1 # Copyright (C) 2016-2019 all contributors <meta@public-inbox.org>
2 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
4 # used for displaying help texts and other non-mail content
5 package PublicInbox::WwwText;
8 use bytes (); # only for bytes::length
9 use PublicInbox::Linkify;
10 use PublicInbox::WwwStream;
11 use PublicInbox::Hval qw(ascii_html);
12 use URI::Escape qw(uri_escape_utf8);
13 our $QP_URL = 'https://xapian.org/docs/queryparser.html';
14 our $WIKI_URL = 'https://en.wikipedia.org/wiki';
16 require PublicInbox::HlMod;
17 PublicInbox::HlMod->new
20 # /$INBOX/_/text/$KEY/ # KEY may contain slashes
21 # For now, "help" is the only supported $KEY
26 $key = 'help' if !defined $key; # this 302s to _/text/help/
28 # get the raw text the same way we get mboxrds
29 my $raw = ($key =~ s!/raw\z!!);
30 my $have_tslash = ($key =~ s!/\z!!) if !$raw;
33 my $hdr = [ 'Content-Type', 'text/plain', 'Content-Length', undef ];
34 if (!_default_text($ctx, $key, $hdr, \$txt)) {
36 $txt = "404 Not Found ($key)\n";
39 $hdr->[3] = bytes::length($txt);
40 return [ $code, $hdr, [ $txt ] ]
43 # enforce trailing slash for "wget -r" compatibility
44 if (!$have_tslash && $code == 200) {
45 my $url = $ctx->{-inbox}->base_url($ctx->{env});
46 $url .= "_/text/$key/";
48 return [ 302, [ 'Content-Type', 'text/plain',
50 [ "Redirecting to $url\n" ] ];
53 # Follow git commit message conventions,
54 # first line is the Subject/title
55 my ($title) = ($txt =~ /\A([^\n]*)/s);
57 $ctx->{-title_html} = ascii_html($title);
58 my $nslash = ($key =~ tr!/!/!);
59 $ctx->{-upfx} = '../../../' . ('../' x $nslash);
60 PublicInbox::WwwStream->response($ctx, $code, \&_do_linkify);
65 return unless $nr == 1;
66 my $l = PublicInbox::Linkify->new;
67 my $txt = delete $ctx->{txt};
70 $hl->do_hl_text($txt);
72 $$txt = ascii_html($$txt);
74 '<pre>' . $l->linkify_2($$txt) . '</pre>';
77 sub _srch_prefix ($$) {
78 my ($srch, $txt) = @_;
81 my $help = $srch->help;
83 for ($i = 0; $i < @$help; $i += 2) {
84 my $pfx = $help->[$i];
86 $pad = $n if $n > $pad;
88 $htxt .= $help->[$i + 1];
92 my $padding = ' ' x ($pad + 8);
93 $htxt =~ s/^/$padding/gms;
94 $htxt =~ s/^$padding(\S+)\0/" $1".
95 (' ' x ($pad - length($1)))/egms;
96 $htxt =~ s/\f\n/\n/gs;
101 sub _colors_help ($$) {
102 my ($ctx, $txt) = @_;
103 my $ibx = $ctx->{-inbox};
104 my $env = $ctx->{env};
105 my $base_url = $ibx->base_url($env);
106 $$txt .= "color customization for $base_url\n";
109 public-inbox provides a stable set of CSS classes for users to
110 customize colors for highlighting diffs and code.
