From d5998cce7dee2c7e4200c3e8e7285095e5ba1a85 Mon Sep 17 00:00:00 2001
From: Nan Zhang <nanzhang@google.com>
Date: Wed, 13 Sep 2017 13:17:43 -0700
Subject: [PATCH] Don't add data_descripters when merging uncompress zip
 entries for merge_zips.

Also filter out META-INF/TRANSITIVE dir, and report warnings when
merge_zips see duplicates entries with different CRC hash.

Bug: b/65455145
Test: m clean && m -j java (locally)

Change-Id: I47172ffa27df71f3280f35f6b540a7b5a0c14550
---
 cmd/merge_zips/merge_zips.go | 35 +++++++++++++++++++++++++++++++---
 third_party/zip/android.go   | 37 ++++++++++++++++++------------------
 2 files changed, 51 insertions(+), 21 deletions(-)

diff --git a/cmd/merge_zips/merge_zips.go b/cmd/merge_zips/merge_zips.go
index 7874a41b7..3c1cb9a18 100644
--- a/cmd/merge_zips/merge_zips.go
+++ b/cmd/merge_zips/merge_zips.go
@@ -26,11 +26,28 @@ import (
 	"android/soong/third_party/zip"
 )
 
+type strip struct{}
+
+func (s *strip) String() string {
+	return `""`
+}
+
+func (s *strip) Set(path_prefix string) error {
+	strippings = append(strippings, path_prefix)
+
+	return nil
+}
+
 var (
 	sortEntries = flag.Bool("s", false, "sort entries (defaults to the order from the input zip files)")
 	emulateJar  = flag.Bool("j", false, "sort zip entries using jar ordering (META-INF first)")
+	strippings  []string
 )
 
+func init() {
+	flag.Var(&strip{}, "strip", "the prefix of file path to be excluded from the output zip")
+}
+
 func main() {
 	flag.Usage = func() {
 		fmt.Fprintln(os.Stderr, "usage: merge_zips [-j] output [inputs...]")
@@ -115,7 +132,13 @@ func mergeZips(readers []namedZipReader, writer *zip.Writer, sortEntries bool, e
 	orderedMappings := []fileMapping{}
 
 	for _, namedReader := range readers {
+	FileLoop:
 		for _, file := range namedReader.reader.File {
+			for _, path_prefix := range strippings {
+				if strings.HasPrefix(file.Name, path_prefix) {
+					continue FileLoop
+				}
+			}
 			// check for other files or directories destined for the same path
 			dest := file.Name
 			mapKey := dest
@@ -142,8 +165,15 @@ func mergeZips(readers []namedZipReader, writer *zip.Writer, sortEntries bool, e
 					continue
 				}
 				if !isDir {
-					return fmt.Errorf("Duplicate path %v found in %v and %v\n",
-						dest, existingMapping.source.path, newMapping.source.path)
+					if emulateJar {
+						if existingMapping.source.content.CRC32 != newMapping.source.content.CRC32 {
+							fmt.Fprintf(os.Stdout, "WARNING: Duplicate path %v found in %v and %v\n",
+								dest, existingMapping.source.path, newMapping.source.path)
+						}
+					} else {
+						return fmt.Errorf("Duplicate path %v found in %v and %v\n",
+							dest, existingMapping.source.path, newMapping.source.path)
+					}
 				}
 			} else {
 				// save entry
@@ -151,7 +181,6 @@ func mergeZips(readers []namedZipReader, writer *zip.Writer, sortEntries bool, e
 				orderedMappings = append(orderedMappings, newMapping)
 			}
 		}
-
 	}
 
 	if emulateJar {
diff --git a/third_party/zip/android.go b/third_party/zip/android.go
index bde3afa4f..d35513fd7 100644
--- a/third_party/zip/android.go
+++ b/third_party/zip/android.go
@@ -32,7 +32,6 @@ func (w *Writer) CopyFrom(orig *File, newName string) error {
 	fileHeader := orig.FileHeader
 	fileHeader.Name = newName
 	fh := &fileHeader
-	fh.Flags |= DataDescriptorFlag
 
 	// The zip64 extras change between the Central Directory and Local File Header, while we use
 	// the same structure for both. The Local File Haeder is taken care of by us writing a data
@@ -57,24 +56,26 @@ func (w *Writer) CopyFrom(orig *File, newName string) error {
 	}
 	io.Copy(w.cw, io.NewSectionReader(orig.zipr, dataOffset, int64(orig.CompressedSize64)))
 
-	// Write data descriptor.
-	var buf []byte
-	if fh.isZip64() {
-		buf = make([]byte, dataDescriptor64Len)
-	} else {
-		buf = make([]byte, dataDescriptorLen)
+	if orig.hasDataDescriptor() {
+		// Write data descriptor.
+		var buf []byte
+		if fh.isZip64() {
+			buf = make([]byte, dataDescriptor64Len)
+		} else {
+			buf = make([]byte, dataDescriptorLen)
+		}
+		b := writeBuf(buf)
+		b.uint32(dataDescriptorSignature)
+		b.uint32(fh.CRC32)
+		if fh.isZip64() {
+			b.uint64(fh.CompressedSize64)
+			b.uint64(fh.UncompressedSize64)
+		} else {
+			b.uint32(fh.CompressedSize)
+			b.uint32(fh.UncompressedSize)
+		}
+		_, err = w.cw.Write(buf)
 	}
-	b := writeBuf(buf)
-	b.uint32(dataDescriptorSignature)
-	b.uint32(fh.CRC32)
-	if fh.isZip64() {
-		b.uint64(fh.CompressedSize64)
-		b.uint64(fh.UncompressedSize64)
-	} else {
-		b.uint32(fh.CompressedSize)
-		b.uint32(fh.UncompressedSize)
-	}
-	_, err = w.cw.Write(buf)
 	return err
 }