Support "A column is known to be entirely NULL" in PruningPredicate

datafusion-examples/examples/pruning.rs

@@ -163,6 +163,11 @@ impl PruningStatistics for MyCatalog {
         None
     }
 
+    fn row_counts(&self, _column: &Column) -> Option<ArrayRef> {
+        // In this example, we know nothing about the number of rows in each file
+        None
+    }
+
     fn contained(
         &self,
         _column: &Column,

datafusion/core/src/datasource/physical_plan/parquet/mod.rs

@@ -1870,7 +1870,7 @@ mod tests {
 
         assert_contains!(
             &display,
-            "pruning_predicate=c1_min@0 != bar OR bar != c1_max@1"
+            "pruning_predicate=CASE WHEN c1_null_count@2 = c1_row_count@3 THEN false ELSE c1_min@0 != bar OR bar != c1_max@1 END"
         );
 
         assert_contains!(&display, r#"predicate=c1@0 != bar"#);

datafusion/core/src/datasource/physical_plan/parquet/page_filter.rs

@@ -547,6 +547,10 @@ impl<'a> PruningStatistics for PagesPruningStatistics<'a> {
         }
     }
 
+    fn row_counts(&self, _column: &datafusion_common::Column) -> Option<ArrayRef> {
+        None
+    }
+
     fn contained(
         &self,
         _column: &datafusion_common::Column,

datafusion/core/src/datasource/physical_plan/parquet/row_groups.rs

@@ -195,6 +195,10 @@ impl PruningStatistics for BloomFilterStatistics {
         None
     }
 
+    fn row_counts(&self, _column: &Column) -> Option<ArrayRef> {
+        None
+    }
+
     /// Use bloom filters to determine if we are sure this column can not
     /// possibly contain `values`
     ///
@@ -328,6 +332,10 @@ impl<'a> PruningStatistics for RowGroupPruningStatistics<'a> {
         scalar.to_array().ok()
     }
 
+    fn row_counts(&self, _column: &Column) -> Option<ArrayRef> {
+        None
+    }
+
     fn contained(
         &self,
         _column: &Column,