public-inbox.git  about / heads / tags
an "archives first" approach to mailing lists
blob 275e509f2f1136ad62d129e114b245f9c8b75cd4 19048 bytes (raw)
$ git show v1.4.0:lib/PublicInbox/WWW.pm	# shows this blob on the CLI

  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
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
 
# Copyright (C) 2014-2020 all contributors <meta@public-inbox.org>
# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
#
# Main web interface for mailing list archives
#
# We focus on the lowest common denominators here:
# - targeted at text-only console browsers (w3m, links, etc..)
# - Only basic HTML, CSS only for line-wrapping <pre> text content for GUIs
#   and diff/syntax-highlighting (optional)
# - No JavaScript, graphics or icons allowed.
# - Must not rely on static content
# - UTF-8 is only for user-content, 7-bit US-ASCII for us
package PublicInbox::WWW;
use 5.010_001;
use strict;
use warnings;
use bytes (); # only for bytes::length
use PublicInbox::Config;
use PublicInbox::Hval;
use URI::Escape qw(uri_unescape);
use PublicInbox::MID qw(mid_escape);
use PublicInbox::GitHTTPBackend;
use PublicInbox::UserContent;
use PublicInbox::WwwStatic qw(r path_info_raw);

# TODO: consider a routing tree now that we have more endpoints:
our $INBOX_RE = qr!\A/([\w\-][\w\.\-]*)!;
our $MID_RE = qr!([^/]+)!;
our $END_RE = qr!(T/|t/|t\.mbox(?:\.gz)?|t\.atom|raw|)!;
our $ATTACH_RE = qr!([0-9][0-9\.]*)-($PublicInbox::Hval::FN)!;
our $OID_RE = qr![a-f0-9]{7,40}!;

sub new {
	my ($class, $pi_config) = @_;
	$pi_config ||= PublicInbox::Config->new;
	bless { pi_config => $pi_config }, $class;
}

# backwards compatibility, do not use
sub run {
	my ($req, $method) = @_;
	PublicInbox::WWW->new->call($req->env);
}

sub call {
	my ($self, $env) = @_;
	my $ctx = { env => $env, www => $self };

	# we don't care about multi-value
	%{$ctx->{qp}} = map {
		utf8::decode($_);
		tr/+/ /;
		my ($k, $v) = split('=', $_, 2);
		$v = uri_unescape($v // '');
		# none of the keys we care about will need escaping
		$k => $v;
	} split(/[&;]+/, $env->{QUERY_STRING});

	my $path_info = path_info_raw($env);
	my $method = $env->{REQUEST_METHOD};

	if ($method eq 'POST') {
		if ($path_info =~ m!$INBOX_RE/(?:(?:git/)?([0-9]+)(?:\.git)?/)?
					(git-upload-pack)\z!x) {
			my ($epoch, $path) = ($2, $3);
			return invalid_inbox($ctx, $1) ||
				serve_git($ctx, $epoch, $path);
		} elsif ($path_info =~ m!$INBOX_RE/(\w+)\.sql\.gz\z!o) {
			return get_altid_dump($ctx, $1, $2);
		} elsif ($path_info =~ m!$INBOX_RE/!o) {
			return invalid_inbox($ctx, $1) || mbox_results($ctx);
		}
	}
	elsif ($method !~ /\A(?:GET|HEAD)\z/) {
		return r(405);
	}

