-# 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)
package PublicInbox::Feed;
use strict;
use warnings;
use Email::Address;
use Email::MIME;
-use Date::Parse qw(strptime str2time);
+use Date::Parse qw(strptime);
use PublicInbox::Hval;
use PublicInbox::GitCatFile;
use PublicInbox::View;
+use PublicInbox::MID qw/mid_clean mid_compress mid2path/;
+use POSIX qw/strftime/;
use constant {
DATEFMT => '%Y-%m-%dT%H:%M:%SZ', # atom standard
MAX_PER_PAGE => 25, # this needs to be tunable
- PRE_WRAP => "<pre\nstyle=\"white-space:pre-wrap\">",
};
+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) };
+}
- 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),
- );
- $feed->no_generator;
+sub generate_thread_atom {
+ my ($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) = @_;
+ # try to avoid the type attribute in title:
+ $title = PublicInbox::Hval->new_oneline($title)->as_html;
+ 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\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/xml']]);
+ my $max = $ctx->{max} || MAX_PER_PAGE;
+ my $feed_opts = get_feedopts($ctx);
+ my $x = atom_header($feed_opts);
my $git = PublicInbox::GitCatFile->new($ctx->{git_dir});
each_recent_blob($ctx, sub {
- my ($add) = @_;
- add_to_feed($feed_opts, $feed, $add, $git);
+ my ($path, undef, $ts) = @_;
+ if (defined $x) {
+ $fh->write($x . '<updated>' .
+ strftime(DATEFMT, gmtime($ts)) .
+ '</updated>');
+ $x = undef;
+ }
+ add_to_feed($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, $ctx) = @_;
- require PublicInbox::Thread;
+sub emit_atom_thread {
+ my ($cb, $ctx) = @_;
+ my $res = $ctx->{srch}->get_thread($ctx->{mid});
+ return _no_thread($cb) unless $res->{total};
+ my $fh = $cb->([200, ['Content-Type' => 'application/xml']]);
+ my $feed_opts = get_feedopts($ctx);
+
+ my $html_url = $feed_opts->{atomurl} = $ctx->{self_url};
+ $html_url =~ s!/atom\z!/!;
+ $feed_opts->{url} = $html_url;
+ $feed_opts->{emit_header} = 1;
+
+ my $git = PublicInbox::GitCatFile->new($ctx->{git_dir});
+ foreach my $msg (@{$res->{msgs}}) {
+ add_to_feed($feed_opts, $fh, mid2path($msg->mid), $git);
+ }
+ $git = undef; # destroy pipes
+ _end_feed($fh);
+}
+
+sub emit_html_index {
+ my ($cb, $ctx) = @_;
+ my $fh = $cb->([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 @messages;
- my $git_dir = $ctx->{git_dir};
- my $git = PublicInbox::GitCatFile->new($git_dir);
- my ($first, $last) = each_recent_blob($ctx, sub {
- mime_load_for_sort($git, $_[0], \@messages);
- });
- $git = undef; # destroy pipes.
+ $fh->write("<html><head><title>$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 $th = PublicInbox::Thread->new(@messages);
- $th->thread;
- my $html = "<html><head><title>$title</title>" .
- '<link rel="alternate" title="Atom feed"' . "\nhref=\"" .
- $feed_opts->{atomurl} . "\"\ntype=\"application/atom+xml\"/>" .
- '</head><body>' . PRE_WRAP;
-
- # sort child messages in chronological order
- $th->order(*PublicInbox::Thread::sort_ts);
-
- # except we sort top-level messages reverse chronologically
- my $state = [ $ctx->{srch}, {}, $first, 0 ];
- for (PublicInbox::Thread::rsort_ts($th->rootset)) {
- dump_msg($_, 0, \$html, $state)
- }
+ 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 ||= {
+ ctx => $ctx,
+ seen => {},
+ first_commit => $commit,
+ anchor_idx => 0,
+ };
+
+ if ($srch) {
+ add_topic($git, $srch, $topics, $path, $ts, $u, $subj);
+ } else {
+ my $mime = do_cat_mail($git, $path) or return 0;
+ PublicInbox::View::index_entry($fh, $mime, 0, $state);
+ 1;
+ }
+ });
Email::Address->purge_cache;
+ $git = undef; # destroy pipes.
