next, inline-style links: [link text](url "optional title")
    #
    $text =~ s{
        (               # wrap whole match in $1
          \[
            ($g_nested_brackets)    # link text = $2
          \]
          \(            # literal paren
            [ \t]*
            ($g_nested_parens)   # href = $3
            [ \t]*
            (           # $4
              (&quot;|')    # quote char = $5
              (.*?)     # Title = $6
              \5        # matching quote
              [ \t]*    # ignore any spaces/tabs between closing quote and )
            )?          # title is optional
          \)
        )
    }{
        my $result;
        my $whole_match = $1;
        my $link_text   = $2;
        my $url         = $3;
        my $title       = $6;

        # Remove Bugzilla quoteUrls() linkification
        if ($url =~ /^a href="/ && $url =~ m|</a$|) {
            $url =~ s/^[^>]+>//;
            $url =~ s@</a$@@;
        }

        my $safe_url_regexp = Bugzilla::Template::SAFE_URL_REGEXP();
        $url = "http://$url" unless $url =~ /^$safe_url_regexp$/;

        $self->_GenerateAnchor($whole_match, $link_text, undef, $url, $title);
    }xsge;

    #
    # Handle reference-style shortcuts: [link text]
    # These must come last in case you've also got [link test][1]
    # or [link test](/foo)
    #
    $text =~ s{
        (                    # wrap whole match in $1
          \[
            ([^\[\]]+)        # link text = $2; can't contain '[' or ']'
          \]
        )
    }{
        my $result;
        my $whole_match = $1;
        my $link_text   = $2;
        (my $link_id = lc $2) =~ s{[ ]*\n}{ }g; # lower-case and turn embedded newlines into spaces

        $self->_GenerateAnchor($whole_match, $link_text, $link_id);
    }xsge;

    # Last, handle "naked" references
    # Caveat, does not handle ;http://amazon.com
    my $safe_url_regexp = Bugzilla::Template::SAFE_URL_REGEXP();
    $text =~ s{
        (
          (^|(?<![;^"'<>]))  # negative lookbehind, including ';' in '&lt;'
          (                  # wrap url in $3
             (?:https?|ftp):[^'">\s]+\w
          )
        )
    }{
        my $whole_match = $3;
        my $url = $whole_match;
        $self->_GenerateAnchor($whole_match, $url, undef, $url, undef);
    }xsge;

    return $text;
}

# The purpose of overriding this function is to add support
# for a Github Flavored Markdown (GFM) feature called 'Multiple
# underscores in words'. The standard markdown specification
# specifies the underscore for making the text emphasized/bold.
# However, some variable names in programming languages contain underscores
# and we do not want a part of those variables to look emphasized/bold.
# Instead, we render them as the way they originally are.
sub _DoItalicsAndBold {
    my ($self, $text) = @_;

    # Handle at beginning of lines:
    $text =~ s{ (^__ (?=\S) (.+?[*_]*) (?<=\S) __ (?!\S)) }
              {
                  my $result = _has_multiple_underscores($2) ? $1 : "<strong>$2</strong>";
                  $result;
              }gsxe;

    $text =~ s{ ^\*\* (?=\S) (.+?[*_]*) (?<=\S) \*\* }{<strong>$1</strong>}gsx;

    $text =~ s{ (^_ (?=\S) (.+?) (?<=\S) _ (?!\S)) }
              {
                  my $result = _has_multiple_underscores($2) ? $1 : "<em>$2</em>";
                  $result;
              }gsxe;

    $text =~ s{ ^\* (?=\S) (.+?) (?<=\S) \* }{<em>$1</em>}gsx;

    # <strong> must go first:
    $text =~ s{ ( (?<=\W) __ (?=\S) (.+?[*_]*) (?<=\S) __ (?!\S) ) }
              {
                  my $result = _has_multiple_underscores($2) ? $1 : "<strong>$2</strong>";
                  $result;
              }gsxe;


