-# Copyright (C) 2016-2018 all contributors <meta@public-inbox.org>
+# Copyright (C) 2016-2020 all contributors <meta@public-inbox.org>
# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
#
# Atom body stream for which yields getline+close methods
use POSIX qw(strftime);
use Digest::SHA qw(sha1_hex);
use PublicInbox::Address;
-use PublicInbox::Hval qw(ascii_html);
-use PublicInbox::MID qw/mid_clean mid_escape/;
+use PublicInbox::Hval qw(ascii_html mid_href);
use PublicInbox::MsgTime qw(msg_timestamp);
+use PublicInbox::GzipFilter qw(gzf_maybe);
+use PublicInbox::GitAsyncCat;
-# called by PSGI server after getline:
-sub close {}
+# called by generic PSGI server after getline,
+# and also by PublicInbox::HTTP::close
+sub close { !!delete($_[0]->{http_out}) }
sub new {
my ($class, $ctx, $cb) = @_;
- $ctx->{emit_header} = 1;
$ctx->{feed_base_url} = $ctx->{-inbox}->base_url($ctx->{env});
- bless { cb => $cb || *close, ctx => $ctx }, $class;
+ $ctx->{cb} = $cb || \&close;
+ $ctx->{emit_header} = 1;
+ bless $ctx, $class;
+}
+
+# called by PublicInbox::DS::write
+sub atom_async_next {
+ my ($http) = @_; # PublicInbox::HTTP
+ atom_async_step($http->{forward});
+}
+
+# this is public-inbox-httpd-specific
+sub atom_blob_cb { # git->cat_async callback
+ my ($bref, $oid, $type, $size, $ctx) = @_;
+ my $http = $ctx->{env}->{'psgix.io'} or return; # client abort
+ my $smsg = delete $ctx->{smsg} or die 'BUG: no smsg';
+ if (!defined($oid)) {
+ # it's possible to have TOCTOU if an admin runs
+ # public-inbox-(edit|purge), just move onto the next message
+ return $http->next_step(\&atom_async_next);
+ } else {
+ $smsg->{blob} eq $oid or die "BUG: $smsg->{blob} != $oid";
+ }
+ my $buf = feed_entry($ctx, $smsg, PublicInbox::Eml->new($bref));
+ if (my $gzf = $ctx->{gzf}) {
+ $buf = $gzf->translate($buf);
+ }
+ # PublicInbox::HTTP::{Chunked,Identity}::write
+ $ctx->{http_out}->write($buf);
+
+ $http->next_step(\&atom_async_next);
+}
+
+sub atom_async_step { # this is public-inbox-httpd-specific
+ my ($ctx) = @_;
+ if (my $smsg = $ctx->{smsg} = $ctx->{cb}->($ctx)) {
+ git_async_cat($ctx->{-inbox}->git, $smsg->{blob},
+ \&atom_blob_cb, $ctx);
+ } elsif (my $out = delete $ctx->{http_out}) {
+ if (my $gzf = delete $ctx->{gzf}) {
+ $out->write($gzf->zflush);
+ }
+ $out->close;
+ }
}
sub response {
my ($class, $ctx, $code, $cb) = @_;
- [ $code, [ 'Content-Type', 'application/atom+xml' ],
- $class->new($ctx, $cb) ]
+ my $res_hdr = [ 'Content-Type' => 'application/atom+xml' ];
+ $class->new($ctx, $cb);
+ $ctx->{gzf} = gzf_maybe($res_hdr, $ctx->{env});
+ if ($ctx->{env}->{'pi-httpd.async'}) {
+ sub {
+ my ($wcb) = @_; # -httpd provided write callback
+ $ctx->{http_out} = $wcb->([200, $res_hdr]);
+ $ctx->{env}->{'psgix.io'}->{forward} = $ctx;
+ atom_async_step($ctx); # start stepping
+ };
+ } else {
+ [ $code, $res_hdr, $ctx ];
+ }
}
# called once for each message by PSGI server
sub getline {
my ($self) = @_;
- if (my $middle = $self->{cb}) {
- my $smsg = $middle->();
- return feed_entry($self, $smsg) if $smsg;
- }
- delete $self->{cb} ? '</feed>' : undef;
+ my $buf = do {
+ if (my $middle = $self->{cb}) {
+ if (my $smsg = $middle->($self)) {
+ my $eml = $self->{-inbox}->smsg_eml($smsg) or
+ return '';
+ feed_entry($self, $smsg, $eml);
+ } else {
+ undef;
+ }
+ }
+ } // (delete($self->{cb}) ? '</feed>' : undef);
+
+ # gzf may be GzipFilter, `undef' or `0'
+ my $gzf = $self->{gzf} or return $buf;
+
