Update ZipString to deal with reads that do not return all data

Refine the logic in `ZipString.hash` and `ZipString.compare` to deal with the fact a read operation may not return all available bytes. Fixes gh-38751
spring-projects · Dec 13, 2023 · b4a4e91 · b4a4e91
1 parent afad358
commit b4a4e91
Show file tree

Hide file tree

Showing 3 changed files with 41 additions and 22 deletions.
diff --git a/...ing-boot-loader/src/main/java/org/springframework/boot/loader/zip/ByteArrayDataBlock.java b/...ing-boot-loader/src/main/java/org/springframework/boot/loader/zip/ByteArrayDataBlock.java
@@ -28,12 +28,19 @@ class ByteArrayDataBlock implements CloseableDataBlock {
 
 	private final byte[] bytes;
 
+	private final int maxReadSize;
+
 	/**
 	 * Create a new {@link ByteArrayDataBlock} backed by the given bytes.
 	 * @param bytes the bytes to use
 	 */
 	ByteArrayDataBlock(byte... bytes) {
+		this(bytes, -1);
+	}
+
+	ByteArrayDataBlock(byte[] bytes, int maxReadSize) {
 		this.bytes = bytes;
+		this.maxReadSize = maxReadSize;
 	}
 
 	@Override
@@ -49,6 +56,9 @@ public int read(ByteBuffer dst, long pos) throws IOException {
 	private int read(ByteBuffer dst, int pos) {
 		int remaining = dst.remaining();
 		int length = Math.min(this.bytes.length - pos, remaining);
+		if (this.maxReadSize > 0 && length > this.maxReadSize) {
+			length = this.maxReadSize;
+		}
 		dst.put(this.bytes, pos, length);
 		return length;
 	}

diff --git a/...tools/spring-boot-loader/src/main/java/org/springframework/boot/loader/zip/ZipString.java b/...tools/spring-boot-loader/src/main/java/org/springframework/boot/loader/zip/ZipString.java
@@ -108,19 +108,15 @@ static int hash(ByteBuffer buffer, DataBlock dataBlock, long pos, int len, boole
 		byte[] bytes = buffer.array();
 		int hash = 0;
 		char lastChar = 0;
+		int codePointSize = 1;
 		while (len > 0) {
-			int count = readInBuffer(dataBlock, pos, buffer, len);
-			len -= count;
-			pos += count;
+			int count = readInBuffer(dataBlock, pos, buffer, len, codePointSize);
 			for (int byteIndex = 0; byteIndex < count;) {
-				int codePointSize = getCodePointSize(bytes, byteIndex);
+				codePointSize = getCodePointSize(bytes, byteIndex);
 				if (!hasEnoughBytes(byteIndex, codePointSize, count)) {
-					pos--;
-					len++;
 					break;
 				}
 				int codePoint = getCodePoint(bytes, byteIndex, codePointSize);
-				byteIndex += codePointSize;
 				if (codePoint <= 0xFFFF) {
 					lastChar = (char) (codePoint & 0xFFFF);
 					hash = 31 * hash + lastChar;
@@ -130,6 +126,10 @@ static int hash(ByteBuffer buffer, DataBlock dataBlock, long pos, int len, boole
 					hash = 31 * hash + Character.highSurrogate(codePoint);
 					hash = 31 * hash + Character.lowSurrogate(codePoint);
 				}
+				byteIndex += codePointSize;
+				pos += codePointSize;
+				len -= codePointSize;
+				codePointSize = 1;
 			}
 		}
 		hash = (addEndSlash && lastChar != '/') ? 31 * hash + '/' : hash;
@@ -198,19 +198,15 @@ private static int compare(ByteBuffer buffer, DataBlock dataBlock, long pos, int
 		int maxCharSequenceLength = (!addSlash) ? charSequence.length() : charSequence.length() + 1;
 		int result = 0;
 		byte[] bytes = buffer.array();
+		int codePointSize = 1;
 		while (len > 0) {
-			int count = readInBuffer(dataBlock, pos, buffer, len);
-			len -= count;
-			pos += count;
+			int count = readInBuffer(dataBlock, pos, buffer, len, codePointSize);
 			for (int byteIndex = 0; byteIndex < count;) {
-				int codePointSize = getCodePointSize(bytes, byteIndex);
+				codePointSize = getCodePointSize(bytes, byteIndex);
 				if (!hasEnoughBytes(byteIndex, codePointSize, count)) {
-					pos--;
-					len++;
 					break;
 				}
 				int codePoint = getCodePoint(bytes, byteIndex, codePointSize);
-				result += codePointSize;
 				if (codePoint <= 0xFFFF) {
 					char ch = (char) (codePoint & 0xFFFF);
 					if (charSequenceIndex >= maxCharSequenceLength
@@ -230,10 +226,14 @@ private static int compare(ByteBuffer buffer, DataBlock dataBlock, long pos, int
 						return -1;
 					}
 				}
+				byteIndex += codePointSize;
+				pos += codePointSize;
+				len -= codePointSize;
+				result += codePointSize;
+				codePointSize = 1;
 				if (compareType == CompareType.STARTS_WITH && charSequenceIndex >= charSequence.length()) {
 					return result;
 				}
-				byteIndex += codePointSize;
 			}
 		}
 		return (charSequenceIndex >= charSequence.length()) ? result : -1;
@@ -273,16 +273,22 @@ static String readString(DataBlock data, long pos, long len) {
 		}
 	}
 