	# top-level indices and feeds
	if ($path_info eq '/' || $path_info eq '/manifest.js.gz') {
		www_listing($self)->call($env);
	} elsif ($path_info =~ m!$INBOX_RE\z!o) {
		invalid_inbox($ctx, $1) || r301($ctx, $1);
	} elsif ($path_info =~ m!$INBOX_RE(?:/|/index\.html)?\z!o) {
		invalid_inbox($ctx, $1) || get_index($ctx);
	} elsif ($path_info =~ m!$INBOX_RE/(?:atom\.xml|new\.atom)\z!o) {
		invalid_inbox($ctx, $1) || get_atom($ctx);
	} elsif ($path_info =~ m!$INBOX_RE/new\.html\z!o) {
		invalid_inbox($ctx, $1) || get_new($ctx);
	} elsif ($path_info =~ m!$INBOX_RE/description\z!o) {
		get_description($ctx, $1);
	} elsif ($path_info =~ m!$INBOX_RE/(?:(?:git/)?([0-9]+)(?:\.git)?/)?
				($PublicInbox::GitHTTPBackend::ANY)\z!ox) {
		my ($epoch, $path) = ($2, $3);
		invalid_inbox($ctx, $1) || serve_git($ctx, $epoch, $path);
	} elsif ($path_info =~ m!$INBOX_RE/([a-zA-Z0-9_\-]+).mbox\.gz\z!o) {
		serve_mbox_range($ctx, $1, $2);
	} elsif ($path_info =~ m!$INBOX_RE/$MID_RE/$END_RE\z!o) {
		msg_page($ctx, $1, $2, $3);

	} elsif ($path_info =~ m!$INBOX_RE/$MID_RE/$ATTACH_RE\z!o) {
		my ($idx, $fn) = ($3, $4);
		invalid_inbox_mid($ctx, $1, $2) || get_attach($ctx, $idx, $fn);
	# in case people leave off the trailing slash:
	} elsif ($path_info =~ m!$INBOX_RE/$MID_RE/(T|t)\z!o) {
		my ($inbox, $mid_ue, $suffix) = ($1, $2, $3);
		$suffix .= $suffix =~ /\A[tT]\z/ ? '/#u' : '/';
		r301($ctx, $inbox, $mid_ue, $suffix);

	} elsif ($path_info =~ m!$INBOX_RE/$MID_RE/R/?\z!o) {
		my ($inbox, $mid_ue) = ($1, $2);
		r301($ctx, $inbox, $mid_ue, '#R');

	} elsif ($path_info =~ m!$INBOX_RE/$MID_RE/f/?\z!o) {
		r301($ctx, $1, $2);
	} elsif ($path_info =~ m!$INBOX_RE/_/text(?:/(.*))?\z!o) {
		get_text($ctx, $1, $2);
	} elsif ($path_info =~ m!$INBOX_RE/([a-zA-Z0-9_\-\.]+)\.css\z!o) {
		get_css($ctx, $1, $2);
	} elsif ($path_info =~ m!$INBOX_RE/manifest\.js\.gz\z!o) {
		get_inbox_manifest($ctx, $1, $2);
	} elsif ($path_info =~ m!$INBOX_RE/($OID_RE)/s/\z!o) {
		get_vcs_object($ctx, $1, $2);
	} elsif ($path_info =~ m!$INBOX_RE/($OID_RE)/s/
				($PublicInbox::Hval::FN)\z!ox) {
		get_vcs_object($ctx, $1, $2, $3);
	} elsif ($path_info =~ m!$INBOX_RE/($OID_RE)/s\z!o) {
		r301($ctx, $1, $2, 's/');
	} elsif ($path_info =~ m!$INBOX_RE/(\w+)\.sql\.gz\z!o) {
		get_altid_dump($ctx, $1, $2);
	# convenience redirects order matters
	} elsif ($path_info =~ m!$INBOX_RE/([^/]{2,})\z!o) {
		r301($ctx, $1, $2);

