-# Copyright (C) 2013, Eric Wong <normalperson@yhbt.net> and all contributors
+# Copyright (C) 2013-2015 all contributors <meta@public-inbox.org>
# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+#
+# Used for generating Atom feeds for web-accessible mailing list archives.
package PublicInbox::Feed;
use strict;
use warnings;
use Email::Address;
use Email::MIME;
use Date::Parse qw(strptime);
-use PublicInbox::Hval;
-use PublicInbox::GitCatFile;
+use PublicInbox::Hval qw/ascii_html/;
+use PublicInbox::Git;
use PublicInbox::View;
-use PublicInbox::MID qw/mid_clean mid_compressed/;
+use PublicInbox::MID qw/mid_clean mid2path/;
+use POSIX qw/strftime/;
use constant {
- DATEFMT => '%Y-%m-%dT%H:%M:%SZ', # atom standard
+ DATEFMT => '%Y-%m-%dT%H:%M:%SZ', # Atom standard
MAX_PER_PAGE => 25, # this needs to be tunable
};
-use Encode qw/find_encoding/;
-my $enc_utf8 = find_encoding('UTF-8');
-
# main function
sub generate {
- my ($class, $ctx) = @_;
- require XML::Atom::SimpleFeed;
- require POSIX;
- my $max = $ctx->{max} || MAX_PER_PAGE;
+ my ($ctx) = @_;
+ sub { emit_atom($_[0], $ctx) };
+}
+
+sub generate_thread_atom {
+ my ($ctx) = @_;
+ sub { emit_atom_thread($_[0], $ctx) };
+}
+
+sub generate_html_index {
+ my ($ctx) = @_;
+ sub { emit_html_index($_[0], $ctx) };
+}
+
+# private subs
+
+sub title_tag {
+ my ($title) = @_;
+ $title =~ tr/\t\n / /s; # squeeze spaces
+ # try to avoid the type attribute in title:
+ $title = ascii_html($title);
+ my $type = index($title, '&') >= 0 ? "\ntype=\"html\"" : '';
+ "<title$type>$title</title>";
+}
+sub atom_header {
+ my ($feed_opts, $title) = @_;
+
+ $title = title_tag($feed_opts->{description}) unless (defined $title);
+
+ qq(<?xml version="1.0" encoding="us-ascii"?>\n) .
+ qq{<feed\nxmlns="http://www.w3.org/2005/Atom">} .
+ qq{$title} .
+ qq(<link\nrel="alternate"\ntype="text/html") .
+ qq(\nhref="$feed_opts->{url}"/>) .
+ qq(<link\nrel="self"\nhref="$feed_opts->{atomurl}"/>) .
+ qq(<id>mailto:$feed_opts->{id_addr}</id>);
+}
+
+sub emit_atom {
+ my ($cb, $ctx) = @_;
+ my $fh = $cb->([ 200, ['Content-Type' => 'application/atom+xml']]);
+ my $max = $ctx->{max} || MAX_PER_PAGE;
my $feed_opts = get_feedopts($ctx);
- my $addr = $feed_opts->{address};
- $addr = $addr->[0] if ref($addr);
- my $feed = XML::Atom::SimpleFeed->new(
- title => $feed_opts->{description} || "unnamed feed",
- link => $feed_opts->{url} || "http://example.com/",
- link => {
- rel => 'self',
- href => $feed_opts->{atomurl} ||
- "http://example.com/atom.xml",
- },
- id => 'mailto:' . ($addr || 'public-inbox@example.com'),
- updated => POSIX::strftime(DATEFMT, gmtime),
- );
- $feed->no_generator;
-
- my $git = PublicInbox::GitCatFile->new($ctx->{git_dir});
+ my $x = atom_header($feed_opts);
+ my $git = $ctx->{git} ||= PublicInbox::Git->new($ctx->{git_dir});
each_recent_blob($ctx, sub {
- my ($add, undef) = @_;
- add_to_feed($feed_opts, $feed, $add, $git);
+ my ($path, undef, $ts) = @_;
