huge refactor of encoding handling

[public-inbox.git] / public-inbox.cgi

diff --git a/public-inbox.cgi b/public-inbox.cgi
index 03202c61736afa7887887ab7a8786a8584bc2105..3c51c314ebe6c4452e84d8fbd3849357c559456f 100755 (executable)
--- a/public-inbox.cgi
+++ b/public-inbox.cgi
@@ -12,10 +12,11 @@
 use 5.008;
 use strict;
 use warnings;
-use CGI qw(:cgi :escapeHTML -nosticky); # PSGI/FastCGI/mod_perl compat
-use Encode qw(decode_utf8);
+use CGI qw(:cgi -nosticky); # PSGI/FastCGI/mod_perl compat
+use Encode qw(find_encoding);
 use PublicInbox::Config;
 use URI::Escape qw(uri_escape uri_unescape);
+my $enc_utf8 = find_encoding('UTF-8');
 our $LISTNAME_RE = qr!\A/([\w\.\-]+)!;
 our $pi_config;
 BEGIN {
@@ -56,7 +57,7 @@ sub main {
        if ($cgi->request_method !~ /\AGET|HEAD\z/) {
                return r("405 Method Not Allowed");
        }
-       my $path_info = decode_utf8($cgi->path_info);
+       my $path_info = $enc_utf8->decode($cgi->path_info);
 
        # top-level indices and feeds
        if ($path_info eq "/") {
@@ -65,9 +66,7 @@ sub main {
                invalid_list(\%ctx, $1) || redirect_list_index(\%ctx, $cgi);
        } elsif ($path_info =~ m!$LISTNAME_RE(?:/|/index\.html)?\z!o) {
                invalid_list(\%ctx, $1) || get_index(\%ctx, $cgi, 0);
-       } elsif ($path_info =~ m!$LISTNAME_RE/index\.atom\.xml\z!o) {
-               invalid_list(\%ctx, $1) || get_atom(\%ctx, $cgi, 1);
-       } elsif ($path_info =~ m!$LISTNAME_RE/all\.atom\.xml\z!o) {
+       } elsif ($path_info =~ m!$LISTNAME_RE/atom\.xml\z!o) {
                invalid_list(\%ctx, $1) || get_atom(\%ctx, $cgi, 0);
 
        # single-message pages
@@ -80,7 +79,7 @@ sub main {
        } elsif ($path_info =~ m!$LISTNAME_RE/f/(\S+)\.html\z!o) {
                invalid_list_mid(\%ctx, $1, $2) || get_full_html(\%ctx, $cgi);
 
-       # convenience redirect
+       # convenience redirects, order matters
        } elsif ($path_info =~ m!$LISTNAME_RE/(?:m|f)/(\S+)\z!o) {
                invalid_list_mid(\%ctx, $1, $2) || redirect_mid(\%ctx, $cgi);
 
@@ -114,8 +113,7 @@ sub invalid_list_mid {
        undef;
 }
 
-# /$LISTNAME/index.atom.xml                     -> Atom feed
-# /$LISTNAME/all.atom.xml                       -> Atom feed, includes replies
+# /$LISTNAME/atom.xml                       -> Atom feed, includes replies
 sub get_atom {
        my ($ctx, $cgi, $top) = @_;
        require PublicInbox::Feed;
@@ -170,8 +168,9 @@ sub get_mid_html {
        my $x = mid2blob($ctx);
        return r404() unless $x;
 
-       my $pfx = "../f/" . uri_escape($ctx->{mid}) . ".html";
        require PublicInbox::View;
+       my $mid_href = PublicInbox::Hval::ascii_html(uri_escape($ctx->{mid}));
+       my $pfx = "../f/$mid_href.html";
        require Email::MIME;
        [ "200 OK", {'Content-Type' => 'text/html'},
                PublicInbox::View->as_html(Email::MIME->new($$x), $pfx)];
@@ -215,6 +214,6 @@ sub set_binmode {
                # no way to validate raw messages, mixed encoding is possible.
                binmode STDOUT;
        } else { # strict encoding for HTML and XML
-               binmode STDOUT, ':encoding(UTF-8)';
+               binmode STDOUT, ':encoding(us-ascii)';
        }
 }