X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FSolverGit.pm;h=8629f0dafeb4216b41ac2c0951c866c8e76c2dcf;hb=55b707d788ce13696e4411389583e720ea6dab01;hp=a78360fd6bac59a956b7a6c31305fd5fa1b19fc8;hpb=3ac73bd16d9fa7beba6a1735dd5c505d021e227a;p=public-inbox.git

diff --git a/lib/PublicInbox/SolverGit.pm b/lib/PublicInbox/SolverGit.pm
index a78360fd..8629f0da 100644
--- a/lib/PublicInbox/SolverGit.pm
+++ b/lib/PublicInbox/SolverGit.pm
@@ -11,7 +11,7 @@ package PublicInbox::SolverGit;
 use strict;
 use warnings;
 use 5.010_001;
-use File::Temp 0.19 ();
+use File::Temp 0.19 (); # 0.19 for ->newdir
 use Fcntl qw(SEEK_SET);
 use PublicInbox::Git qw(git_unquote git_quote);
 use PublicInbox::MsgIter qw(msg_iter msg_part_text);
@@ -102,7 +102,6 @@ sub extract_diff ($$) {
 	my $ct = $part->content_type || 'text/plain';
 	my ($s, undef) = msg_part_text($part, $ct);
 	defined $s or return;
-	my $di = {};
 
 	# Email::MIME::Encodings forces QP to be CRLF upon decoding,
 	# change it back to LF:
@@ -112,6 +111,8 @@ sub extract_diff ($$) {
 	}
 
 	state $LF = qr!\r?\n!;
+	state $ANY = qr![^\r\n]+!;
+	state $MODE = '100644|120000|100755';
 	state $FN = qr!(?:("?[^/\n]+/[^\r\n]+)|/dev/null)!;
 
 	$s =~ m!( # $1 start header lines we save for debugging:
@@ -124,17 +125,16 @@ sub extract_diff ($$) {
 		# try to get the pre-and-post filenames as $2 and $3
 		(?:^diff\x20--git\x20$FN\x20$FN$LF)
 
-		# old mode $4
-		(?:^old mode\x20(100644|120000|100755)$LF)?
-
-		# ignore other info
-		(?:^(?:copy|rename|deleted|dissimilarity|similarity).*$LF)?
-
-		# new mode (possibly new file) ($5)
-		(?:^new\x20(?:file\x20)?mode\x20(100644|120000|100755)$LF)?
-
-		# ignore other info
-		(?:^(?:copy|rename|deleted|dissimilarity|similarity).*$LF)?
+		(?:^(?: # pass all this to git-apply:
+			# old mode $4
+			(?:old\x20mode\x20($MODE))
+			|
+			# new mode (possibly new file) ($5)
+			(?:new\x20(?:file\x20)?mode\x20($MODE))
+			|
+			(?:(?:copy|rename|deleted|
+				dissimilarity|similarity)$ANY)
+		)$LF)*
 
 		)? # end of optional stuff, everything below is required
 
@@ -155,15 +155,17 @@ sub extract_diff ($$) {
 		# the meat of the diff, including "^\\No newline ..."
 		# We also allow for totally blank lines w/o leading spaces,
 		# because git-apply(1) handles that case, too
-		(?:^(?:[\@\+\x20\-\\][^\r\n]*|)$LF)+
+		(?:^(?:[\@\+\x20\-\\][^\n]*|)$LF)+
 	)!smx or return;
 
-	my $hdr_lines = $1;
+	my $di = {
+		hdr_lines => $1,
+		oid_a => $6,
+		oid_b => $7,
+		mode_a => $5 // $8 // $4, # new (file) // unchanged // old
+	};
 	my $path_a = $2 // $10;
 	my $path_b = $3 // $11;
-	$di->{oid_a} = $6;
-	$di->{oid_b} = $7;
-	$di->{mode_a} = $5 // $8 // $4; # new (file) // unchanged // old
 	my $patch = $9;
 
 	# don't care for leading 'a/' and 'b/'
@@ -178,19 +180,16 @@ sub extract_diff ($$) {
 	$di->{path_a} = join('/', @a);
 	$di->{path_b} = join('/', @b);
 
-	utf8::encode($hdr_lines);
-	utf8::encode($patch);
 	my $path = ++$self->{tot};
 	$di->{n} = $path;
-	open(my $tmp, '>', $self->{tmp}->dirname . "/$path") or
+	open(my $tmp, '>:utf8', $self->{tmp}->dirname . "/$path") or
 		die "open(tmp): $!";
-	print $tmp $hdr_lines, $patch or die "print(tmp): $!";
+	print $tmp $di->{hdr_lines}, $patch or die "print(tmp): $!";
 	close $tmp or die "close(tmp): $!";
 
 	# for debugging/diagnostics:
 	$di->{ibx} = $ibx;
 	$di->{smsg} = $smsg;
-	$di->{hdr_lines} = $hdr_lines;
 
 	push @$diffs, $di;
 }