[arrow] branch master updated: PARQUET-1582: [C++] Add ToString method to ColumnDescriptor

[wesm]

This is an automated email from the ASF dual-hosted git repository.

wesm pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/arrow.git


The following commit(s) were added to refs/heads/master by this push:
     new 5870080  PARQUET-1582: [C++] Add ToString method to ColumnDescriptor
5870080 is described below

commit 587008066724c2dd59be273e14f5de74edad3d4d
Author: Micah Kornfield <emkornfi...@gmail.com>
AuthorDate: Mon May 20 16:41:37 2019 -0500

    PARQUET-1582: [C++] Add ToString method to ColumnDescriptor
    
    Author: Micah Kornfield <emkornfi...@gmail.com>
    Author: Micah Kornfield <mic...@google.com>
    
    Closes #4338 from emkornfield/add_debug_string and squashes the following 
commits:
    
    25980a37a <Micah Kornfield> Add unit test
    9568a6744 <Micah Kornfield> Add ToString method to ColumnDescripotor
---
 cpp/src/parquet/schema-test.cc | 24 ++++++++++++++++++++++++
 cpp/src/parquet/schema.cc      | 27 ++++++++++++++++++++++++---
 cpp/src/parquet/schema.h       |  2 ++
 3 files changed, 50 insertions(+), 3 deletions(-)

diff --git a/cpp/src/parquet/schema-test.cc b/cpp/src/parquet/schema-test.cc
index 36fe8e6..80293c1 100644
--- a/cpp/src/parquet/schema-test.cc
+++ b/cpp/src/parquet/schema-test.cc
@@ -566,6 +566,16 @@ TEST(TestColumnDescriptor, TestAttrs) {
   ASSERT_EQ(Type::BYTE_ARRAY, descr.physical_type());
 
   ASSERT_EQ(-1, descr.type_length());
+  ASSERT_EQ(
+      R"(column descriptor = {
+  name: name
+  path: 
+  physical_type: BYTE_ARRAY
+  logical_type: UTF8
+  max_definition_level: 4
+  max_repetition_level: 1
+})",
+      descr.ToString());
 
   // Test FIXED_LEN_BYTE_ARRAY
   node = PrimitiveNode::Make("name", Repetition::OPTIONAL, 
Type::FIXED_LEN_BYTE_ARRAY,
@@ -574,6 +584,20 @@ TEST(TestColumnDescriptor, TestAttrs) {
 
   ASSERT_EQ(Type::FIXED_LEN_BYTE_ARRAY, descr.physical_type());
   ASSERT_EQ(12, descr.type_length());
+
+  ASSERT_EQ(
+      R"(column descriptor = {
+  name: name
+  path: 
+  physical_type: FIXED_LEN_BYTE_ARRAY
+  logical_type: DECIMAL
+  max_definition_level: 4
+  max_repetition_level: 1
+  length: 12
+  precision: 10
+  scale: 4
+})",
+      descr.ToString());
 }
 
 class TestSchemaDescriptor : public ::testing::Test {
diff --git a/cpp/src/parquet/schema.cc b/cpp/src/parquet/schema.cc
index 0a5668d..9206aba 100644
--- a/cpp/src/parquet/schema.cc
+++ b/cpp/src/parquet/schema.cc
@@ -16,17 +16,15 @@
 // under the License.
 
 #include "parquet/schema.h"
-#include "parquet/schema-internal.h"
 
 #include <algorithm>
 #include <cstring>
 #include <memory>
 #include <string>
 #include <utility>
-
 #include "arrow/util/logging.h"
-
 #include "parquet/exception.h"
+#include "parquet/schema-internal.h"
 #include "parquet/thrift.h"
 
 using parquet::format::SchemaElement;
@@ -746,6 +744,29 @@ std::string SchemaDescriptor::ToString() const {
   return ss.str();
 }
 
+std::string ColumnDescriptor::ToString() const {
+  std::ostringstream ss;
+  ss << "column descriptor = {" << std::endl
+     << "  name: " << name() << std::endl
+     << "  path: " << path()->ToDotString() << std::endl
+     << "  physical_type: " << TypeToString(physical_type()) << std::endl
+     << "  logical_type: " << LogicalTypeToString(logical_type()) << std::endl
+     << "  max_definition_level: " << max_definition_level() << std::endl
+     << "  max_repetition_level: " << max_repetition_level() << std::endl;
+
+  if (physical_type() == ::parquet::Type::FIXED_LEN_BYTE_ARRAY) {
+    ss << "  length: " << type_length() << std::endl;
+  }
+
+  if (logical_type() == parquet::LogicalType::DECIMAL) {
+    ss << "  precision: " << type_precision() << std::endl
+       << "  scale: " << type_scale() << std::endl;
+  }
+
+  ss << "}";
+  return ss.str();
+}
+
 int ColumnDescriptor::type_scale() const {
   return primitive_node_->decimal_metadata().scale;
 }
diff --git a/cpp/src/parquet/schema.h b/cpp/src/parquet/schema.h
index 76920c0..62cf95c 100644
--- a/cpp/src/parquet/schema.h
+++ b/cpp/src/parquet/schema.h
@@ -353,6 +353,8 @@ class PARQUET_EXPORT ColumnDescriptor {
 
   const schema::NodePtr& schema_node() const { return node_; }
 
+  std::string ToString() const;
+
   int type_length() const;
 
   int type_precision() const;