apacheGH-42100: [C++][Parquet] ParquetFilePrinter::JSONPrint print le…

…ngth of FLBA (apache#41981) ### Rationale for this change Print FLBA length in `ParquetFilePrinter::JSONPrint` ### What changes are included in this PR? Print FLBA length in `ParquetFilePrinter::JSONPrint` ### Are these changes tested? no need ### Are there any user-facing changes? no * GitHub Issue: apache#42100 Authored-by: mwish <[email protected]> Signed-off-by: mwish <[email protected]>
mathworks · Jun 11, 2024 · 6597467 · 6597467
1 parent 64b1109
commit 6597467
Show file tree

Hide file tree

Showing 2 changed files with 40 additions and 10 deletions.
diff --git a/cpp/src/parquet/printer.cc b/cpp/src/parquet/printer.cc
@@ -246,7 +246,8 @@ void ParquetFilePrinter::JSONPrint(std::ostream& stream, std::list<int> selected
     const ColumnDescriptor* descr = file_metadata->schema()->Column(i);
     stream << "     { \"Id\": \"" << i << "\","
            << " \"Name\": \"" << descr->path()->ToDotString() << "\","
-           << " \"PhysicalType\": \"" << TypeToString(descr->physical_type()) << "\","
+           << " \"PhysicalType\": \""
+           << TypeToString(descr->physical_type(), descr->type_length()) << "\","
            << " \"ConvertedType\": \"" << ConvertedTypeToString(descr->converted_type())
            << "\","
            << " \"LogicalType\": " << (descr->logical_type())->ToJSON() << " }";

diff --git a/cpp/src/parquet/reader_test.cc b/cpp/src/parquet/reader_test.cc
@@ -1077,7 +1077,24 @@ Column 1
   ASSERT_EQ(header_output + dump_output, ss_dump.str());
 }
 
-TEST(TestJSONWithLocalFile, JSONOutput) {
+class TestJSONWithLocalFile : public ::testing::Test {
+ public:
+  static std::string ReadFromLocalFile(std::string_view local_file_name) {
+    std::stringstream ss;
+    // empty list means print all
+    std::list<int> columns;
+
+    auto reader =
+        ParquetFileReader::OpenFile(data_file(local_file_name.data()),
+                                    /*memory_map=*/false, default_reader_properties());
+    ParquetFilePrinter printer(reader.get());
+    printer.JSONPrint(ss, columns, local_file_name.data());
+
+    return ss.str();
+  }
+};
+
+TEST_F(TestJSONWithLocalFile, JSONOutput) {
   std::string json_output = R"###({
   "FileName": "alltypes_plain.parquet",
   "Version": "1.0",
@@ -1131,16 +1148,28 @@ TEST(TestJSONWithLocalFile, JSONOutput) {
 }
 )###";
 
-  std::stringstream ss;
-  // empty list means print all
-  std::list<int> columns;
+  std::string json_content = ReadFromLocalFile("alltypes_plain.parquet");
+  ASSERT_EQ(json_output, json_content);
+}
 
-  auto reader =
-      ParquetFileReader::OpenFile(alltypes_plain(), false, default_reader_properties());
-  ParquetFilePrinter printer(reader.get());
-  printer.JSONPrint(ss, columns, "alltypes_plain.parquet");
+TEST_F(TestJSONWithLocalFile, JSONOutputFLBA) {
+  // min-max stats for FLBA contains non-utf8 output, so we don't check
+  // the whole json output.
+  std::string json_content = ReadFromLocalFile("fixed_length_byte_array.parquet");
+
+  std::string json_contains = R"###({
+  "FileName": "fixed_length_byte_array.parquet",
+  "Version": "1.0",
+  "CreatedBy": "parquet-mr version 1.13.0-SNAPSHOT (build d057b39d93014fe40f5067ee4a33621e65c91552)",
+  "TotalRows": "1000",
+  "NumberOfRowGroups": "1",
+  "NumberOfRealColumns": "1",
+  "NumberOfColumns": "1",
+  "Columns": [
+     { "Id": "0", "Name": "flba_field", "PhysicalType": "FIXED_LEN_BYTE_ARRAY(4)", "ConvertedType": "NONE", "LogicalType": {"Type": "None"} }
+  ])###";
 
-  ASSERT_EQ(json_output, ss.str());
+  EXPECT_THAT(json_content, testing::HasSubstr(json_contains));
 }
 
 TEST(TestFileReader, BufferedReadsWithDictionary) {