Xxhash integration

azure/pom.xml

@@ -6,12 +6,12 @@
     <parent>
         <groupId>com.instaclustr</groupId>
         <artifactId>esop-parent</artifactId>
-        <version>4.0.1</version>
+        <version>4.0.2-SNAPSHOT</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 
     <artifactId>esop-azure</artifactId>
-    <version>4.0.1</version>
+    <version>4.0.2-SNAPSHOT</version>
 
     <name>esop-azure</name>
     <description>Backup and restoration tooling for Cassandra for Azure</description>

core/pom.xml

@@ -6,12 +6,12 @@
     <parent>
         <groupId>com.instaclustr</groupId>
         <artifactId>esop-parent</artifactId>
-        <version>4.0.1</version>
+        <version>4.0.2-SNAPSHOT</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 
     <artifactId>esop-core</artifactId>
-    <version>4.0.1</version>
+    <version>4.0.2-SNAPSHOT</version>
 
     <name>esop-core</name>
     <description>Core of backup and restoration tooling for Cassandra</description>
@@ -25,6 +25,7 @@
         <commons-io.version>2.20.0</commons-io.version>
         <awaitility.version>3.1.6</awaitility.version>
         <jackson.bom.version>2.19.2</jackson.bom.version>
+        <lz4.version>1.8.0</lz4.version>
 
         <slf4j.version>2.0.17</slf4j.version>
         <logback.version>1.5.19</logback.version>
@@ -120,6 +121,12 @@
             <version>${awaitility.version}</version>
         </dependency>
 
+        <dependency>
+            <groupId>org.lz4</groupId>
+            <artifactId>lz4-java</artifactId>
+            <version>${lz4.version}</version>
+        </dependency>
+
         <!-- logging -->
 
         <dependency>

core/src/main/java/com/instaclustr/esop/impl/hash/HashSpec.java

@@ -7,6 +7,8 @@
 import java.util.zip.CRC32;
 import java.util.zip.Checksum;
 
+import net.jpountz.xxhash.StreamingXXHash64;
+import net.jpountz.xxhash.XXHashFactory;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 import picocli.CommandLine;
@@ -116,8 +118,35 @@ public String getHash(byte[] digest) throws Exception {
         }
     }
 
+    /**
+     * Wraps the xxHash64 algorithm. Used for fast hashing of large files as an alternative to SHA-256.
+     */
+    public static class XXHasher implements Hasher {
+
+        @Override
+        public String getHash(final InputStream is) throws Exception {
+            try (StreamingXXHash64 xxHash64 = XXHashFactory.fastestJavaInstance().newStreamingHash64(0)) {
+                byte[] byteArray = new byte[1024];
+                int bytesCount = 0;
+
+                while ((bytesCount = is.read(byteArray)) != -1) {
+                    xxHash64.update(byteArray, 0, bytesCount);
+                }
+
+                return Long.toString(xxHash64.getValue());
+            }
+        }
+
+        @Override
+        public String getHash(final byte[] digest) throws Exception {
+            // TODO do we actually need this?
+            throw new UnsupportedOperationException();
+        }
+    }
+
     public enum HashAlgorithm {
         SHA_256("SHA-256", () -> new SHAHasher("SHA-256")),
+        XXHASH64("xxHash64", () -> new XXHasher()),
         CRC("CRC", () -> new CRCHasher()),
         NONE("NONE", () -> new NoOp());
 
@@ -146,7 +175,7 @@ public static HashAlgorithm parse(final String value) {
             }
 
             for (final HashAlgorithm algorithm : HashAlgorithm.values()) {
-                if (algorithm.name.equals(value)) {
+                if (algorithm.name.equalsIgnoreCase(value)) {
                     return algorithm;
                 }
             }

core/src/test/java/com/instaclustr/esop/backup/HashServiceTest.java

@@ -6,16 +6,54 @@
 import com.instaclustr.esop.impl.hash.HashService;
 import com.instaclustr.esop.impl.hash.HashServiceImpl;
 import com.instaclustr.esop.impl.hash.HashSpec;
+import org.junit.jupiter.api.AfterAll;
+import org.junit.jupiter.api.BeforeAll;
 import org.junit.jupiter.api.Test;
 
 public class HashServiceTest {
 
+    private static File testFile;
+
+    @BeforeAll
+    public static void setup() throws Exception {
+        testFile = File.createTempFile("hashingTest", ".tmp");
+        Files.write(testFile.toPath(), "testdata".getBytes());
+    }
+
+    @AfterAll
+    public static void teardown() throws Exception {
+        if (testFile != null && testFile.exists()) {
+            testFile.delete();
+        }
+    }
+
+    @Test
+    public void testHashing_DefaultAlgorithm() throws Exception {
+        testHashing(new HashSpec());
+    }
+
+    @Test
+    public void testHashing_SHA256() throws Exception {
+        testHashing(new HashSpec(HashSpec.HashAlgorithm.SHA_256));
+    }
+
+    @Test
+    public void testHashing_CRC32() throws Exception {
+        testHashing(new HashSpec(HashSpec.HashAlgorithm.CRC));
+    }
+
     @Test
-    public void testHashing() throws Exception {
-        final File f = File.createTempFile("hashingTest", ".tmp");
-        Files.write(f.toPath(), "".getBytes());
-        final HashService hashService = new HashServiceImpl(new HashSpec());
-        hashService.verify(f.toPath(), hashService.hash(f.toPath()));
+    public void testHashing_xxHash64() throws Exception {
+        testHashing(new HashSpec(HashSpec.HashAlgorithm.XXHASH64));
     }
 
+    @Test
+    public void testHashing_None() throws Exception {
+        testHashing(new HashSpec(HashSpec.HashAlgorithm.NONE));
+    }
+
+    private void testHashing(HashSpec hashSpec) throws Exception {
+        final HashService hashService = new HashServiceImpl(hashSpec);
+        hashService.verify(testFile.toPath(), hashService.hash(testFile.toPath()));
+    }
 }

gcp/pom.xml

@@ -6,12 +6,12 @@
     <parent>
         <groupId>com.instaclustr</groupId>
         <artifactId>esop-parent</artifactId>
-        <version>4.0.1</version>
+        <version>4.0.2-SNAPSHOT</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 
     <artifactId>esop-gcp</artifactId>
-    <version>4.0.1</version>
+    <version>4.0.2-SNAPSHOT</version>
 
     <name>esop-gcp</name>
     <description>Backup and restoration tooling for Cassandra for GCP</description>

pom.xml

@@ -5,7 +5,7 @@
 
     <groupId>com.instaclustr</groupId>
     <artifactId>esop-parent</artifactId>
-    <version>4.0.1</version>
+    <version>4.0.2-SNAPSHOT</version>
     <packaging>pom</packaging>
 
     <name>esop-parent</name>

s3/pom.xml

@@ -6,12 +6,12 @@
     <parent>
         <groupId>com.instaclustr</groupId>
         <artifactId>esop-parent</artifactId>
-        <version>4.0.1</version>
+        <version>4.0.2-SNAPSHOT</version>
         <relativePath>../pom.xml</relativePath>
     </parent>
 
     <artifactId>esop-s3</artifactId>
-    <version>4.0.1</version>
+    <version>4.0.2-SNAPSHOT</version>
 
     <name>esop-s3</name>
     <description>Backup and restoration tooling for Cassandra for AWS S3</description>