+ return $gzf->translate($buf) if defined $buf;
+ $self->{gzf} = 0; # next call to ->getline returns $buf (== undef)
+ $gzf->translate(undef);
}
# private
"<title$type>$title</title>";
}
+sub to_uuid ($) {
+ my ($any) = @_;
+ utf8::encode($any); # really screwed up In-Reply-To fields exist
+ $any = sha1_hex($any);
+ my $h = '[a-f0-9]';
+ my (@uuid5) = ($any =~ m!\A($h{8})($h{4})($h{4})($h{4})($h{12})!o);
+ 'urn:uuid:' . join('-', @uuid5);
+}
+
sub atom_header {
my ($ctx, $title) = @_;
my $ibx = $ctx->{-inbox};
my $search_q = $ctx->{search_query};
my $self_url = $base_url;
my $mid = $ctx->{mid};
+ my $page_id;
if (defined $mid) { # per-thread
- $self_url .= mid_escape($mid).'/t.atom';
+ $self_url .= mid_href($mid).'/t.atom';
+ $page_id = to_uuid("t\n".$mid)
} elsif (defined $search_q) {
my $query = $search_q->{'q'};
$title = title_tag("$query - search results");
$base_url .= '?' . $search_q->qs_html(x => undef);
$self_url .= '?' . $search_q->qs_html;
+ $page_id = to_uuid("q\n".$query);
} else {
$title = title_tag($ibx->description);
$self_url .= 'new.atom';
+ $page_id = "mailto:$ibx->{-primary_address}";
}
- my $mtime = (stat($ibx->{mainrepo}))[9] || time;
-
qq(<?xml version="1.0" encoding="us-ascii"?>\n) .
qq(<feed\nxmlns="http://www.w3.org/2005/Atom"\n) .
qq(xmlns:thr="http://purl.org/syndication/thread/1.0">) .
qq(<link\nrel="alternate"\ntype="text/html") .
qq(\nhref="$base_url"/>) .
qq(<link\nrel="self"\nhref="$self_url"/>) .
- qq(<id>mailto:$ibx->{-primary_address}</id>) .
- feed_updated(gmtime($mtime));
-}
-
-sub mid2uuid ($) {
- my ($mid) = @_;
- utf8::encode($mid); # really screwed up In-Reply-To fields exist
- $mid = sha1_hex($mid);
- my $h = '[a-f0-9]';
- my (@uuid5) = ($mid =~ m!\A($h{8})($h{4})($h{4})($h{4})($h{12})!o);
- 'urn:uuid:' . join('-', @uuid5);
+ qq(<id>$page_id</id>) .
+ feed_updated($ibx->modified);
}
# returns undef or string
sub feed_entry {
- my ($self, $smsg) = @_;
- my $ctx = $self->{ctx};
- my $mime = $smsg->{mime};
- my $hdr = $mime->header_obj;
- my $mid = $smsg->mid;
+ my ($ctx, $smsg, $eml) = @_;
+ my $hdr = $eml->header_obj;
+ my $mid = $smsg->{mid};
my $irt = PublicInbox::View::in_reply_to($hdr);
- my $uuid = mid2uuid($mid);
+ my $uuid = to_uuid($mid);
my $base = $ctx->{feed_base_url};
if (defined $irt) {
- my $irt_uuid = mid2uuid($irt);
- $irt = mid_escape($irt);
+ my $irt_uuid = to_uuid($irt);
+ $irt = mid_href($irt);
$irt = qq(<thr:in-reply-to\nref="$irt_uuid"\n).
qq(href="$base$irt/"/>);
} else {
$irt = '';
}
- my $href = $base . mid_escape($mid) . '/';
- my $t = msg_timestamp($hdr);
- my @t = gmtime(defined $t ? $t : time);
- my $updated = feed_updated(@t);
+ my $href = $base . mid_href($mid) . '/';
+ my $updated = feed_updated(msg_timestamp($hdr));
my $title = $hdr->header('Subject');
$title = '(no subject)' unless defined $title && $title ne '';
"<id>$uuid</id>$irt" .
qq{<content\ntype="xhtml">} .
qq{<div\nxmlns="http://www.w3.org/1999/xhtml">} .
- qq(<pre\nstyle="white-space:pre-wrap">) .
- PublicInbox::View::multipart_text_as_html($mime, $href) .
- '</pre></div></content></entry>';
+ qq(<pre\nstyle="white-space:pre-wrap">);
+ $ctx->{obuf} = \$s;
+ $ctx->{mhref} = $href;
+ PublicInbox::View::multipart_text_as_html($eml, $ctx);
+ delete $ctx->{obuf};
+ $s .= '</pre></div></content></entry>';
}
sub feed_updated {
- '<updated>' . strftime('%Y-%m-%dT%H:%M:%SZ', @_) . '</updated>';
+ '<updated>' . strftime('%Y-%m-%dT%H:%M:%SZ', gmtime(@_)) . '</updated>';
}
1;