[Git][reproducible-builds/strip-nondeterminism][master] Add support for custom .zip filename filtering and exclude two patterns of...

[Chris Lamb]

Chris Lamb pushed to branch master at Reproducible Builds / strip-nondeterminism


Commits:
2578d3dc by Chris Lamb at 2020-04-17T11:28:38+01:00
Add support for custom .zip filename filtering and exclude two patterns of files generated by Maven projects in "fork" mode. (Closes: reproducible-builds/strip-nondeterminism#13)

- - - - -


2 changed files:

- lib/File/StripNondeterminism/handlers/jar.pm
- lib/File/StripNondeterminism/handlers/zip.pm


Changes:

=====================================
lib/File/StripNondeterminism/handlers/jar.pm
=====================================
@@ -1,6 +1,6 @@
 #
 # Copyright 2014 Andrew Ayer
-# Copyright 2016-2018 Chris Lamb <lamby at debian.org>
+# Copyright 2016-2018, 2020 Chris Lamb <lamby at debian.org>
 #
 # This file is part of strip-nondeterminism.
 #
@@ -42,6 +42,28 @@ sub _jar_filename_cmp($$) {
 	return $a cmp $b;
 }
 
+sub _jar_filename_filter($) {
+	my @filenames = @_;
+
+	return grep {
+		# When a Maven project is built with the fork mode enabled for
+		# the maven-compiler.plugin, two extra files containing non
+		# reproducible elements are generated in the output directory
+		# and subsequently packaged in the jar.
+		#
+		# Remove executable script that could be run to recompile the
+		# project. It captures the build path.
+		$_ = undef if $_ eq "javac.sh";
+
+		# Contains the compiler arguments. The file name uses a random
+		# value.
+		$_ = undef if $_ =~ m/^org\.codehaus\.plexus\.compiler\.javac\.JavacCompiler\d+arguments$/;
+
+		# Allow all other files
+		$_;
+	} @filenames;
+}
+
 sub _jar_normalize_manifest($) {
 	my ($filename) = @_;
 
@@ -127,6 +149,7 @@ sub normalize {
 		$jar_filename,
 		archive_filter => \&_jar_archive_filter,
 		filename_cmp => \&_jar_filename_cmp,
+		filename_filter => \&_jar_filename_filter,
 		member_normalizer => \&_jar_normalize_member
 	);
 }


=====================================
lib/File/StripNondeterminism/handlers/zip.pm
=====================================
@@ -1,6 +1,6 @@
 #
 # Copyright 2014 Andrew Ayer
-# Copyright 2016-2019 Chris Lamb <lamby at debian.org>
+# Copyright 2016-2020 Chris Lamb <lamby at debian.org>
 #
 # This file is part of strip-nondeterminism.
 #
@@ -201,6 +201,16 @@ sub normalize {
 	$canonical_time = SAFE_EPOCH
 	  if not defined $canonical_time or $canonical_time < SAFE_EPOCH;
 	my @filenames = sort $filename_cmp $zip->memberNames();
+	if (exists($options{filename_filter})) {
+		@filenames = $options{filename_filter}->(@filenames);
+		# Remove any members deleted by the custom filter
+		my %seen = map { $_ => 1 } @filenames;
+		for my $filename ($zip->memberNames()) {
+			next if $seen{$filename};
+			print STDERR "Removing $filename from $zip_filename\n";
+			$zip->removeMember($filename);
+		}
+	}
 	for my $filename (@filenames) {
 		my $member = $zip->removeMember($filename);
 		if ($member->isEncrypted()) {



View it on GitLab: https://salsa.debian.org/reproducible-builds/strip-nondeterminism/-/commit/2578d3dccaf562f1ecb61ea4c68a1cef1ce11259

-- 
View it on GitLab: https://salsa.debian.org/reproducible-builds/strip-nondeterminism/-/commit/2578d3dccaf562f1ecb61ea4c68a1cef1ce11259
You're receiving this email because of your account on salsa.debian.org.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.reproducible-builds.org/pipermail/rb-commits/attachments/20200417/e0b36b72/attachment.htm>