X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FWWW.pm;h=2acb4c83b0175ad1d2f7550bda4058673a600366;hb=c3509baca0b103a0c9d034de06f3ddc5fc9426fd;hp=ee414e85d1d5a9442d6a34281df43834e0753cea;hpb=4c2c2325d2948ec5340e2fcafbee798cf568f5fd;p=public-inbox.git diff --git a/lib/PublicInbox/WWW.pm b/lib/PublicInbox/WWW.pm index ee414e85..2acb4c83 100644 --- a/lib/PublicInbox/WWW.pm +++ b/lib/PublicInbox/WWW.pm @@ -13,48 +13,76 @@ package PublicInbox::WWW; use 5.008; use strict; use warnings; +use Plack::Request; use PublicInbox::Config qw(try_cat); use URI::Escape qw(uri_escape_utf8 uri_unescape); -use constant SSOMA_URL => 'http://ssoma.public-inbox.org/'; -use constant PI_URL => 'http://public-inbox.org/'; +use constant SSOMA_URL => '//ssoma.public-inbox.org/'; +use constant PI_URL => '//public-inbox.org/'; require PublicInbox::Git; +use PublicInbox::GitHTTPBackend; our $LISTNAME_RE = qr!\A/([\w\.\-]+)!; our $MID_RE = qr!([^/]+)!; -our $END_RE = qr!(f/|T/|t/|t\.mbox(?:\.gz)?|t\.atom|raw|)!; -our $pi_config; +our $END_RE = qr!(f/|T/|t/|R/|t\.mbox(?:\.gz)?|t\.atom|raw|)!; -sub run { - my ($cgi, $method) = @_; +sub new { + my ($class, $pi_config) = @_; $pi_config ||= PublicInbox::Config->new; - my $ctx = { cgi => $cgi, pi_config => $pi_config }; - if ($method !~ /\AGET|HEAD\z/) { + 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 $cgi = Plack::Request->new($env); + my $ctx = { cgi => $cgi, pi_config => $self->{pi_config} }; + my $path_info = $cgi->path_info; + + my $method = $cgi->method; + if ($method eq 'POST' && + $path_info =~ m!$LISTNAME_RE/(git-upload-pack)\z!) { + my $path = $2; + return (invalid_list($self, $ctx, $1) || + serve_git($cgi, $ctx->{git}, $path)); + } + elsif ($method !~ /\AGET|HEAD\z/) { return r(405, 'Method Not Allowed'); } - my $path_info = $cgi->path_info; # top-level indices and feeds if ($path_info eq '/') { r404(); } elsif ($path_info =~ m!$LISTNAME_RE\z!o) { - invalid_list($ctx, $1) || r301($ctx, $1); + invalid_list($self, $ctx, $1) || r301($ctx, $1); } elsif ($path_info =~ m!$LISTNAME_RE(?:/|/index\.html)?\z!o) { - invalid_list($ctx, $1) || get_index($ctx); + invalid_list($self, $ctx, $1) || get_index($ctx); } elsif ($path_info =~ m!$LISTNAME_RE/(?:atom\.xml|new\.atom)\z!o) { - invalid_list($ctx, $1) || get_atom($ctx); + invalid_list($self, $ctx, $1) || get_atom($ctx); + } elsif ($path_info =~ m!$LISTNAME_RE/ + ($PublicInbox::GitHTTPBackend::ANY)\z!ox) { + my $path = $2; + invalid_list($self, $ctx, $1) || + serve_git($cgi, $ctx->{git}, $path); } elsif ($path_info =~ m!$LISTNAME_RE/$MID_RE/$END_RE\z!o) { - msg_page($ctx, $1, $2, $3); + msg_page($self, $ctx, $1, $2, $3); # in case people leave off the trailing slash: - } elsif ($path_info =~ m!$LISTNAME_RE/$MID_RE/(f|T|t)\z!o) { - r301($ctx, $1, $2, $3 eq 't' ? 't/#u' : $3); + } elsif ($path_info =~ m!$LISTNAME_RE/$MID_RE/(f|T|t|R)\z!o) { + my ($listname, $mid, $suffix) = ($1, $2, $3); + $suffix .= $suffix =~ /\A[tT]\z/ ? '/#u' : '/'; + r301($ctx, $listname, $mid, $suffix); # convenience redirects order matters } elsif ($path_info =~ m!