	} else {
		legacy_redirects($ctx, $path_info);
	}
}

# for CoW-friendliness, MOOOOO!  Even for single-process setups,
# we want to get all immortal allocations done early to avoid heap
# fragmentation since common allocators favor a large contiguous heap.
sub preload {
	my ($self) = @_;
	require PublicInbox::ExtMsg;
	require PublicInbox::Feed;
	require PublicInbox::View;
	require PublicInbox::SearchThread;
	require PublicInbox::MIME;
	require PublicInbox::Mbox;
	require PublicInbox::ViewVCS;
	require PublicInbox::WwwText;
	require PublicInbox::WwwAttach;
	eval {
		require PublicInbox::Search;
		PublicInbox::Search::load_xapian();
	};
	for (qw(SearchView MboxGz WwwAltId)) {
		eval "require PublicInbox::$_;";
	}
	if (ref($self)) {
		my $pi_config = $self->{pi_config};
		if (defined($pi_config->{'publicinbox.cgitrc'})) {
			$pi_config->limiter('-cgit');
		}
		$self->cgit;
		$self->stylesheets_prepare($_) for ('', '../', '../../');
		$self->www_listing;
		$self->news_www;
		$pi_config->each_inbox(\&preload_inbox);
	}
}

sub preload_inbox {
	my $ibx = shift;
	$ibx->altid_map;
	$ibx->cloneurl;
	$ibx->description;
	$ibx->base_url;
}

# private functions below

sub r404 {
	my ($ctx) = @_;
	if ($ctx && $ctx->{mid}) {
		require PublicInbox::ExtMsg;
		return PublicInbox::ExtMsg::ext_msg($ctx);
	}
	r(404);
}

sub news_cgit_fallback ($) {
	my ($ctx) = @_;
	my $www = $ctx->{www};
	my $env = $ctx->{env};
	my $res = $www->news_www->call($env);
	$res->[0] == 404 ? $www->cgit->call($env) : $res;
}

# returns undef if valid, array ref response if invalid
sub invalid_inbox ($$) {
	my ($ctx, $inbox) = @_;
	my $ibx = $ctx->{www}->{pi_config}->lookup_name($inbox);
	if (defined $ibx) {
		$ctx->{-inbox} = $ibx;
		return;
	}

	# sometimes linkifiers (not ours!) screw up automatic link
	# generation and link things intended for nntp:// to https?://,
	# so try to infer links and redirect them to the appropriate
	# list URL.
	news_cgit_fallback($ctx);
}

# returns undef if valid, array ref response if invalid
sub invalid_inbox_mid {
	my ($ctx, $inbox, $mid_ue) = @_;
	my $ret = invalid_inbox($ctx, $inbox);
	return $ret if $ret;

	my $mid = $ctx->{mid} = uri_unescape($mid_ue);
	my $ibx = $ctx->{-inbox};
	if ($mid =~ m!\A([a-f0-9]{2})([a-f0-9]{38})\z!) {
		my ($x2, $x38) = ($1, $2);
		# this is horrifically wasteful for legacy URLs:
		my $str = $ctx->{-inbox}->msg_by_path("$x2/$x38") or return;
		require Email::Simple;
		my $s = Email::Simple->new($str);
		$mid = PublicInbox::MID::mid_clean($s->header('Message-ID'));
		return r301($ctx, $inbox, mid_escape($mid));
	}
	undef;
}

# /$INBOX/new.atom                     -> Atom feed, includes replies
sub get_atom {
	my ($ctx) = @_;
	require PublicInbox::Feed;
	PublicInbox::Feed::generate($ctx);
}

# /$INBOX/new.html			-> HTML only
sub get_new {
	my ($ctx) = @_;
	require PublicInbox::Feed;
	PublicInbox::Feed::new_html($ctx);
}

# /$INBOX/?r=$GIT_COMMIT                 -> HTML only
sub get_index {
	my ($ctx) = @_;
	require PublicInbox::Feed;
	if ($ctx->{env}->{QUERY_STRING} =~ /(?:\A|[&;])q=/) {
		require PublicInbox::SearchView;
		PublicInbox::SearchView::sres_top_html($ctx);
	} else {
		PublicInbox::Feed::generate_html_index($ctx);
	}
}