-	private static int readInBuffer(DataBlock dataBlock, long pos, ByteBuffer buffer, int maxLen) throws IOException {
+	private static int readInBuffer(DataBlock dataBlock, long pos, ByteBuffer buffer, int maxLen, int minLen)
+			throws IOException {
 		buffer.clear();
 		if (buffer.remaining() > maxLen) {
 			buffer.limit(maxLen);
 		}
-		int count = dataBlock.read(buffer, pos);
-		if (count <= 0) {
-			throw new EOFException();
+		int result = 0;
+		while (result < minLen) {
+			int count = dataBlock.read(buffer, pos);
+			if (count <= 0) {
+				throw new EOFException();
+			}
+			result += count;
+			pos += count;
 		}
-		return count;
+		return result;
 	}
 
 	private static int getCodePointSize(byte[] bytes, int i) {

diff --git a/.../spring-boot-loader/src/test/java/org/springframework/boot/loader/zip/ZipStringTests.java b/.../spring-boot-loader/src/test/java/org/springframework/boot/loader/zip/ZipStringTests.java
@@ -86,7 +86,10 @@ void testHash(HashSourceType sourceType, String source, boolean addEndSlash, int
 			case DATA_BLOCK -> {
 				ByteArrayDataBlock dataBlock = new ByteArrayDataBlock(source.getBytes(StandardCharsets.UTF_8));
 				assertThat(ZipString.hash(null, dataBlock, 0, (int) dataBlock.size(), addEndSlash)).isEqualTo(expected);
-
+			}
+			case SINGLE_BYTE_READ_DATA_BLOCK -> {
+				ByteArrayDataBlock dataBlock = new ByteArrayDataBlock(source.getBytes(StandardCharsets.UTF_8), 1);
+				assertThat(ZipString.hash(null, dataBlock, 0, (int) dataBlock.size(), addEndSlash)).isEqualTo(expected);
 			}
 		}
 	}
@@ -187,7 +190,7 @@ private AbstractIntegerAssert<?> assertStartsWith(String source, CharSequence ch
 
 	enum HashSourceType {
 
-		STRING, CHAR_SEQUENCE, DATA_BLOCK
+		STRING, CHAR_SEQUENCE, DATA_BLOCK, SINGLE_BYTE_READ_DATA_BLOCK
 
 	}