    $text =~ s{ (?<=\W) \*\* (?=\S) (.+?[*_]*) (?<=\S) \*\* }{<strong>$1</strong>}gsx;

    $text =~ s{ ( (?<=\W) _ (?=\S) (.+?) (?<=\S) _ (?!\S) ) }
              {
                  my $result = _has_multiple_underscores($2) ? $1 : "<em>$2</em>";
                  $result;
              }gsxe;

    $text =~ s{ (?<=\W) \* (?=\S) (.+?) (?<=\S) \* }{<em>$1</em>}gsx;

    # And now, a second pass to catch nested strong and emphasis special cases
    $text =~ s{ ( (?<=\W) __ (?=\S) (.+?[*_]*) (?<=\S) __ (\S*) ) }
              {
                  my $result = _has_multiple_underscores($3) ? $1 : "<strong>$2</strong>$3";
                  $result;
              }gsxe;

    $text =~ s{ (?<=\W) \*\* (?=\S) (.+?[*_]*) (?<=\S) \*\* }{<strong>$1</strong>}gsx;
    $text =~ s{ ( (?<=\W) _ (?=\S) (.+?) (?<=\S) _ (\S*) ) }
              {
                  my $result = _has_multiple_underscores($3) ? $1 : "<em>$2</em>$3";
                  $result;
              }gsxe;

    $text =~ s{ (?<=\W) \* (?=\S) (.+?) (?<=\S) \* }{<em>$1</em>}gsx;

    return $text;
}

sub _DoStrikethroughs {
    my ($self, $text) = @_;

    $text =~ s{ ^ ~~ (?=\S) ([^~]+?) (?<=\S) ~~ (?!~) }{<del>$1</del>}gsx;
    $text =~ s{ (?<=_|[^~\w]) ~~ (?=\S) ([^~]+?) (?<=\S) ~~ (?!~) }{<del>$1</del>}gsx;

    return $text;
}

# The original _DoCodeSpans() uses the 's' modifier in its regex
# which prevents _DoCodeBlocks() to match GFM fenced code blocks.
# We copy the code from the original implementation and remove the
# 's' modifier from it.
sub _DoCodeSpans {
    my ($self, $text) = @_;

    $text =~ s@
            (?<!\\)     # Character before opening ` can't be a backslash
            (`+)        # $1 = Opening run of `
            (.+?)       # $2 = The code block
            (?<!`)
            \1          # Matching closer
            (?!`)
        @
             my $c = "$2";
             $c =~ s/^[ \t]*//g; # leading whitespace
             $c =~ s/[ \t]*$//g; # trailing whitespace
             $c = $self->_EncodeCode($c);
            "<code>$c</code>";
        @egx;

    return $text;
}

# Override to add GFM Fenced Code Blocks
sub _DoCodeBlocks {
    my ($self, $text) = @_;

    # First, do the standard code blocks to avoid generating nested code blocks
    # if the block is both indented and is surrounded by backticks.
    $text = $self->SUPER::_DoCodeBlocks($text);

    $text =~ s{
        ^ %%%FENCED_BLOCK%%%
        }{
            my $codeblock = shift @{$self->_code_blocks};
            my $result;

            $codeblock = $self->_EncodeCode($codeblock);
            $codeblock = $self->_Detab($codeblock);
            $codeblock =~ s/\n\z//; # remove the trailing newline

            $result = "\n\n<pre><code>" . $codeblock . "</code></pre>\n\n";
            $result;
        }egmx;

    return $text;
}

sub _DoBlockQuotes {
    my ($self, $text) = @_;