# /$INBOX/$MESSAGE_ID/raw                    -> raw mbox
sub get_mid_txt {
	my ($ctx) = @_;
	require PublicInbox::Mbox;
	PublicInbox::Mbox::emit_raw($ctx) || r404($ctx);
}

# /$INBOX/$MESSAGE_ID/                   -> HTML content (short quotes)
sub get_mid_html {
	my ($ctx) = @_;
	require PublicInbox::View;
	PublicInbox::View::msg_page($ctx) || r404($ctx);
}

# /$INBOX/$MESSAGE_ID/t/
sub get_thread {
	my ($ctx, $flat) = @_;
	$ctx->{-inbox}->over or return need($ctx, 'Overview');
	$ctx->{flat} = $flat;
	require PublicInbox::View;
	PublicInbox::View::thread_html($ctx);
}

# /$INBOX/_/text/$KEY/
# /$INBOX/_/text/$KEY/raw
# KEY may contain slashes
sub get_text {
	my ($ctx, $inbox, $key) = @_;
	my $r404 = invalid_inbox($ctx, $inbox);
	return $r404 if $r404;

	require PublicInbox::WwwText;
	PublicInbox::WwwText::get_text($ctx, $key);
}

# show git objects (blobs and commits)
# /$INBOX/_/$OBJECT_ID/show
# /$INBOX/_/${OBJECT_ID}_${FILENAME}
# KEY may contain slashes
sub get_vcs_object ($$$;$) {
	my ($ctx, $inbox, $oid, $filename) = @_;
	my $r404 = invalid_inbox($ctx, $inbox);
	return $r404 if $r404;
	require PublicInbox::ViewVCS;
	PublicInbox::ViewVCS::show($ctx, $oid, $filename);
}

sub get_altid_dump {
	my ($ctx, $inbox, $altid_pfx) =@_;
	my $r404 = invalid_inbox($ctx, $inbox);
	return $r404 if $r404;
	eval { require PublicInbox::WwwAltId } or return need($ctx, 'sqlite3');
	PublicInbox::WwwAltId::sqldump($ctx, $altid_pfx);
}

sub need {
	my ($ctx, $extra) = @_;
	my $msg = <<EOF;
<html><head><title>$extra not available for this
public-inbox</title><body><pre>$extra is not available for this public-inbox
<a href="../">Return to index</a></pre></body></html>
EOF
	[ 501, [ 'Content-Type' => 'text/html; charset=UTF-8' ], [ $msg ] ];
}

# /$INBOX/$MESSAGE_ID/t.mbox           -> thread as mbox
# /$INBOX/$MESSAGE_ID/t.mbox.gz        -> thread as gzipped mbox
# note: I'm not a big fan of other compression formats since they're
# significantly more expensive on CPU than gzip and less-widely available,
# especially on older systems.  Stick to zlib since that's what git uses.
sub get_thread_mbox {
	my ($ctx, $sfx) = @_;
	my $over = $ctx->{-inbox}->over or return need($ctx, 'Overview');
	require PublicInbox::Mbox;
	PublicInbox::Mbox::thread_mbox($ctx, $over, $sfx);
}


# /$INBOX/$MESSAGE_ID/t.atom		  -> thread as Atom feed
sub get_thread_atom {
	my ($ctx) = @_;
	$ctx->{-inbox}->over or return need($ctx, 'Overview');
	require PublicInbox::Feed;
	PublicInbox::Feed::generate_thread_atom($ctx);
}

sub legacy_redirects {
	my ($ctx, $path_info) = @_;

	# single-message pages
	if ($path_info =~ m!$INBOX_RE/m/(\S+)/\z!o) {
		r301($ctx, $1, $2);
	} elsif ($path_info =~ m!$INBOX_RE/m/(\S+)/raw\z!o) {
		r301($ctx, $1, $2, 'raw');

	} elsif ($path_info =~ m!$INBOX_RE/f/(\S+)/\z!o) {
		r301($ctx, $1, $2);