+ if (defined $x) {
+ $fh->write($x . feed_updated(undef, $ts));
+ $x = undef;
+ }
+ add_to_feed($feed_opts, $fh, $path, $git);
});
- $git = undef; # destroy pipes
+ end_feed($fh);
+}
+
+sub _no_thread {
+ my ($cb) = @_;
+ my $fh = $cb->([404, ['Content-Type' => 'text/plain']]);
+ $fh->write("No feed found for thread\n");
+ $fh->close;
+}
+
+sub end_feed {
+ my ($fh) = @_;
Email::Address->purge_cache;
- $feed->as_string;
+ $fh->write('</feed>');
+ $fh->close;
}
-sub generate_html_index {
- my ($class, $ctx) = @_;
+sub emit_atom_thread {
+ my ($cb, $ctx) = @_;
+ my $res = $ctx->{srch}->get_thread($ctx->{mid});
+ return _no_thread($cb) unless $res->{total};
+ my $fh = $cb->([200, ['Content-Type' => 'application/atom+xml']]);
+ my $feed_opts = get_feedopts($ctx);
+
+ my $html_url = $feed_opts->{atomurl} = $ctx->{self_url};
+ $html_url =~ s!/t\.atom\z!/!;
+ $feed_opts->{url} = $html_url;
+ $feed_opts->{emit_header} = 1;
+
+ my $git = $ctx->{git} ||= PublicInbox::Git->new($ctx->{git_dir});
+ foreach my $msg (@{$res->{msgs}}) {
+ add_to_feed($feed_opts, $fh, mid2path($msg->mid), $git);
+ }
+ end_feed($fh);
+}
+
+sub emit_html_index {
+ my ($res, $ctx) = @_;
+ my $fh = $res->([200,['Content-Type'=>'text/html; charset=UTF-8']]);
my $max = $ctx->{max} || MAX_PER_PAGE;
my $feed_opts = get_feedopts($ctx);
- my $title = $feed_opts->{description} || '';
- $title = PublicInbox::Hval->new_oneline($title)->as_html;
- my $atom_url = $feed_opts->{atomurl};
+ my $title = ascii_html($feed_opts->{description} || '');
+ my ($footer, $param, $last);
+ my $state = { ctx => $ctx, seen => {}, anchor_idx => 0, fh => $fh };
+ my $srch = $ctx->{srch};
- my $html = "<html><head><title>$title</title>" .
- "<link\nrel=alternate\ntitle=\"Atom feed\"\n".
- "href=\"$atom_url\"\"\ntype=\"application/atom+xml\"/>" .
- '</head><body>' . PublicInbox::View::PRE_WRAP .
- "<b>$title</b> (<a\nhref=\"$atom_url\">Atom feed</a>)\n";
+ my $top = "<b>$title</b> (<a\nhref=\"new.atom\">Atom feed</a>)";
- my $state;
- my $git = PublicInbox::GitCatFile->new($ctx->{git_dir});
- my $topics;
- my $srch = $ctx->{srch};
- $srch and $topics = [ [], {} ];
- my (undef, $last) = each_recent_blob($ctx, sub {
- my ($path, $commit, $ts, $u, $subj) = @_;
- $state ||= [ undef, {}, $commit, 0 ];
+ if ($srch) {
+ $top = qq{<form\naction=""><pre>$top} .
+ qq{ <input\nname=q\ntype=text />} .
+ qq{<input\ntype=submit\nvalue=search />} .
+ q{</pre></form><pre>}
+ } else {
+ $top = '<pre>' . $top . "\n";
+ }
- if ($srch) {
- add_topic($git, $srch, $topics, $path, $ts, $u, $subj);
- } else {
- my $mime = do_cat_mail($git, $path) or return 0;
- $html .=
- PublicInbox::View->index_entry($mime, 0, $state);
- 1;
- }
- });
- Email::Address->purge_cache;
- $git = undef; # destroy pipes.
+ $fh->write("<html><head><title>$title</title>" .
+ "<link\nrel=alternate\ntitle=\"Atom feed\"\n".
+ "href=\"new.atom\"\ntype=\"application/atom+xml\"/>" .
+ PublicInbox::Hval::STYLE .