my $footer = nav_footer($ctx->{cgi}, $last, $feed_opts, $state);
if ($footer) {
my $list_footer = $ctx->{footer};
$footer .= "\n" . $list_footer if $list_footer;
- $footer = "<hr />" . PRE_WRAP . "$footer</pre>";
+ $footer = "<hr /><pre>$footer</pre>";
}
- $html . "</pre>$footer</body></html>";
+ $fh->write(dump_topics($topics)) if $topics;
+ $fh->write("$footer</body></html>");
+ $fh->close;
}
-# private subs
-
sub nav_footer {
my ($cgi, $last, $feed_opts, $state) = @_;
$cgi or return '';
my $old_r = $cgi->param('r');
my $head = ' ';
my $next = ' ';
- my $first = $state->[2];
- my $anchor = $state->[3];
+ my $first = $state->{first_commit};
+ my $anchor = $state->{anchor_idx};
if ($last) {
$next = qq!<a\nhref="?r=$last">next</a>!;
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 = $ctx->{cgi};
# revision ranges may be specified
# leave us with filenames with spaces in them..
my @cmd = ('git', "--git-dir=$ctx->{git_dir}",
qw/log --no-notes --no-color --raw -r
- --abbrev=16 --abbrev-commit/);
+ --abbrev=16 --abbrev-commit/,
+ "--format=%h%x00%ct%x00%an%x00%s%x00");
push @cmd, $range;
my $pid = open(my $log, '-|', @cmd) or
my %deleted; # only an optimization at this point
my $last;
my $nr = 0;
- my @commits = ();
- while (my $line = <$log>) {
+ my ($cur_commit, $first_commit, $last_commit);
+ my ($ts, $subj, $u);
+ while (defined(my $line = <$log>)) {
if ($line =~ /$addmsg/o) {
my $add = $1;
- next if $deleted{$add};
- $nr += $cb->($add);
+ next if $deleted{$add}; # optimization-only
+ $nr += $cb->($add, $cur_commit, $ts, $u, $subj);
if ($nr >= $max) {
$last = 1;
last;
}
} elsif ($line =~ /$delmsg/o) {
$deleted{$1} = 1;
- } elsif ($line =~ /^commit (${hex}{7,40})/o) {
- push @commits, $1;
+ } elsif ($line =~ /^${hex}{7,40}/o) {
+ ($cur_commit, $ts, $u, $subj) = split("\0", $line);
+ unless (defined $first_commit) {
+ $first_commit = $cur_commit;
+ }
}
}
if ($last) {
while (my $line = <$log>) {
- if ($line =~ /^commit (${hex}{7,40})/o) {
- push @commits, $1;
+ if ($line =~ /^(${hex}{7,40})/o) {
+ $last_commit = $1;
last;
}
}
- } else {
- push @commits, undef;
}
close $log; # we may EPIPE here
# for pagination
- ($commits[0], $commits[-1]);
+ ($first_commit, $last_commit);
}
# private functions below
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);
$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";
+ $url_base = "$base/$listname";
+ if (my $mid = $ctx->{mid}) { # per-thread feed:
+ $rv{atomurl} = "$url_base/t/$mid/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/";
my ($date) = @_;
my @t = eval { strptime($date) };
- scalar(@t) ? POSIX::strftime(DATEFMT, @t) : 0;
+ scalar(@t) ? strftime(DATEFMT, @t) : 0;
}
# 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 $mid = $mime->header_obj->header_raw('Message-ID');
+ my $header_obj = $mime->header_obj;
+ my $mid = $header_obj->header('Message-ID');
defined $mid or return 0;
$mid = PublicInbox::Hval->new_msgid($mid);
- my $href = $mid->as_href . '.html';
+ my $href = $mid->as_href . '/';
my $content = PublicInbox::View->feed_entry($mime, $fullurl . $href);
