Two improvements to make hashing large files much faster

1. Use an index when updating the hasher instead of recopying the input array every time
2. Load the file into the hasher one piece at a time instead of all at
once to avoid memory issues

src/com/github/rctcwyvrn/blake3/Blake3.java

@@ -1,15 +1,16 @@
 package com.github.rctcwyvrn.blake3;
 
 import java.io.File;
+import java.io.FileInputStream;
 import java.io.IOException;
+import java.io.InputStream;
 import java.nio.ByteBuffer;
 import java.nio.ByteOrder;
 import java.nio.charset.StandardCharsets;
-import java.nio.file.Files;
 import java.util.Arrays;
 
 /**
- * Translation of the Blake3 reference implemenetation from Rust to Java
+ * Translation of the Blake3 reference implementation from Rust to Java
  * BLAKE3 Source: https://github.com/BLAKE3-team/BLAKE3
  * Translator: rctcwyvrn
  */
@@ -207,7 +208,8 @@ private int startFlag(){
         }
 
         private void update(byte[] input) {
-            while (input.length != 0) {
+            int currPos = 0;
+            while (currPos < input.length) {
 
                 // Chain the next 64 byte block into this chunk/node
                 if (blockLen == BLOCK_LEN) {
@@ -222,11 +224,11 @@ private void update(byte[] input) {
 
                 // Take bytes out of the input and update
                 int want = BLOCK_LEN - this.blockLen; // How many bytes we need to fill up the current block
-                int canTake = Math.min(want, input.length);
-                if (canTake >= 0) System.arraycopy(input, 0, block, blockLen, canTake);
+                int canTake = Math.min(want, input.length - currPos);
 
+                System.arraycopy(input, currPos, block, blockLen, canTake);
                 blockLen += canTake;
-                input = Arrays.copyOfRange(input, canTake, input.length);
+                currPos+=canTake;
             }
         }
 
@@ -283,17 +285,29 @@ public Blake3(String context){
      * @throws IOException If the file does not exist
      */
     public void update(File file) throws IOException {
-        update(Files.readAllBytes(file.toPath()));
+        // Update the hasher 4kb at a time to avoid memory issues when hashing large files
+        try(InputStream ios = new FileInputStream(file)){
+            byte[] buffer = new byte[4096];
+            int read = 0;
+            while((read = ios.read(buffer)) != -1){
+                if(read == buffer.length) {
+                    update(buffer);
+                } else {
+                    update(Arrays.copyOfRange(buffer, 0, read));
+                }
+            }
+        }
     }
 
     /**
      * Appends new data to the hash tree
      * @param input Data to be added
      */
     public void update(byte[] input){
-        while(input.length != 0) {
+        int currPos = 0;
+        while(currPos < input.length) {
 
-            // If this chunk has chained in 16 64 bytes of input, add it's CV to the stack
+            // If this chunk has chained in 16 64 bytes of input, add its CV to the stack
             if (chunkState.len() == CHUNK_LEN) {
                 int[] chunkCV = chunkState.createNode().chainingValue();
                 long totalChunks = chunkState.chunkCounter + 1;
@@ -302,9 +316,9 @@ public void update(byte[] input){
             }
 
             int want = CHUNK_LEN - chunkState.len();
-            int take = Math.min(want, input.length);
-            chunkState.update(Arrays.copyOfRange(input, 0, take));
-            input = Arrays.copyOfRange(input, take, input.length);
+            int take = Math.min(want, input.length - currPos);
+            chunkState.update(Arrays.copyOfRange(input, currPos, currPos + take));
+            currPos+=take;
         }
     }
 

src/com/github/rctcwyvrn/blake3/Examples.java

@@ -14,14 +14,13 @@ public static void main(String[] args){
         System.out.println("Success: " + hexhash);
 
         // Hashing files
-        // Warning: Very slow for large files due to lack of optimizations :c
-        String filename = "LICENSE";
+        String filename = "License";
         try {
             Blake3 fileHasher = new Blake3();
             fileHasher.update(new File(filename));
             String filehash = fileHasher.hexdigest();
             if (!filehash.equals("381f3baeddb0ce5202ac9528ecc787c249901e74528ba2cbc5546567a2e0bd33"))
-                throw new AssertionError();
+                throw new AssertionError(filehash);
             System.out.println("Success: " + filehash);
         } catch (IOException e) {
             System.err.println("File not found: " + filename);