From 3a8738316294f26262a13740835ad81a09ca6628 Mon Sep 17 00:00:00 2001
From: mwish <maplewish117@gmail.com>
Date: Fri, 24 Nov 2023 21:49:05 +0800
Subject: [PATCH] Resolve comments

---
 cpp/src/parquet/bloom_filter.cc |  9 +++++++++
 cpp/src/parquet/bloom_filter.h  |  4 ++--
 cpp/src/parquet/printer.cc      | 16 ++++++++--------
 3 files changed, 19 insertions(+), 10 deletions(-)

diff --git a/cpp/src/parquet/bloom_filter.cc b/cpp/src/parquet/bloom_filter.cc
index fb543ea086455..b1d552accc36e 100644
--- a/cpp/src/parquet/bloom_filter.cc
+++ b/cpp/src/parquet/bloom_filter.cc
@@ -144,6 +144,15 @@ BlockSplitBloomFilter BlockSplitBloomFilter::Deserialize(
     bloom_filter.Init(header_buf->data() + header_size, bloom_filter_size);
     return bloom_filter;
   }
+  if (bloom_filter_length && *bloom_filter_length < bloom_filter_size + header_size) {
+    // We know the bloom filter data size, but the length is not enough to read the
+    // entire bloom filter.
+    std::stringstream ss;
+    ss << "Bloom filter length (" << bloom_filter_length.value()
+       << ") is not enough to read the entire bloom filter (size: "
+       << bloom_filter_size + header_size << ").";
+    throw ParquetException(ss.str());
+  }
   // We have read a part of the bloom filter already, copy it to the target buffer
   // and read the remaining part from the InputStream.
   auto buffer = AllocateBuffer(properties.memory_pool(), bloom_filter_size);
diff --git a/cpp/src/parquet/bloom_filter.h b/cpp/src/parquet/bloom_filter.h
index c5eed0ec002e9..909563d013fed 100644
--- a/cpp/src/parquet/bloom_filter.h
+++ b/cpp/src/parquet/bloom_filter.h
@@ -310,8 +310,8 @@ class PARQUET_EXPORT BlockSplitBloomFilter : public BloomFilter {
   /// a Bloom filter from a parquet filter.
   ///
   /// @param properties The parquet reader properties.
-  /// @param input_stream The input stream from which to construct the Bloom filter.
-  /// @param bloom_filter_length The length of the Serialized Bloom filter including
+  /// @param input_stream The input stream from which to construct the bloom filter.
+  /// @param bloom_filter_length The length of the serialized bloom filter including
   /// header.
   /// @return The BlockSplitBloomFilter.
   static BlockSplitBloomFilter Deserialize(
diff --git a/cpp/src/parquet/printer.cc b/cpp/src/parquet/printer.cc
index 46085d19e4ad1..f11397ab96ed8 100644
--- a/cpp/src/parquet/printer.cc
+++ b/cpp/src/parquet/printer.cc
@@ -324,19 +324,19 @@ void ParquetFilePrinter::JSONPrint(std::ostream& stream, std::list<int> selected
         stream << "\"}";
       }
 
-      if (column_chunk->GetOffsetIndexLocation()) {
-        auto location = column_chunk->GetOffsetIndexLocation().value();
-        // Output OffsetIndex {offset, length}
-        stream << "\", OffsetIndex {"
+      if (column_chunk->GetColumnIndexLocation()) {
+        auto location = column_chunk->GetColumnIndexLocation().value();
+        // Output ColumnIndex {offset, length}
+        stream << "\", ColumnIndex {"
                << "\"offset\": \"" << location.offset;
         stream << "\", \"length\": \"" << location.length;
         stream << "\"}";
       }
 
-      if (column_chunk->GetColumnIndexLocation()) {
-        auto location = column_chunk->GetColumnIndexLocation().value();
-        // Output ColumnIndex {offset, length}
-        stream << "\", ColumnIndex {"
+      if (column_chunk->GetOffsetIndexLocation()) {
+        auto location = column_chunk->GetOffsetIndexLocation().value();
+        // Output OffsetIndex {offset, length}
+        stream << "\", OffsetIndex {"
                << "\"offset\": \"" << location.offset;
         stream << "\", \"length\": \"" << location.length;
         stream << "\"}";