    $text =~ s{
          (                             # Wrap whole match in $1
            (?:
              ^[ \t]*&gt;[ \t]?         # '>' at the start of a line
                .+\n                    # rest of the first line
              (?:.+\n)*                 # subsequent consecutive lines
              \n*                       # blanks
            )+
          )
        }{
            my $bq = $1;
            $bq =~ s/^[ \t]*&gt;[ \t]?//gm; # trim one level of quoting
            $bq =~ s/^[ \t]+$//mg;          # trim whitespace-only lines
            $bq = $self->_removeFencedCodeBlocks($bq);
            $bq = $self->_RunBlockGamut($bq, {wrap_in_p_tags => 1});      # recurse
            $bq =~ s/^/  /mg;
            # These leading spaces screw with <pre> content, so we need to fix that:
            $bq =~ s{(\s*<pre>.+?</pre>)}{
                        my $pre = $1;
                        $pre =~ s/^  //mg;
                        $pre;
                    }egs;
            "<blockquote class=\"markdown\">\n$bq\n</blockquote>\n\n";
        }egmx;

    return $text;
}

sub _EncodeCode {
    my ($self, $text) = @_;

    # We need to unescape the escaped HTML characters in code blocks.
    # These are the reverse of the escapings done in Bugzilla::Util::html_quote()
    $text =~ s/&lt;/</g;
    $text =~ s/&gt;/>/g;
    $text =~ s/&quot;/"/g;
    $text =~ s/&#64;/@/g;
    # '&amp;' substitution must be the last one, otherwise a literal like '&gt;'
    # will turn to '>' because '&' is already changed to '&amp;' in Bugzilla::Util::html_quote().
    # In other words, html_quote() will change '&gt;' to '&amp;gt;' and then we will
    # change '&amp;gt' -> '&gt;' -> '>' if we write this substitution as the first one.
    $text =~ s/&amp;/&/g;
    $text =~ s{<a \s+ href="(?:mailto:)? (.+?)"> \1 </a>}{$1}xmgi;
    $text = $self->SUPER::_EncodeCode($text);
    $text =~ s/~/$g_escape_table{'~'}/go;
    # Encode '&lt;' to prevent URLs from getting linkified in code spans
    $text =~ s/&lt;/$g_escape_table{'&lt;'}/go;

    return $text;
}

sub _EncodeBackslashEscapes {
    my ($self, $text) = @_;

    $text = $self->SUPER::_EncodeBackslashEscapes($text);
    $text =~ s/\\~/$g_escape_table{'~'}/go;

    return $text;
}

sub _UnescapeSpecialChars {
    my ($self, $text) = @_;

    $text = $self->SUPER::_UnescapeSpecialChars($text);
    $text =~ s/$g_escape_table{'~'}/~/go;
    $text =~ s/$g_escape_table{'&lt;'}/&lt;/go;

    return $text;
}

# Check if the passed string is of the form multiple_underscores_in_a_word.
# To check that, we first need to make sure that the string does not contain
# any white-space. Then, if the string is composed of non-space chunks which
# are bound together with underscores, the string has the desired form.
sub _has_multiple_underscores {
    my $string = shift;
    return 0 unless $string;
    return 0 if $string =~ /\s/;
    return 1 if $string =~ /_/;
    return 0;
}

1;

__END__

=head1 NAME

Bugzilla::Markdown - Generates HTML output from structured plain-text input.

=head1 SYNOPSIS

 use Bugzilla::Markdown;

 my $markdown = Bugzilla::Markdown->new();
 print $markdown->markdown($text);

=head1 DESCRIPTION

Bugzilla::Markdown implements a Markdown engine that produces
an HTML-based output from a given plain-text input.

The majority of the implementation is done by C<Text::MultiMarkdown>
CPAN module. It also applies the linkifications done in L<Bugzilla::Template>
to the input resulting in an output which is a combination of both Markdown
structures and those defined by Bugzilla itself.

=head2 Accessors

=over

=item C<markdown>

C<string> Produces an HTML-based output string based on the structures
and format defined in the given plain-text input.

=over

=item B<Params>

=over

=item C<text>

C<string> A plain-text string which includes Markdown structures.

=back

=back

=back