$LISTNAME_RE/([^/]{2,})\z!o) { r301($ctx, $1, $2); } else { - legacy_redirects($ctx, $path_info); + legacy_redirects($self, $ctx, $path_info); } } @@ -67,12 +95,11 @@ sub preload { require Digest::SHA; require POSIX; - eval { - require PublicInbox::Search; - require PublicInbox::SearchView; - require PublicInbox::Mbox; - require IO::Compress::Gzip; - }; + foreach (qw(PublicInbox::Search PublicInbox::SearchView + PublicInbox::Mbox IO::Compress::Gzip + PublicInbox::NewsWWW PublicInbox::NewsGroup)) { + eval "require $_;"; + } } # private functions below @@ -92,25 +119,31 @@ sub r { [ $_[0], ['Content-Type' => 'text/plain'], [ join(' ', @_, "\n") ] ] } # returns undef if valid, array ref response if invalid sub invalid_list { - my ($ctx, $listname) = @_; - my $git_dir = $pi_config->get($listname, "mainrepo"); + my ($self, $ctx, $listname, $mid) = @_; + my $git_dir = $ctx->{pi_config}->get($listname, "mainrepo"); if (defined $git_dir) { $ctx->{git_dir} = $git_dir; $ctx->{git} = PublicInbox::Git->new($git_dir); $ctx->{listname} = $listname; return; } - r404(); + + # 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. + $self->news_www->call($ctx->{cgi}->{env}); } # returns undef if valid, array ref response if invalid sub invalid_list_mid { - my ($ctx, $listname, $mid) = @_; - my $ret = invalid_list($ctx, $listname, $mid); + my ($self, $ctx, $listname, $mid) = @_; + my $ret = invalid_list($self, $ctx, $listname, $mid); return $ret if $ret; $ctx->{mid} = $mid = uri_unescape($mid); if ($mid =~ /\A[a-f0-9]{40}\z/) { + # this is horiffically wasteful for legacy URLs: if ($mid = mid2blob($ctx)) { require Email::Simple; use PublicInbox::MID qw/mid_clean/; @@ -186,6 +219,19 @@ sub get_full_html { [ PublicInbox::View::msg_html($ctx, $mime, undef, $foot)] ]; } +# /$LISTNAME/$MESSAGE_ID/R/ -> HTML content (fullquotes) +sub get_reply_html { + my ($ctx) = @_; + my $x = mid2blob($ctx) or return r404($ctx); + + require PublicInbox::View; + my $foot = footer($ctx); + require Email::MIME; + my $hdr = Email::MIME->new($x)->header_obj; + [ 200, [ 'Content-Type' => 'text/html; charset=UTF-8' ], + [ PublicInbox::View::msg_reply($ctx, $hdr, $foot)] ]; +} + # /$LISTNAME/$MESSAGE_ID/t/ sub get_thread { my ($ctx, $flat) = @_; @@ -196,15 +242,10 @@ sub get_thread { PublicInbox::View::thread_html($ctx, $foot, $srch); } -sub self_url { - my ($cgi) = @_; - ref($cgi) eq 'CGI' ? $cgi->self_url : $cgi->uri->as_string; -} - sub ctx_get { my ($ctx, $key) = @_; my $val = $ctx->{$key}; - (defined $val && $val ne '') or die "BUG: bad ctx, $key unusable\n"; + (defined $val && $val ne '') or die "BUG: bad ctx, $key unusable"; $val; } @@ -229,18 +270,22 @@ sub footer { my $urls = try_cat("$git_dir/cloneurl"); my @urls = split(/\r?\n/, $urls || ''); - my $nurls = scalar @urls; - if ($nurls == 0) { - $urls = '($GIT_DIR/cloneurl missing)'; - } elsif ($nurls == 1) { - $urls = "git URL for ssoma: ' . $urls[0]; + my %seen = map { $_ => 1 } @urls; + my $cgi = $ctx->{cgi}; + my $http = $cgi->base->as_string . $listname; + $seen{$http} or unshift @urls, $http; + my $ssoma_url = PublicInbox::Hval::prurl($cgi->{env}, SSOMA_URL); + if (scalar(@urls) == 1) { + $urls = "URL for ssoma or git clone --mirror \$URL :) . + $urls[0]; } else { - $urls = "git URLs for ssoma:\n" . join("\n", map { "\t$_" } @urls); + $urls = "URLs for ssoma or git clone --mirror \$URL\n) . + join("\n", map { "\t$_" } @urls); } - my $addr = $pi_config->get($listname, 'address'); + my $addr = $ctx->{pi_config}->get($listname, 'address'); if (ref($addr) eq 'ARRAY') { $addr = $addr->[0]; # first address is primary } @@ -249,7 +294,9 @@ sub footer { $ctx->{footer} = join("\n", '- ' . $desc, - "A public-inbox, ' . + "A {cgi}->{env}, PI_URL) . + '">public-inbox, ' . 'anybody may post in plain-text (not HTML):', $addr, $urls @@ -293,13 +340,13 @@ sub get_thread_mbox { sub get_thread_atom { my ($ctx) = @_; searcher($ctx) or return need_search($ctx); - $ctx->{self_url} = self_url($ctx->{cgi}); + $ctx->{self_url} = $ctx->{cgi}->uri->as_string; require PublicInbox::Feed; PublicInbox::Feed::generate_thread_atom($ctx); } sub legacy_redirects { - my ($ctx, $path_info) = @_; + my ($self, $ctx, $path_info) = @_; # single-message pages if ($path_info =~ m!$LISTNAME_RE/m/(\S+)/\z!o) { @@ -344,13 +391,13 @@ sub legacy_redirects { # some Message-IDs have slashes in them and the HTTP server # may try to be clever and unescape them :< } elsif ($path_info =~ m!$LISTNAME_RE/(\S+/\S+)/$END_RE\z!o) { - msg_page($ctx, $1, $2, $3); + msg_page($self, $ctx, $1, $2, $3); # in case people leave off the trailing slash: } elsif ($path_info =~ m!$LISTNAME_RE/(\S+/\S+)/(f|T|t)\z!o) { r301($ctx, $1, $2, $3 eq 't' ? 't/#u' : $3); } else { - r404(); + $self->news_www->call($ctx->{cgi}->{env}); } } @@ -358,15 +405,11 @@ sub r301 { my ($ctx, $listname, $mid, $suffix) = @_; my $cgi = $ctx->{cgi}; my $url; - if (ref($cgi) eq 'CGI') { - $url = $cgi->url(-base) . '/'; - } else { - $url = $cgi->base->as_string; - } - - $url .= $listname . '/'; + my $qs = $cgi->env->{QUERY_STRING}; + $url = $cgi->base->as_string . $listname . '/'; $url .= (uri_escape_utf8($mid) . '/') if (defined $mid); $url .= $suffix if (defined $suffix); + $url .= "?$qs" if $qs ne ''; [ 301, [ Location => $url, 'Content-Type' => 'text/plain' ], @@ -374,18 +417,32 @@ sub r301 { } sub msg_page { - my ($ctx, $list, $mid, $e) = @_; - unless (invalid_list_mid($ctx, $list, $mid)) { - '' eq $e and return get_mid_html($ctx); - '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'); - 'T/' eq $e and return get_thread($ctx, 1); - 'raw' eq $e and return get_mid_txt($ctx); - 'f/' eq $e and return get_full_html($ctx); - } + my ($self, $ctx, $list, $mid, $e) = @_; + my $ret; + $ret = invalid_list_mid($self, $ctx, $list, $mid) and return $ret; + '' eq $e and return get_mid_html($ctx); + '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'); + 'T/' eq $e and return get_thread($ctx, 1); + 'raw' eq $e and return get_mid_txt($ctx); + 'f/' eq $e and return get_full_html($ctx); + 'R/' eq $e and return get_reply_html($ctx); r404($ctx); } +sub serve_git { + my ($cgi, $git, $path) = @_; + PublicInbox::GitHTTPBackend::serve($cgi, $git, $path); +} + +sub news_www { + my ($self) = @_; + my $nw = $self->{news_www}; + return $nw if $nw; + require PublicInbox::NewsWWW; + $self->{news_www} = PublicInbox::NewsWWW->new($self->{pi_config}); +} + 1;