fix reading empty parquet (#2615)

pola-rs · Feb 11, 2022 · e4d6f6d · e4d6f6d
1 parent c65a6d5
commit e4d6f6d
Show file tree

Hide file tree

Showing 3 changed files with 18 additions and 3 deletions.
diff --git a/polars/polars-io/src/parquet/read_impl.rs b/polars/polars-io/src/parquet/read_impl.rs
@@ -4,6 +4,7 @@ use crate::parquet::predicates::collect_statistics;
 use crate::predicates::{apply_predicate, arrow_schema_to_empty_df, PhysicalIoExpr};
 use crate::utils::apply_projection;
 use crate::RowCount;
+use arrow::array::new_empty_array;
 use arrow::io::parquet::read;
 use arrow::io::parquet::read::{to_deserializer, FileMetaData};
 use polars_core::prelude::*;
@@ -85,8 +86,11 @@ pub fn read_parquet<R: MmapBytesReader>(
                             remaining_rows,
                             Some(chunk_size),
                         )?;
-
-                        Series::try_from((field.name.as_str(), iter.next().unwrap()?))
+                        let arr = match iter.next() {
+                            Some(arr) => arr?,
+                            None => Arc::from(new_empty_array(field.data_type.clone())),
+                        };
+                        Series::try_from((field.name.as_str(), arr))
                     })
                     .collect::<Result<Vec<_>>>()
             })?
@@ -99,7 +103,11 @@ pub fn read_parquet<R: MmapBytesReader>(
                     let mut iter =
                         to_deserializer(columns, field.clone(), remaining_rows, Some(chunk_size))?;
 
-                    Series::try_from((field.name.as_str(), iter.next().unwrap()?))
+                    let arr = match iter.next() {
+                        Some(arr) => arr?,
+                        None => Arc::from(new_empty_array(field.data_type.clone())),
+                    };
+                    Series::try_from((field.name.as_str(), arr))
                 })
                 .collect::<Result<Vec<_>>>()?
         };

diff --git a/py-polars/null.parquet b/py-polars/null.parquet
diff --git a/py-polars/tests/lazy_io/test_parquet.py b/py-polars/tests/lazy_io/test_parquet.py
@@ -33,3 +33,10 @@ def test_categorical_parquet_statistics() -> None:
             .collect()
         )
     assert df.shape == (4, 3)
+
+
+def test_null_parquet() -> None:
+    df = pl.DataFrame([pl.Series("foo", [], dtype=pl.Int8)])
+    df.to_parquet("null.parquet")
+    out = pl.read_parquet("null.parquet")
+    assert out.frame_equal(df)