+ "</head><body>$top");
- my $footer = nav_footer($ctx->{cgi}, $last, $feed_opts, $state);
+ # if the 'r' query parameter is given, it is a legacy permalink
+ # which we must continue supporting:
+ my $cgi = $ctx->{cgi};
+ if ($cgi && !$cgi->param('r') && $srch) {
+ $state->{srch} = $srch;
+ $last = PublicInbox::View::emit_index_topics($state);
+ $param = 'o';
+ } else {
+ $last = emit_index_nosrch($ctx, $state);
+ $param = 'r';
+ }
+ $footer = nav_footer($cgi, $last, $feed_opts, $state, $param);
if ($footer) {
my $list_footer = $ctx->{footer};
- $footer .= "\n" . $list_footer if $list_footer;
+ $footer .= "\n\n" . $list_footer if $list_footer;
$footer = "<hr /><pre>$footer</pre>";
}
- dump_topics(\$html, $topics) if $topics;
- $html .= "$footer</body></html>";
+ $fh->write("$footer</body></html>");
+ $fh->close;
}
-# private subs
+sub emit_index_nosrch {
+ my ($ctx, $state) = @_;
+ my $git = $ctx->{git} ||= PublicInbox::Git->new($ctx->{git_dir});
+ my (undef, $last) = each_recent_blob($ctx, sub {
+ my ($path, $commit, $ts, $u, $subj) = @_;
+ $state->{first} ||= $commit;
+
+ my $mime = do_cat_mail($git, $path) or return 0;
+ PublicInbox::View::index_entry($mime, 0, $state);
+ 1;
+ });
+ Email::Address->purge_cache;
+ $last;
+}
sub nav_footer {
- my ($cgi, $last, $feed_opts, $state) = @_;
+ my ($cgi, $last, $feed_opts, $state, $param) = @_;
$cgi or return '';
- my $old_r = $cgi->param('r');
+ my $old_r = $cgi->param($param);
my $head = ' ';
my $next = ' ';
- # $state = [ undef, {}, $first_commit, $last_anchor ];
- my $first = $state->[2];
- my $anchor = $state->[3];
+ my $first = $state->{first};
+ my $anchor = $state->{anchor_idx};
if ($last) {
- $next = qq!<a\nhref="?r=$last">next</a>!;
+ $next = qq!<a\nhref="?$param=$last"\nrel=next>next</a>!;
}
if ($old_r) {
$head = $cgi->path_info;
$head = qq!<a\nhref="$head">head</a>!;
}
- my $atom = "<a\nhref=\"$feed_opts->{atomurl}\">atom</a>";
- my $permalink = "<a\nhref=\"?r=$first\">permalink</a>";
- "<a\nname=\"s$anchor\">page:</a> $next $head $atom $permalink";
+ my $atom = "<a\nhref=\"$feed_opts->{atomurl}\">Atom feed</a>";
+ "<a\nname=\"s$anchor\">page:</a> $next $head $atom";
}
sub each_recent_blob {
my $hex = '[a-f0-9]';
my $addmsg = qr!^:000000 100644 \S+ \S+ A\t(${hex}{2}/${hex}{38})$!;
my $delmsg = qr!^:100644 000000 \S+ \S+ D\t(${hex}{2}/${hex}{38})$!;
- my $refhex = qr/${hex}{4,40}(?:~\d+)?/;
+ my $refhex = qr/(?:HEAD|${hex}{4,40})(?:~\d+)?/;
my $cgi = $ctx->{cgi};