	# thread display
	} elsif ($path_info =~ m!$INBOX_RE/t/(\S+)/\z!o) {
		r301($ctx, $1, $2, 't/#u');

	} elsif ($path_info =~ m!$INBOX_RE/t/(\S+)/mbox(\.gz)?\z!o) {
		r301($ctx, $1, $2, "t.mbox$3");

	# even older legacy redirects
	} elsif ($path_info =~ m!$INBOX_RE/m/(\S+)\.html\z!o) {
		r301($ctx, $1, $2);

	} elsif ($path_info =~ m!$INBOX_RE/t/(\S+)\.html\z!o) {
		r301($ctx, $1, $2, 't/#u');

	} elsif ($path_info =~ m!$INBOX_RE/f/(\S+)\.html\z!o) {
		r301($ctx, $1, $2);

	} elsif ($path_info =~ m!$INBOX_RE/(?:m|f)/(\S+)\.txt\z!o) {
		r301($ctx, $1, $2, 'raw');

	} elsif ($path_info =~ m!$INBOX_RE/t/(\S+)(\.mbox(?:\.gz)?)\z!o) {
		r301($ctx, $1, $2, "t$3");

	# legacy convenience redirects, order still matters
	} elsif ($path_info =~ m!$INBOX_RE/m/(\S+)\z!o) {
		r301($ctx, $1, $2);
	} elsif ($path_info =~ m!$INBOX_RE/t/(\S+)\z!o) {
		r301($ctx, $1, $2, 't/#u');
	} elsif ($path_info =~ m!$INBOX_RE/f/(\S+)\z!o) {
		r301($ctx, $1, $2);

	# some Message-IDs have slashes in them and the HTTP server
	# may try to be clever and unescape them :<
	} elsif ($path_info =~ m!$INBOX_RE/(\S+/\S+)/$END_RE\z!o) {
		msg_page($ctx, $1, $2, $3);

	# in case people leave off the trailing slash:
	} elsif ($path_info =~ m!$INBOX_RE/(\S+/\S+)/(T|t)\z!o) {
		r301($ctx, $1, $2, $3 eq 't' ? 't/#u' : $3);
	} elsif ($path_info =~ m!$INBOX_RE/(\S+/\S+)/f\z!o) {
		r301($ctx, $1, $2);
	} else {
		news_cgit_fallback($ctx);
	}
}

sub r301 {
	my ($ctx, $inbox, $mid_ue, $suffix) = @_;
	my $ibx = $ctx->{-inbox};
	unless ($ibx) {
		my $r404 = invalid_inbox($ctx, $inbox);
		return $r404 if $r404;
		$ibx = $ctx->{-inbox};
	}
	my $url = $ibx->base_url($ctx->{env});
	my $qs = $ctx->{env}->{QUERY_STRING};
	if (defined $mid_ue) {
		# common, and much nicer as '@' than '%40':
		$mid_ue =~ s/%40/@/g;
		$url .= $mid_ue . '/';
	}
	$url .= $suffix if (defined $suffix);
	$url .= "?$qs" if $qs ne '';

	[ 301,
	  [ Location => $url, 'Content-Type' => 'text/plain' ],
	  [ "Redirecting to $url\n" ] ]
}

sub msg_page {
	my ($ctx, $inbox, $mid_ue, $e) = @_;
	my $ret;
	$ret = invalid_inbox_mid($ctx, $inbox, $mid_ue) and return $ret;
	'' eq $e and return get_mid_html($ctx);
	'T/' eq $e and return get_thread($ctx, 1);
	't/' eq $e and return get_thread($ctx);
	't.atom' eq $e and return get_thread_atom($ctx);
	't.mbox' eq $e and return get_thread_mbox($ctx);
	't.mbox.gz' eq $e and return get_thread_mbox($ctx, '.gz');
	'raw' eq $e and return get_mid_txt($ctx);