112 Users of browsers such as dillo, Firefox, or some browser
113 extensions may start by downloading the following sample CSS file
114 to control the colors they see:
116 ${base_url}userContent.css
122 $$txt .= PublicInbox::UserContent::sample($ibx, $env) . "```\n";
125 # git-config section names are quoted in the config file, so escape them
128 $name =~ s/\\/\\\\/g;
133 sub URI_PATH () { '^A-Za-z0-9\-\._~/' }
135 # n.b. this is a perfect candidate for memoization
136 sub inbox_config ($$$) {
137 my ($ctx, $hdr, $txt) = @_;
138 my $ibx = $ctx->{-inbox};
139 push @$hdr, 'Content-Disposition', 'inline; filename=inbox.config';
140 my $name = dq_escape($ibx->{name});
142 ; example public-inbox config snippet for "$name"
143 ; see public-inbox-config(5) manpage for more details:
144 ; https://public-inbox.org/public-inbox-config.html
145 [publicinbox "$name"]
146 inboxdir = /path/to/top-level-inbox
147 ; note: public-inbox before v1.2.0 used "mainrepo"
148 ; instead of "inboxdir", both remain supported after 1.2
149 mainrepo = /path/to/top-level-inbox
151 for my $k (qw(address listid)) {
152 defined(my $v = $ibx->{$k}) or next;
153 $$txt .= "\t$k = $_\n" for @$v;
156 for my $k (qw(filter infourl newsgroup obfuscate replyto watchheader)) {
157 defined(my $v = $ibx->{$k}) or next;
158 $$txt .= "\t$k = $v\n";
160 $$txt .= "\tnntpmirror = $_\n" for (@{$ibx->nntp_url});
162 # note: this doesn't preserve cgitrc layout, since we parse cgitrc
163 # and drop the original structure
164 if (defined(my $cr = $ibx->{coderepo})) {
165 $$txt .= "\tcoderepo = $_\n" for @$cr;
167 my $pi_config = $ctx->{www}->{pi_config};
168 for my $cr_name (@$cr) {
169 my $url = $pi_config->{"coderepo.$cr_name.cgiturl"};
170 my $path = "/path/to/$cr_name";
171 $cr_name = dq_escape($cr_name);
173 $$txt .= qq([coderepo "$cr_name"]\n);
176 if ($path !~ m![a-z0-9_/\.\-]!i) {
177 $cpath = dq_escape($cpath);
179 $$txt .= qq(\t; git clone $url "$cpath"\n);
181 $$txt .= "\tdir = $path\n";
182 $$txt .= "\tcgiturl = https://example.com/";
183 $$txt .= uri_escape_utf8($cr_name, URI_PATH)."\n";
189 sub _default_text ($$$$) {
190 my ($ctx, $key, $hdr, $txt) = @_;
191 return _colors_help($ctx, $txt) if $key eq 'color';
192 return inbox_config($ctx, $hdr, $txt) if $key eq 'config';
193 return if $key ne 'help'; # TODO more keys?
195 my $ibx = $ctx->{-inbox};
196 my $base_url = $ibx->base_url($ctx->{env});
197 $$txt .= "public-inbox help for $base_url\n";
203 public-inbox uses Message-ID identifiers in URLs.
204 One may look up messages by substituting Message-IDs
205 (without the leading '<' or trailing '>') into the URL.
206 Forward slash ('/') characters in the Message-IDs
207 need to be escaped as "%2F" (without quotes).
209 Thus, it is possible to retrieve any message by its
210 Message-ID by going to:
212 $base_url<Message-ID>/
214 (without the '<' or '>')
216 Message-IDs are described at:
222 # n.b. we use the Xapian DB for any regeneratable,
223 # order-of-arrival-independent data.
224 my $srch = $ibx->search;
230 This public-inbox has search functionality provided by Xapian.
232 It supports typical AND, OR, NOT, '+', '-' queries present
233 in other search engines.
235 We also support search prefixes to limit the scope of the
236 search to certain fields.
238 Prefixes supported in this installation include:
241 _srch_prefix($srch, $txt);
245 Most prefixes are probabilistic, meaning they support stemming
246 and wildcards ('*'). Ranges (such as 'd:') and boolean prefixes
247 do not support stemming or wildcards.
248 The upstream Xapian query parser documentation fully explains
256 Message threading is enabled for this public-inbox,
257 additional endpoints for message threads are available:
259 * $base_url<Message-ID>/T/#u
261 Loads the thread belonging to the given <Message-ID>
262 in flat chronological order. The "#u" anchor
263 focuses the browser on the given <Message-ID>.
265 * $base_url<Message-ID>/t/#u
267 Loads the thread belonging to the given <Message-ID>
268 in threaded order with nesting. For deep threads,
269 this requires a wide display or horizontal scrolling.
271 Both of these HTML endpoints are suitable for offline reading
272 using the thread overview at the bottom of each page.
274 Users of feed readers may follow a particular thread using:
276 * $base_url<Message-ID>/t.atom
278 Which loads the thread in Atom Syndication Standard
279 described at Wikipedia and RFC4287:
281 $WIKI_URL/Atom_(standard)
282 https://tools.ietf.org/html/rfc4287
284 Atom Threading Extensions (RFC4685) is supported:
286 https://tools.ietf.org/html/rfc4685
288 Finally, the gzipped mbox for a thread is available for
289 downloading and importing into your favorite mail client:
291 * $base_url<Message-ID>/t.mbox.gz
293 We use the mboxrd variant of the mbox format described
301 This help text is maintained by public-inbox developers
302 reachable via plain-text email at: meta\@public-inbox.org
305 # TODO: support admin contact info in ~/.public-inbox/config