PARQUET-620: Ensure metadata is written only once

Author: Uwe L. Korn <uwelk@xhochy.com> Closes apache#108 from xhochy/parquet-620 and squashes the following commits: da122ad [Uwe L. Korn] Ensure metadata is written only once Change-Id: I7653597fdf69c961545d6c978fdc1367267adee7
wesm · May 18, 2016 · e1e0d28 · e1e0d28
1 parent 3ff3b58
commit e1e0d28
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 6 deletions.
diff --git a/cpp/src/parquet/file/writer-internal.cc b/cpp/src/parquet/file/writer-internal.cc
@@ -173,13 +173,16 @@ std::unique_ptr<ParquetFileWriter::Contents> FileSerializer::Open(
 }
 
 void FileSerializer::Close() {
-  if (row_group_writer_) { row_group_writer_->Close(); }
-  row_group_writer_.reset();
+  if (is_open_) {
+    if (row_group_writer_) { row_group_writer_->Close(); }
+    row_group_writer_.reset();
 
-  // Write magic bytes and metadata
-  WriteMetaData();
+    // Write magic bytes and metadata
+    WriteMetaData();
 
-  sink_->Close();
+    sink_->Close();
+    is_open_ = false;
+  }
 }
 
 int FileSerializer::num_columns() const {
@@ -238,7 +241,11 @@ void FileSerializer::WriteMetaData() {
 
 FileSerializer::FileSerializer(std::shared_ptr<OutputStream> sink,
     std::shared_ptr<GroupNode>& schema, MemoryAllocator* allocator = default_allocator())
-    : sink_(sink), allocator_(allocator), num_row_groups_(0), num_rows_(0) {
+    : sink_(sink),
+      allocator_(allocator),
+      num_row_groups_(0),
+      num_rows_(0),
+      is_open_(true) {
   schema_.Init(schema);
   StartFile();
 }

diff --git a/cpp/src/parquet/file/writer-internal.h b/cpp/src/parquet/file/writer-internal.h
@@ -130,6 +130,7 @@ class FileSerializer : public ParquetFileWriter::Contents {
   MemoryAllocator* allocator_;
   int num_row_groups_;
   int num_rows_;
+  bool is_open_;
   std::unique_ptr<RowGroupWriter> row_group_writer_;
 
   void StartFile();