	# legacy, but no redirect for compatibility:
	'f/' eq $e and return get_mid_html($ctx);
	r404($ctx);
}

sub serve_git {
	my ($ctx, $epoch, $path) = @_;
	my $env = $ctx->{env};
	my $ibx = $ctx->{-inbox};
	my $git = defined $epoch ? $ibx->git_epoch($epoch) : $ibx->git;
	$git ? PublicInbox::GitHTTPBackend::serve($env, $git, $path) : r404();
}

sub mbox_results {
	my ($ctx) = @_;
	if ($ctx->{env}->{QUERY_STRING} =~ /(?:\A|[&;])q=/) {
		$ctx->{-inbox}->search or return need($ctx, 'search');
		require PublicInbox::SearchView;
		return PublicInbox::SearchView::mbox_results($ctx);
	}
	r404();
}

sub serve_mbox_range {
	my ($ctx, $inbox, $range) = @_;
	invalid_inbox($ctx, $inbox) || eval {
		require PublicInbox::Mbox;
		PublicInbox::Mbox::emit_range($ctx, $range);
	}
}

sub news_www {
	my ($self) = @_;
	$self->{news_www} ||= do {
		require PublicInbox::NewsWWW;
		PublicInbox::NewsWWW->new($self->{pi_config});
	}
}

sub cgit {
	my ($self) = @_;
	$self->{cgit} ||= do {
		my $pi_config = $self->{pi_config};

		if (defined($pi_config->{'publicinbox.cgitrc'})) {
			require PublicInbox::Cgit;
			PublicInbox::Cgit->new($pi_config);
		} else {
			require Plack::Util;
			Plack::Util::inline_object(call => sub { r404() });
		}
	}
}

sub www_listing {
	my ($self) = @_;
	$self->{www_listing} ||= do {
		require PublicInbox::WwwListing;
		PublicInbox::WwwListing->new($self);
	}
}

# GET $INBOX/manifest.js.gz
sub get_inbox_manifest ($$$) {
	my ($ctx, $inbox, $key) = @_;
	my $r404 = invalid_inbox($ctx, $inbox);
	return $r404 if $r404;
	require PublicInbox::WwwListing;
	PublicInbox::WwwListing::js($ctx->{env}, [$ctx->{-inbox}]);
}

sub get_attach {
	my ($ctx, $idx, $fn) = @_;
	require PublicInbox::WwwAttach;
	PublicInbox::WwwAttach::get_attach($ctx, $idx, $fn);
}

# User-generated content (UGC) may have excessively long lines
# and screw up rendering on some browsers, so we use pre-wrap.
#
# We also force everything to the same scaled font-size because GUI
# browsers (tested both Firefox and surf (webkit)) uses a larger font
# for the Search <form> element than the rest of the page.  Font size
# uniformity is important to people who rely on gigantic fonts.
# Finally, we use monospace to ensure the Search field and button
# has the same size and spacing as everything else which is
# <pre>-formatted anyways.
our $STYLE = 'pre{white-space:pre-wrap}*{font-size:100%;font-family:monospace}';

sub stylesheets_prepare ($$) {
	my ($self, $upfx) = @_;
	my $mini = eval {
		require CSS::Minifier;
		sub { CSS::Minifier::minify(input => $_[0]) };
	} || eval {
		require CSS::Minifier::XS;
		sub { CSS::Minifier::XS::minify($_[0]) };
	} || sub { $_[0] };

	my $css_map = {};
	my $stylesheets = $self->{pi_config}->{css} || [];
	my $links = [];
	my $inline_ok = 1;

	foreach my $s (@$stylesheets) {
		my $attr = {};
		local $_ = $s;
		foreach my $k (qw(media title href)) {
			if (s/\s*$k='([^']+)'// || s/\s*$k=(\S+)//) {
				$attr->{$k} = $1;
			}
		}

