X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FImport.pm;h=250a2db31e979054c94df5a0e1dd6ca2d8a5f883;hb=e350de7fdbe12dafc36e893e66ce8c93ec6dc3f2;hp=9e8900f3e6f86d2ce50576039654e75cad104120;hpb=dd80811bdaddc40520cb37abda23624674022e37;p=public-inbox.git

diff --git a/lib/PublicInbox/Import.pm b/lib/PublicInbox/Import.pm
index 9e8900f3..250a2db3 100644
--- a/lib/PublicInbox/Import.pm
+++ b/lib/PublicInbox/Import.pm
@@ -14,6 +14,7 @@ use PublicInbox::Address;
 use PublicInbox::MsgTime qw(msg_timestamp msg_datestamp);
 use PublicInbox::ContentId qw(content_digest);
 use PublicInbox::MDA;
+use POSIX qw(strftime);
 
 sub new {
 	my ($class, $git, $name, $email, $ibx) = @_;
@@ -116,6 +117,9 @@ sub _cat_blob ($$$) {
 	$n = read($r, my $lf, 1);
 	defined($n) or die "read final byte of cat-blob failed: $!";
 	die "bad read on final byte: <$lf>" if $lf ne "\n";
+
+	# fixup some bugginess in old versions:
+	$buf =~ s/\A[\r\n]*From [^\r\n]*\r?\n//s;
 	\$buf;
 }
 
@@ -330,7 +334,7 @@ sub v1_mid0 ($) {
 	my $mids = mids($hdr);
 
 	if (!scalar(@$mids)) { # spam often has no Message-Id
-		my $mid0 = digest2mid(content_digest($mime));
+		my $mid0 = digest2mid(content_digest($mime), $hdr);
 		append_mid($hdr, $mid0);
 		return $mid0;
 	}
@@ -436,6 +440,8 @@ sub done {
 	_update_git_info($self, 1) if delete $self->{nchg};
 
 	$self->lock_release;
+
+	$self->{git}->cleanup;
 }
 
 sub atfork_child {
@@ -445,18 +451,19 @@ sub atfork_child {
 	}
 }
 
-sub digest2mid ($) {
-	my ($dig) = @_;
+sub digest2mid ($$) {
+	my ($dig, $hdr) = @_;
 	my $b64 = $dig->clone->b64digest;
 	# Make our own URLs nicer:
 	# See "Base 64 Encoding with URL and Filename Safe Alphabet" in RFC4648
 	$b64 =~ tr!+/=!-_!d;
 
-	# We can make this more meaningful with a date prefix or other things,
-	# but this is only needed for crap that fails to generate a Message-ID
-	# or reuses one.  In other words, it's usually spammers who hit this
-	# so they don't deserve nice Message-IDs :P
-	$b64 . '@localhost';
+	# Add a date prefix to prevent a leading '-' in case that trips
+	# up some tools (e.g. if a Message-ID were a expected as a
+	# command-line arg)
+	my $dt = msg_datestamp($hdr);
+	$dt = POSIX::strftime('%Y%m%d%H%M%S', gmtime($dt));
+	"$dt.$b64" . '@z';
 }
 
 sub clean_purge_buffer {