ApplyCommand: use byte arrays for text patches, not strings

Instead of converting the patch bytes to strings apply the patch on
byte level, like C git does. Converting the input lines and the hunk
lines from bytes to strings and then applying the patch based on
strings may give surprising results if a patch converts a text file
from one encoding to another. Moreover, in the end we don't know which
encoding to use to write the result.

Previous code just wrote the result as UTF-8, which forcibly changed
the encoding if the original input had some other encoding (even if the
patch had the same non-UTF-8 encoding). It was also wrong if the input
was UTF-8, and the patch should have changed the encoding to something
else.

So use ByteBuffers instead of Strings. This has the additional advantage
that all these ByteBuffers can share the underlying byte arrays of the
input and of the patch, so it also reduces memory consumption.

Change-Id: I450975f2ba0e7d0bec8973e3113cc2e7aea187ee
Signed-off-by: Thomas Wolf <thomas.wolf@paranor.ch>

org.eclipse.jgit.test/.settings/org.eclipse.core.resources.prefs

@@ -1,5 +1,6 @@
-#Sat Dec 20 21:21:24 CET 2008
 eclipse.preferences.version=1
+encoding//tst-rsrc/org/eclipse/jgit/diff/umlaut.patch=ISO-8859-1
+encoding//tst-rsrc/org/eclipse/jgit/diff/umlaut_PostImage=ISO-8859-1
 encoding//tst-rsrc/org/eclipse/jgit/patch/testGetText_BothISO88591.patch=ISO-8859-1
 encoding//tst-rsrc/org/eclipse/jgit/patch/testGetText_Convert.patch=ISO-8859-1
 encoding//tst-rsrc/org/eclipse/jgit/patch/testGetText_DiffCc.patch=ISO-8859-1

org.eclipse.jgit.test/tst-rsrc/org/eclipse/jgit/diff/umlaut.patch

@@ -0,0 +1,7 @@
+diff --git a/umlaut b/umlaut
+index 003a054..557f72f 100644
+--- a/umlaut
++++ b/umlaut
+@@ -1 +1 @@
+-ÄÖÜ
++ÄÖÜ

org.eclipse.jgit.test/tst-rsrc/org/eclipse/jgit/diff/umlaut_PostImage

@@ -0,0 +1 @@
+ﾄﾖﾜ

org.eclipse.jgit.test/tst-rsrc/org/eclipse/jgit/diff/umlaut_PreImage

@@ -0,0 +1 @@
+ÄÖÜ

org.eclipse.jgit.test/tst/org/eclipse/jgit/api/ApplyCommandTest.java

@@ -287,6 +287,14 @@ public class ApplyCommandTest extends RepositoryTestCase {
 		checkBinary("literal_add", false);
 	}
 
+	@Test
+	public void testEncodingChange() throws Exception {
+		// This is a text patch that changes a file containing ÄÖÜ in UTF-8 to
+		// the same characters in ISO-8859-1. The patch file itself uses mixed
+		// encoding. Since checkFile() works with strings use the binary check.
+		checkBinary("umlaut", true);
+	}
+
 	@Test
 	public void testAddA1() throws Exception {
 		ApplyResult result = init("A1", false, true);

org.eclipse.jgit/src/org/eclipse/jgit/api/ApplyCommand.java

@@ -10,7 +10,6 @@
 package org.eclipse.jgit.api;
 
 import java.io.BufferedInputStream;
-import java.io.BufferedWriter;
 import java.io.ByteArrayInputStream;
 import java.io.File;
 import java.io.FileInputStream;
@@ -18,9 +17,7 @@ import java.io.FileOutputStream;
 import java.io.IOException;
 import java.io.InputStream;
 import java.io.OutputStream;
-import java.io.OutputStreamWriter;
-import java.io.Writer;
-import java.nio.charset.StandardCharsets;
+import java.nio.ByteBuffer;
 import java.nio.file.Files;
 import java.nio.file.StandardCopyOption;
 import java.text.MessageFormat;
@@ -549,11 +546,11 @@ public class ApplyCommand extends GitCommand<ApplyResult> {
 	private void applyText(Repository repository, String path, RawText rt,
 			File f, FileHeader fh, CheckoutMetadata checkOut)
 			throws IOException, PatchApplyException {
-		List<String> oldLines = new ArrayList<>(rt.size());
+		List<ByteBuffer> oldLines = new ArrayList<>(rt.size());
 		for (int i = 0; i < rt.size(); i++) {
-			oldLines.add(rt.getString(i));
+			oldLines.add(rt.getRawString(i));
 		}
-		List<String> newLines = new ArrayList<>(oldLines);
+		List<ByteBuffer> newLines = new ArrayList<>(oldLines);
 		int afterLastHunk = 0;
 		int lineNumberShift = 0;
 		int lastHunkNewLine = -1;
@@ -571,9 +568,9 @@ public class ApplyCommand extends GitCommand<ApplyResult> {
 					b.length);
 			RawText hrt = new RawText(b);
 