		if (defined $attr->{href}) {
			$inline_ok = 0;
		} else {
			my $fn = $_;
			my ($key) = (m!([^/]+?)(?:\.css)?\z!i);
			if ($key !~ /\A[a-zA-Z0-9_\-\.]+\z/) {
				warn "ignoring $fn, non-ASCII word character\n";
				next;
			}
			open(my $fh, '<', $fn) or do {
				warn "failed to open $fn: $!\n";
				next;
			};
			my $ctime = 0;
			my $local = do { local $/; <$fh> };
			if ($local =~ /\S/) {
				$ctime = sprintf('%x',(stat($fh))[10]);
				$local = $mini->($local);
			}

			# do not let BOFHs override userContent.css:
			if ($local =~ /!\s*important\b/i) {
				warn "ignoring $fn since it uses `!important'\n";
				next;
			}

			$css_map->{$key} = $local;
			$attr->{href} = "$upfx$key.css?$ctime";
			if (defined($attr->{title})) {
				$inline_ok = 0;
			} elsif (($attr->{media}||'screen') eq 'screen') {
				$attr->{-inline} = $local;
			}
		}
		push @$links, $attr;
	}

	my $buf = "<style>$STYLE";
	if ($inline_ok) {
		my @ext; # for media=print and whatnot
		foreach my $attr (@$links) {
			if (defined(my $str = delete $attr->{-inline})) {
				$buf .= $str;
			} else {
				push @ext, $attr;
			}
		}
		$links = \@ext;
	}
	$buf .= '</style>';

	if (@$links) {
		foreach my $attr (@$links) {
			delete $attr->{-inline};
			$buf .= "<link\ntype=text/css\nrel=stylesheet";
			while (my ($k, $v) = each %$attr) {
				$v = qq{"$v"} if $v =~ /[\s=]/;
				$buf .= qq{\n$k=$v};
			}
			$buf .= ' />';
		}
		$self->{"-style-$upfx"} = $buf;
	} else {
		$self->{-style_inline} = $buf;
	}
	$self->{-css_map} = $css_map;
}

# returns an HTML fragment with <style> or <link> tags in them
# Called by WwwStream by nearly every HTML page
sub style {
	my ($self, $upfx) = @_;
	$self->{-style_inline} || $self->{"-style-$upfx"} || do {
		stylesheets_prepare($self, $upfx);
		$self->{-style_inline} || $self->{"-style-$upfx"}
	};
}

# /$INBOX/$KEY.css endpoint
# CSS is configured globally for all inboxes, but we access them on
# a per-inbox basis.  This allows administrators to setup per-inbox
# static routes to intercept the request before it hits PSGI
sub get_css ($$$) {
	my ($ctx, $inbox, $key) = @_;
	my $r404 = invalid_inbox($ctx, $inbox);
	return $r404 if $r404;
	my $self = $ctx->{www};
	my $css_map = $self->{-css_map} || stylesheets_prepare($self, '');
	my $css = $css_map->{$key};
	if (!defined($css) && $key eq 'userContent') {
		my $env = $ctx->{env};
		$css = PublicInbox::UserContent::sample($ctx->{-inbox}, $env);
	}
	defined $css or return r404();
	my $h = [ 'Content-Length', bytes::length($css),
		'Content-Type', 'text/css' ];
	PublicInbox::GitHTTPBackend::cache_one_year($h);
	[ 200, $h, [ $css ] ];
}

sub get_description {
	my ($ctx, $inbox) = @_;
	invalid_inbox($ctx, $inbox) || do {
		my $d = $ctx->{-inbox}->description . "\n";
		[ 200, [ 'Content-Length', bytes::length($d),
			'Content-Type', 'text/plain' ], [ $d ] ];
	};
}

1;

git clone https://public-inbox.org/public-inbox.git
git clone http://7fh6tueqddpjyxjmgtdiueylzoqt6pt7hec3pukyptlmohoowvhde4yd.onion/public-inbox.git