fails fast when footer size larger than Int.MaxValue

apache · Jan 24, 2025 · 107cb79 · 107cb79
1 parent be5ada2
commit 107cb79
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 3 deletions.
diff --git a/parquet-hadoop/src/main/java/org/apache/parquet/hadoop/ParquetFileReader.java b/parquet-hadoop/src/main/java/org/apache/parquet/hadoop/ParquetFileReader.java
@@ -589,7 +589,12 @@ private static final ParquetMetadata readFooter(
     long fileMetadataLengthIndex = fileLen - magic.length - FOOTER_LENGTH_SIZE;
     LOG.debug("reading footer index at {}", fileMetadataLengthIndex);
     f.seek(fileMetadataLengthIndex);
-    int fileMetadataLength = readIntLittleEndian(f);
+    long readFileMetadataLength = readIntLittleEndian(f) & 0xFFFFFFFFL;
+    if (readFileMetadataLength > Integer.MAX_VALUE) {
+      throw new RuntimeException("footer is too large: " + readFileMetadataLength + "to be read");
+    }
+    int fileMetadataLength = (int) readFileMetadataLength;
+
     f.readFully(magic);
 
     boolean encryptedFooterMode;

diff --git a/parquet-hadoop/src/main/java/org/apache/parquet/hadoop/ParquetFileWriter.java b/parquet-hadoop/src/main/java/org/apache/parquet/hadoop/ParquetFileWriter.java
@@ -1915,7 +1915,7 @@ private static void serializeFooter(
       out.write(serializedFooter);
       out.write(signature);
       LOG.debug("{}: footer and signature length = {}", out.getPos(), (out.getPos() - footerIndex));
-      BytesUtils.writeIntLittleEndian(out, toIntWithCheck(out.getPos() - footerIndex, "page"));
+      BytesUtils.writeIntLittleEndian(out, toIntWithCheck(out.getPos() - footerIndex, "footer"));
       out.write(MAGIC);
       return;
     }
@@ -1925,7 +1925,7 @@ private static void serializeFooter(
     writeFileCryptoMetaData(fileEncryptor.getFileCryptoMetaData(), out);
     byte[] footerAAD = AesCipher.createFooterAAD(fileEncryptor.getFileAAD());
     writeFileMetaData(parquetMetadata, out, fileEncryptor.getFooterEncryptor(), footerAAD);
-    int combinedMetaDataLength = toIntWithCheck(out.getPos() - cryptoFooterIndex, "page");
+    int combinedMetaDataLength = toIntWithCheck(out.getPos() - cryptoFooterIndex, "footer");
     LOG.debug("{}: crypto metadata and footer length = {}", out.getPos(), combinedMetaDataLength);
     BytesUtils.writeIntLittleEndian(out, combinedMetaDataLength);
     out.write(EFMAGIC);