-			List<String> hunkLines = new ArrayList<>(hrt.size());
+			List<ByteBuffer> hunkLines = new ArrayList<>(hrt.size());
 			for (int i = 0; i < hrt.size(); i++) {
-				hunkLines.add(hrt.getString(i));
+				hunkLines.add(hrt.getRawString(i));
 			}
 
 			if (hh.getNewStartLine() == 0) {
@@ -642,8 +639,8 @@ public class ApplyCommand extends GitCommand<ApplyResult> {
 			lineNumberShift = applyAt - hh.getNewStartLine() + 1;
 			int sz = hunkLines.size();
 			for (int j = 1; j < sz; j++) {
-				String hunkLine = hunkLines.get(j);
-				switch (hunkLine.charAt(0)) {
+				ByteBuffer hunkLine = hunkLines.get(j);
+				switch (hunkLine.array()[hunkLine.position()]) {
 				case ' ':
 					applyAt++;
 					break;
@@ -651,7 +648,7 @@ public class ApplyCommand extends GitCommand<ApplyResult> {
 					newLines.remove(applyAt);
 					break;
 				case '+':
-					newLines.add(applyAt++, hunkLine.substring(1));
+					newLines.add(applyAt++, slice(hunkLine, 1));
 					break;
 				default:
 					break;
@@ -660,28 +657,29 @@ public class ApplyCommand extends GitCommand<ApplyResult> {
 			afterLastHunk = applyAt;
 		}
 		if (!isNoNewlineAtEndOfFile(fh)) {
-			newLines.add(""); //$NON-NLS-1$
+			newLines.add(null);
 		}
 		if (!rt.isMissingNewlineAtEnd()) {
-			oldLines.add(""); //$NON-NLS-1$
+			oldLines.add(null);
 		}
-		if (!isChanged(oldLines, newLines)) {
-			return; // Don't touch the file
+		if (oldLines.equals(newLines)) {
+			return; // Unchanged; don't touch the file
 		}
 
-		// TODO: forcing UTF-8 is a bit strange and may lead to re-coding if the
-		// input was some other encoding, but it's what previous versions of
-		// this code used. (Even earlier the code used the default encoding,
-		// which has the same problem.) Perhaps using bytes instead of Strings
-		// for the lines would be better.
 		TemporaryBuffer buffer = new TemporaryBuffer.LocalFile(null);
 		try {
-			try (Writer w = new BufferedWriter(
-					new OutputStreamWriter(buffer, StandardCharsets.UTF_8))) {
-				for (Iterator<String> l = newLines.iterator(); l.hasNext();) {
-					w.write(l.next());
+			try (OutputStream out = buffer) {
+				for (Iterator<ByteBuffer> l = newLines.iterator(); l
+						.hasNext();) {
+					ByteBuffer line = l.next();
+					if (line == null) {
+						// Must be the marker for the final newline
+						break;
+					}
+					out.write(line.array(), line.position(),
+							line.limit() - line.position());
 					if (l.hasNext()) {
-						w.write('\n');
+						out.write('\n');
 					}
 				}
 			}
@@ -698,18 +696,18 @@ public class ApplyCommand extends GitCommand<ApplyResult> {
 				fh.getNewMode() == FileMode.EXECUTABLE_FILE);
 	}
 
-	private boolean canApplyAt(List<String> hunkLines, List<String> newLines,
-			int line) {
+	private boolean canApplyAt(List<ByteBuffer> hunkLines,
+			List<ByteBuffer> newLines, int line) {
 		int sz = hunkLines.size();
 		int limit = newLines.size();
 		int pos = line;
 		for (int j = 1; j < sz; j++) {
-			String hunkLine = hunkLines.get(j);
-			switch (hunkLine.charAt(0)) {
+			ByteBuffer hunkLine = hunkLines.get(j);
+			switch (hunkLine.array()[hunkLine.position()]) {
 			case ' ':
 			case '-':
 				if (pos >= limit
-						|| !newLines.get(pos).equals(hunkLine.substring(1))) {
+						|| !newLines.get(pos).equals(slice(hunkLine, 1))) {
 					return false;
 				}
 				pos++;
@@ -721,13 +719,9 @@ public class ApplyCommand extends GitCommand<ApplyResult> {
 		return true;
 	}
 
-	private static boolean isChanged(List<String> ol, List<String> nl) {
-		if (ol.size() != nl.size())
-			return true;
-		for (int i = 0; i < ol.size(); i++)
-			if (!ol.get(i).equals(nl.get(i)))
-				return true;
-		return false;
+	private ByteBuffer slice(ByteBuffer b, int off) {
+		int newOffset = b.position() + off;
+		return ByteBuffer.wrap(b.array(), newOffset, b.limit() - newOffset);
 	}
 
 	private boolean isNoNewlineAtEndOfFile(FileHeader fh) {

org.eclipse.jgit/src/org/eclipse/jgit/diff/RawText.java

@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009, Google Inc.
- * Copyright (C) 2008-2009, Johannes E. Schindelin <johannes.schindelin@gmx.de> and others
+ * Copyright (C) 2008-2021, Johannes E. Schindelin <johannes.schindelin@gmx.de> and others
  *
  * This program and the accompanying materials are made available under the
  * terms of the Eclipse Distribution License v. 1.0 which is available at
@@ -16,6 +16,7 @@ import java.io.File;
 import java.io.IOException;
 import java.io.InputStream;
 import java.io.OutputStream;
+import java.nio.ByteBuffer;
 
 import org.eclipse.jgit.errors.BinaryBlobException;
 import org.eclipse.jgit.errors.LargeObjectException;
@@ -164,6 +165,27 @@ public class RawText extends Sequence {
 		return getString(i, i + 1, true);
 	}
 
+	/**
+	 * Get the raw text for a single line.
+	 *
+	 * @param i
+	 *            index of the line to extract. Note this is 0-based, so line
+	 *            number 1 is actually index 0.
+	 * @return the text for the line, without a trailing LF, as a
+	 *         {@link ByteBuffer} that is backed by a slice of the
+	 *         {@link #getRawContent() raw content}, with the buffer's position
+	 *         on the start of the line and the limit at the end.
+	 * @since 5.12
+	 */
+	public ByteBuffer getRawString(int i) {
+		int s = getStart(i);
+		int e = getEnd(i);
+		if (e > 0 && content[e - 1] == '\n') {
+			e--;
+		}
+		return ByteBuffer.wrap(content, s, e - s);
+	}
+
 	/**
 	 * Get the text for a region of lines.
 	 *