# revision ranges may be specified
# get recent messages
# we could use git log -z, but, we already know ssoma will not
# leave us with filenames with spaces in them..
- my @cmd = ('git', "--git-dir=$ctx->{git_dir}",
- qw/log --no-notes --no-color --raw -r
- --abbrev=16 --abbrev-commit/,
- "--format=%h%x00%ct%x00%an%x00%s%x00");
- push @cmd, $range;
-
- my $pid = open(my $log, '-|', @cmd) or
- die('open `'.join(' ', @cmd) . " pipe failed: $!\n");
+ my $git = $ctx->{git} ||= PublicInbox::Git->new($ctx->{git_dir});
+ my $log = $git->popen(qw/log --no-notes --no-color --raw -r
+ --abbrev=16 --abbrev-commit/,
+ "--format=%h%x00%ct%x00%an%x00%s%x00",
+ $range);
my %deleted; # only an optimization at this point
my $last;
my $nr = 0;
}
}
- close $log; # we may EPIPE here
# for pagination
($first_commit, $last_commit);
}
my %rv;
if (open my $fh, '<', "$ctx->{git_dir}/description") {
chomp($rv{description} = <$fh>);
- close $fh;
+ } else {
+ $rv{description} = '($GIT_DIR/description missing)';
}
- if ($pi_config && defined $listname && length $listname) {
- foreach my $key (qw(address)) {
- $rv{$key} = $pi_config->get($listname, $key) || "";
- }
+ if ($pi_config && defined $listname && $listname ne '') {
+ my $addr = $pi_config->get($listname, 'address') || "";
+ $rv{address} = $addr;
+ $addr = $addr->[0] if ref($addr);
+ $rv{id_addr} = $addr;
}
+ $rv{id_addr} ||= 'public-inbox@example.com';
+
my $url_base;
if ($cgi) {
- my $path_info = $cgi->path_info;
- my $base;
- if (ref($cgi) eq 'CGI') {
- $base = $cgi->url(-base);
- } else {
- $base = $cgi->base->as_string;
- $base =~ s!/\z!!;
- }
- $url_base = $path_info;
- if ($url_base =~ s!/(?:|index\.html)?\z!!) {
- $rv{atomurl} = "$base$url_base/atom.xml";
+ $url_base = $cgi->base->as_string . $listname;
+ if (my $mid = $ctx->{mid}) { # per-thread feed:
+ $rv{atomurl} = "$url_base/$mid/t.atom";
} else {
- $url_base =~ s!/atom\.xml\z!!;
- $rv{atomurl} = $base . $path_info;
- $url_base = $base . $url_base; # XXX is this needed?
+ $rv{atomurl} = "$url_base/new.atom";
}
} else {
$url_base = "http://example.com";
- $rv{atomurl} = "$url_base/atom.xml";
+ $rv{atomurl} = "$url_base/new.atom";
}
$rv{url} ||= "$url_base/";
- $rv{midurl} = "$url_base/m/";
- $rv{fullurl} = "$url_base/f/";
+ $rv{midurl} = "$url_base/";
\%rv;
}
-sub mime_header {
- my ($mime, $name) = @_;
- PublicInbox::Hval->new_oneline($mime->header($name))->raw;
-}
-
-sub feed_date {
- my ($date) = @_;
- my @t = eval { strptime($date) };
+sub feed_updated {
+ my ($date, $ts) = @_;
+ my @t = eval { strptime($date) } if defined $date;
+ @t = gmtime($ts || time) unless scalar @t;
- scalar(@t) ? POSIX::strftime(DATEFMT, @t) : 0;
+ '<updated>' . strftime(DATEFMT, @t) . '</updated>';
}
# returns 0 (skipped) or 1 (added)
sub add_to_feed {
- my ($feed_opts, $feed, $add, $git) = @_;
+ my ($feed_opts, $fh, $add, $git) = @_;
my $mime = do_cat_mail($git, $add) or return 0;
- my $midurl = $feed_opts->{midurl} || 'http://example.com/m/';
- my $fullurl = $feed_opts->{fullurl} || 'http://example.com/f/';
+ my $url = $feed_opts->{url};
+ my $midurl = $feed_opts->{midurl};
my $header_obj = $mime->header_obj;
- my $mid = $header_obj->header('Message-ID');
+ my $mid = $header_obj->header_raw('Message-ID');
defined $mid or return 0;
$mid = PublicInbox::Hval->new_msgid($mid);
- my $href = $mid->as_href . '.html';
- my $content = PublicInbox::View->feed_entry($mime, $fullurl . $href);
+ my $href = $mid->as_href;
+ my $content = PublicInbox::View->feed_entry($mime);
defined($content) or return 0;
$mime = undef;
- my $subject = mime_header($header_obj, 'Subject') or return 0;
+ my $date = $header_obj->header('Date');
+ my $updated = feed_updated($date);
+
+ my $title = $header_obj->header('Subject');
+ defined $title or return 0;
+ $title = title_tag($title);
- my $from = mime_header($header_obj, 'From') or return 0;
- my @from = Email::Address->parse($from);
- my $name = $from[0]->name;
- defined $name or $name = "";
+ my $from = $header_obj->header('From') or return 0;
+ my @from = Email::Address->parse($from) or return 0;
+ my $name = ascii_html($from[0]->name);
my $email = $from[0]->address;
- defined $email or $email = "";
+ $email = ascii_html($email);
+
+ if (delete $feed_opts->{emit_header}) {
+ $fh->write(atom_header($feed_opts, $title) . $updated);
+ }
+ $fh->write("<entry><author><name>$name</name><email>$email</email>" .
