apacheGH-35279: [C++][Parquet] Tools: enhancement Parquet print stats (…

…apache#35262) ### Rationale for this change Enhancement printing value for parquet stats ### What changes are included in this PR? 1. Make DebugPrintStats for RecordReader better 2. Make Stats printer better ### Are these changes tested? No ### Are there any user-facing changes? No * Closes: apache#35279 Authored-by: mwish <maplewish117@gmail.com> Signed-off-by: Will Jones <willjones127@gmail.com>
LoongArch-Python · May 11, 2023 · 555c1b0 · 555c1b0
1 parent 20c6665
commit 555c1b0
Show file tree

Hide file tree

Showing 2 changed files with 24 additions and 13 deletions.
diff --git a/cpp/src/parquet/column_reader.cc b/cpp/src/parquet/column_reader.cc
@@ -1989,17 +1989,21 @@ class TypedRecordReader : public TypedColumnReaderImpl<DType>,
 
     const T* vals = reinterpret_cast<const T*>(this->values());
 
-    std::cout << "def levels: ";
-    for (int64_t i = 0; i < total_levels_read; ++i) {
-      std::cout << def_levels[i] << " ";
+    if (leaf_info_.def_level > 0) {
+      std::cout << "def levels: ";
+      for (int64_t i = 0; i < total_levels_read; ++i) {
+        std::cout << def_levels[i] << " ";
+      }
+      std::cout << std::endl;
     }
-    std::cout << std::endl;
 
-    std::cout << "rep levels: ";
-    for (int64_t i = 0; i < total_levels_read; ++i) {
-      std::cout << rep_levels[i] << " ";
+    if (leaf_info_.rep_level > 0) {
+      std::cout << "rep levels: ";
+      for (int64_t i = 0; i < total_levels_read; ++i) {
+        std::cout << rep_levels[i] << " ";
+      }
+      std::cout << std::endl;
     }
-    std::cout << std::endl;
 
     std::cout << "values: ";
     for (int64_t i = 0; i < this->values_written(); ++i) {

diff --git a/cpp/src/parquet/printer.cc b/cpp/src/parquet/printer.cc
@@ -282,11 +282,18 @@ void ParquetFilePrinter::JSONPrint(std::ostream& stream, std::list<int> selected
       if (column_chunk->is_stats_set()) {
         stream << "\"True\", \"Stats\": {";
         std::string min = stats->EncodeMin(), max = stats->EncodeMax();
-        stream << "\"NumNulls\": \"" << stats->null_count() << "\", "
-               << "\"DistinctValues\": \"" << stats->distinct_count() << "\", "
-               << "\"Max\": \"" << FormatStatValue(descr->physical_type(), max) << "\", "
-               << "\"Min\": \"" << FormatStatValue(descr->physical_type(), min)
-               << "\" },";
+        stream << "\"NumNulls\": \"" << stats->null_count();
+        if (stats->HasDistinctCount()) {
+          stream << "\", "
+                 << "\"DistinctValues\": \"" << stats->distinct_count();
+        }
+        if (stats->HasMinMax()) {
+          stream << "\", "
+                 << "\"Max\": \"" << FormatStatValue(descr->physical_type(), max)
+                 << "\", "
+                 << "\"Min\": \"" << FormatStatValue(descr->physical_type(), min);
+        }
+        stream << "\" },";
       } else {
         stream << "\"False\",";
       }