defined($content) or return 0;
+ $mime = undef;
- my $subject = mime_header($mime, 'Subject') or return 0;
+ my $date = $header_obj->header('Date');
+ $date = PublicInbox::Hval->new_oneline($date);
+ $date = feed_date($date->raw) or return 0;
- my $from = mime_header($mime, 'From') or return 0;
- my @from = Email::Address->parse($from);
- my $name = $from[0]->name;
- defined $name or $name = "";
+ my $title = mime_header($header_obj, 'Subject') or return 0;
+ $title = title_tag($title);
+
+ my $from = mime_header($header_obj, 'From') or return 0;
+ my @from = Email::Address->parse($from) or return 0;
+ my $name = PublicInbox::Hval->new_oneline($from[0]->name)->as_html;
my $email = $from[0]->address;
- defined $email or $email = "";
+ $email = PublicInbox::Hval->new_oneline($email)->as_html;
+
+ if (delete $feed_opts->{emit_header}) {
+ $fh->write(atom_header($feed_opts, $title) .
+ "<updated>$date</updated>");
+ }
+ $fh->write("<entry><author><name>$name</name><email>$email</email>" .
+ "</author>$title$date" .
+ qq{<content\ntype="xhtml">} .
+ qq{<div\nxmlns="http://www.w3.org/1999/xhtml">});
+ $fh->write($content);
- my $date = $mime->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);
+ my $midurl = $feed_opts->{midurl};
+ $fh->write(qq{</div></content><link\nhref="$midurl$href"/>}.
+ "<id>$id</id></entry>");
1;
}
-sub dump_msg {
- my ($self, $level, $html, $state) = @_;
- my $mime = $self->message;
- if ($mime) {
- $$html .= PublicInbox::View->index_entry($mime, $level, $state);
- }
- dump_msg($self->child, $level+1, $html, $state) if $self->child;
- dump_msg($self->next, $level, $html, $state) if $self->next;
-}
-
sub do_cat_mail {
my ($git, $path) = @_;
my $mime = eval {
$@ ? undef : $mime;
}
-sub mime_load_for_sort {
- my ($git, $path, $messages) = @_;
- my $mime = do_cat_mail($git, $path) or return 0;
+# accumulate recent topics if search is supported
+sub add_topic {
+ my ($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');
+ }
- my $t = eval { str2time($mime->header('Date')) };
- defined($t) or $t = 0;
- $mime->header_set('X-PI-TS', $t);
- push @$messages, $mime;
- 1;
+ my $topic = $subj = $srch->subject_normalized($subj);
+
+ # kill "[PATCH v2]" etc. for summarization
+ $topic =~ s/\A\s*\[[^\]]+\]\s*//g;
+
+ if (++$subjs->{$topic} == 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');
+ $mid = mid_compress(mid_clean($mid));
+ $u = $enc_utf8->decode($u);
+ push @$order, [ $mid, $ts, $u, $subj, $topic ];
+ return 1;
+ }
+ 0; # old topic, continue going
+}
+
+sub dump_topics {
+ my ($topics) = @_;
+ my ($order, $subjs) = @$topics;
+ my $dst = '';
+ $dst .= "\n[No recent topics]" unless (scalar @$order);
+ while (defined(my $info = shift @$order)) {
+ my ($mid, $ts, $u, $subj, $topic) = @$info;
+ my $n = delete $subjs->{$topic};
+ $mid = PublicInbox::Hval->new($mid)->as_href;
+ $subj = PublicInbox::Hval->new($subj)->as_html;
+ $u = PublicInbox::Hval->new($u)->as_html;
+ $dst .= "\n<a\nhref=\"t/$mid/#u\"><b>$subj</b></a>\n- ";
+ $ts = strftime('%Y-%m-%d %H:%M', gmtime($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;