view: support SHA-1 of Message-IDs for message links

Some Message-IDs are crazy long, so support SHA-1s for them
instead.  This allows shorter URLs to be generated and are
less likely

However, we'll still favor short Message-IDs whenever possible.

diff --git a/lib/PublicInbox/Hval.pm b/lib/PublicInbox/Hval.pm
index a93e8c247ea6cce2fd7b85398cb2405ef3dd06fa..9be163aa70c13b39eb19c326458886ef1a9f75c8 100644 (file)
--- a/lib/PublicInbox/Hval.pm
+++ b/lib/PublicInbox/Hval.pm
@@ -5,28 +5,36 @@
 package PublicInbox::Hval;
 use strict;
 use warnings;
-use fields qw(raw);
+use fields qw(raw href);
 use Encode qw(find_encoding);
 use URI::Escape qw(uri_escape_utf8);
 
 my $enc_ascii = find_encoding('us-ascii');
 
 sub new {
-       my ($class, $raw) = @_;
+       my ($class, $raw, $href) = @_;
        my $self = fields::new($class);
 
        # we never care about leading/trailing whitespace
        $raw =~ s/\A\s*//;
        $raw =~ s/\s*\z//;
        $self->{raw} = $raw;
+       $self->{href} = defined $href ? $href : $raw;
        $self;
 }
 
 sub new_msgid {
-       my ($class, $raw) = @_;
-       $raw =~ s/\A<//;
-       $raw =~ s/>\z//;
-       $class->new($raw);
+       my ($class, $msgid) = @_;
+       $msgid =~ s/\A\s*<?//;
+       $msgid =~ s/>?\s*\z//;
+
+       if (length($msgid) <= 40) {
+               $class->new($msgid);
+       } else {
+               require Digest::SHA;
+               my $hex = Digest::SHA::sha1_hex($msgid);
+               $class->new($msgid, $hex);
+       }
 }
 
 sub new_oneline {
@@ -52,7 +60,7 @@ sub ascii_html {
 }
 
 sub as_html { ascii_html($_[0]->{raw}) }
-sub as_href { ascii_html(uri_escape_utf8($_[0]->{raw})) }
+sub as_href { ascii_html(uri_escape_utf8($_[0]->{href})) }
 
 sub raw {
        if (defined $_[1]) {

diff --git a/lib/PublicInbox/WWW.pm b/lib/PublicInbox/WWW.pm
index f67e72cefbfb9629189fa50ccb572e8fc690179a..1814286b5e7647e923e8d1d3376f94d062829510 100644 (file)
--- a/lib/PublicInbox/WWW.pm
+++ b/lib/PublicInbox/WWW.pm
@@ -130,13 +130,19 @@ sub get_index {
 # just returns a string ref for the blob in the current ctx
 sub mid2blob {
        my ($ctx) = @_;
-       require Digest::SHA;
-       my $hex = Digest::SHA::sha1_hex($ctx->{mid});
-       $hex =~ /\A([a-f0-9]{2})([a-f0-9]{38})\z/i or
-                       die "BUG: not a SHA-1 hex: $hex";
+       my $hex = $ctx->{mid};
+       my ($x2, $x38) = ($hex =~ /\A([a-f0-9]{2})([a-f0-9]{38})\z/);
+
+       unless (defined $x38) {
+               # compatibility with old links
+               require Digest::SHA;
+               $hex = Digest::SHA::sha1_hex($hex);
+               ($x2, $x38) = ($hex =~ /\A([a-f0-9]{2})([a-f0-9]{38})\z/);
+               defined $x38 or die "BUG: not a SHA-1 hex: $hex";
+       }
 
        my @cmd = ('git', "--git-dir=$ctx->{git_dir}",
-                       qw(cat-file blob), "HEAD:$1/$2");
+                       qw(cat-file blob), "HEAD:$x2/$x38");
        my $cmd = join(' ', @cmd);
        my $pid = open my $fh, '-|';
        defined $pid or die "fork failed: $!\n";