Merge pull request #4869 from volodya-lombrozo/4852-outofmemory-fix

fix(#4852): resolve OutOfMemoryError for large files in Cache.java

Author: yegor256

eo-maven-plugin/src/main/java/org/eolang/maven/Cache.java

@@ -98,15 +98,18 @@ private Path hash(final Path tail) {
      * @return Base64-encoded SHA-256 hash
      * @throws NoSuchAlgorithmException If SHA-256 algorithm is not available
      * @throws IOException If an I/O error occurs reading the file
-     * @todo #4846:30min OutOfMemoryError for large files in cache.
-     *  The sha method reads the entire file into memory using Files.readAllBytes(file) which
-     *  could cause OutOfMemoryError for large files. Consider using a streaming approach with
-     *  MessageDigest.update() in a loop to hash the file in chunks, similar to how it's typically
-     *  done for large file hashing operations.
      */
     private static String sha(final Path file) throws NoSuchAlgorithmException, IOException {
         final MessageDigest digest = MessageDigest.getInstance("SHA-256");
-        final byte[] hash = digest.digest(Files.readAllBytes(file));
+        try (var stream = Files.newInputStream(file)) {
+            final byte[] buffer = new byte[8192];
+            int read = stream.read(buffer);
+            while (read != -1) {
+                digest.update(buffer, 0, read);
+                read = stream.read(buffer);
+            }
+        }
+        final byte[] hash = digest.digest();
         return Base64.getEncoder().encodeToString(hash);
     }
 }

eo-maven-plugin/src/test/java/org/eolang/maven/CacheTest.java

@@ -15,6 +15,7 @@
 import java.security.NoSuchAlgorithmException;
 import java.util.Base64;
 import java.util.concurrent.atomic.AtomicInteger;
+import java.util.stream.IntStream;
 import org.hamcrest.MatcherAssert;
 import org.hamcrest.Matchers;
 import org.junit.jupiter.api.Test;
@@ -158,4 +159,63 @@ void writesCorrectShaHash(@Mktmp final Path temp) throws IOException, NoSuchAlgo
             )
         );
     }
+
+    @Test
+    void generatesCorrectHashForLargeFile(
+        @Mktmp final Path temp
+    ) throws IOException, NoSuchAlgorithmException {
+        final var cache = temp.resolve("cache");
+        Files.createDirectories(cache);
+        final var source = temp.resolve("large.txt");
+        final int lines = 100_000;
+        final StringBuilder builder = new StringBuilder(lines * 10);
+        IntStream.range(0, lines).forEach(
+            i -> builder.append("Line ").append(i).append('\n')
+        );
+        final String content = builder.toString();
+        Files.writeString(source, content, StandardCharsets.UTF_8);
+        final var target = temp.resolve("out.txt");
+        final var tail = source.getFileName();
+        new Cache(cache, p -> content).apply(source, target, tail);
+        MatcherAssert.assertThat(
+            "SHA-256 hash file has incorrect content for large file",
+            Files.readString(
+                cache.resolve(String.format("%s.sha256", tail)),
+                StandardCharsets.UTF_8
+            ),
+            Matchers.equalTo(
+                Base64.getEncoder().encodeToString(
+                    MessageDigest.getInstance("SHA-256")
+                        .digest(content.getBytes(StandardCharsets.UTF_8))
+                )
+            )
+        );
+    }
+
+    @Test
+    void generatesCorrectHashForTinyFile(
+        @Mktmp final Path temp
+    ) throws IOException, NoSuchAlgorithmException {
+        final var cache = temp.resolve("cache");
+        Files.createDirectories(cache);
+        final var source = temp.resolve("tiny.txt");
+        final String content = "x";
+        Files.writeString(source, content, StandardCharsets.UTF_8);
+        final var target = temp.resolve("out.txt");
+        final var tail = source.getFileName();
+        new Cache(cache, p -> content).apply(source, target, tail);
+        MatcherAssert.assertThat(
+            "SHA-256 hash file has incorrect content for tiny file",
+            Files.readString(
+                cache.resolve(String.format("%s.sha256", tail)),
+                StandardCharsets.UTF_8
+            ),
+            Matchers.equalTo(
+                Base64.getEncoder().encodeToString(
+                    MessageDigest.getInstance("SHA-256")
+                        .digest(content.getBytes(StandardCharsets.UTF_8))
+                )
+            )
+        );
+    }
 }