+ "</author>$title$updated" .
+ qq{<content\ntype="xhtml">} .
+ qq{<div\nxmlns="http://www.w3.org/1999/xhtml">});
+ $fh->write($content);
- my $date = $header_obj->header('Date');
- $date = PublicInbox::Hval->new_oneline($date);
- $date = feed_date($date->raw) or return 0;
$add =~ tr!/!!d;
my $h = '[a-f0-9]';
my (@uuid5) = ($add =~ m!\A($h{8})($h{4})($h{4})($h{4})($h{12})!o);
-
- $feed->add_entry(
- author => { name => $name, email => $email },
- title => $subject,
- updated => $date,
- content => { type => 'xhtml', content => $content },
- link => $midurl . $href,
- id => 'urn:uuid:' . join('-', @uuid5),
- );
+ my $id = 'urn:uuid:' . join('-', @uuid5);
+ $fh->write(qq!</div></content><link\nhref="$midurl$href/"/>!.
+ "<id>$id</id></entry>");
1;
}
$@ ? undef : $mime;
}
-# accumulate recent topics if search is supported
-sub add_topic {
- my ($git, $srch, $topics, $path, $ts, $u, $subj) = @_;
- my ($order, $subjs) = @$topics;
- my $header_obj;
-
- # legacy ssoma did not set commit titles based on Subject
- $subj = $enc_utf8->decode($subj);
- if ($subj eq 'mda') {
- my $mime = do_cat_mail($git, $path) or return 0;
- $header_obj = $mime->header_obj;
- $subj = mime_header($header_obj, 'Subject');
- }
-
- $subj = $srch->subject_normalized($subj);
- if (++$subjs->{$subj} == 1) {
- unless ($header_obj) {
- my $mime = do_cat_mail($git, $path) or return 0;
- $header_obj = $mime->header_obj;
- }
- my $mid = $header_obj->header('Message-ID');
- $mid = mid_compressed(mid_clean($mid));
- $u = $enc_utf8->decode($u);
- push @$order, [ $mid, $ts, $u, $subj ];
- return 1;
- }
- 0; # old topic, continue going
-}
-
-sub dump_topics {
- my ($dst, $topics) = @_;
- my ($order, $subjs) = @$topics;
- $$dst .= "\n[No recent topics]" unless (scalar @$order);
- while (defined(my $info = shift @$order)) {
- my ($mid, $ts, $u, $subj) = @$info;
- my $n = delete $subjs->{$subj};
- $mid = PublicInbox::Hval->new($mid)->as_href;
- $subj = PublicInbox::Hval->new($subj)->as_html;
- $u = PublicInbox::Hval->new($u)->as_html;
- $$dst .= "\n<a\nhref=\"t/$mid.html#u\"><b>$subj</b></a>\n- ";
- $ts = POSIX::strftime('%Y-%m-%d %H:%M', gmtime($ts));
- if ($n == 1) {
- $$dst .= "created by $u @ $ts UTC\n"
- } else {
- # $n isn't the total number of posts on the topic,
- # just the number of posts in the current "git log"
- # window, so leave it unlabeled
- $$dst .= "updated by $u @ $ts UTC ($n)\n"
- }
- }
- $$dst .= '</pre>'
-}
-
1;