Merge pull request #1394 from mabel-dev/#1393

#1393

opteryx/__version__.py

@@ -1,4 +1,4 @@
-__build__ = 224
+__build__ = 226
 
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.

opteryx/operators/function_dataset_node.py

@@ -31,7 +31,8 @@
 
 def _generate_series(**kwargs):
     value_array = series.generate_series(*kwargs["args"])
-    return [{kwargs["columns"][0].schema_column.identity: value} for value in value_array]
+    column_name = kwargs["columns"][0].schema_column.identity
+    return pyarrow.Table.from_arrays([value_array], [column_name])
 
 
 def _unnest(**kwargs):
@@ -41,7 +42,8 @@ def _unnest(**kwargs):
     else:
         list_items = kwargs["args"][0].value
     column_name = kwargs["columns"][0].schema_column.identity
-    return [{column_name: row} for row in list_items]
+
+    return pyarrow.Table.from_arrays([list_items], [column_name])
 
 
 def _values(**parameters):
@@ -108,8 +110,10 @@ def execute(self) -> Generator:
 
         if isinstance(data, list):
             table = pyarrow.Table.from_pylist(data)
-        if hasattr(data, "arrow"):
+        elif hasattr(data, "arrow"):
             table = data.arrow()
+        else:
+            table = data
 
         self.statistics.rows_read += table.num_rows
         self.statistics.columns_read += len(table.column_names)

opteryx/utils/series.py

@@ -13,6 +13,7 @@
 import numpy
 from orso.types import OrsoTypes
 
+from opteryx.exceptions import InvalidFunctionParameterError
 from opteryx.exceptions import SqlError
 from opteryx.utils import dates
 
@@ -39,7 +40,9 @@ def generate_series(*args):
             raise SqlError("generate_series for dates needs start, end, and interval parameters")
         return dates.date_range(*arg_vals)
 
-    raise SqlError("Unsupported value for GENERATE_SERIES")
+    raise InvalidFunctionParameterError(
+        "Unsupported value for GENERATE_SERIES, must be date or numeric series."
+    )
 
 
 def numeric_range(*args) -> numpy.ndarray:

opteryx/virtual_datasets/no_table_data.py

@@ -26,10 +26,9 @@
 def read(*args):
     import pyarrow
 
-    # Create a PyArrow schema with one column called '$column' of integer type
-    _schema = pyarrow.schema([("$column", pyarrow.int64())])
-    # Create a PyArrow table with the given schema and one row
-    return pyarrow.Table.from_arrays([[None]], schema=_schema)
+    # Create a PyArrow table with one column and one row
+
+    return pyarrow.Table.from_arrays([[0]], ["$column"])  # schema=_schema)
 
 
 def schema():

opteryx/virtual_datasets/planet_data.py

@@ -71,21 +71,20 @@ def read(end_date=None, *args):
     full_set = pyarrow.Table.from_arrays(data, column_names)
 
     if end_date is None:
-        end_date = datetime.datetime.utcnow()
+        return full_set
 
     # Make the planet data act like it supports temporality
-    mask = [True, True, True, True, True, True, True, True, True]
-    if end_date < datetime.datetime(1930, 3, 13):
-        # March 13, 1930 - Pluto discovered by Clyde William Tombaugh
-        mask = [True, True, True, True, True, True, True, True, False]
-    if end_date < datetime.datetime(1846, 11, 13):
-        # November 13, 1846 - Neptune
-        mask = [True, True, True, True, True, True, True, False, False]
     if end_date < datetime.datetime(1781, 4, 26):
         # April 26, 1781 - Uranus discovered by Sir William Herschel
-        mask = [True, True, True, True, True, True, False, False, False]
+        return full_set.take([0, 1, 2, 3, 4, 5])
+    if end_date < datetime.datetime(1846, 11, 13):
+        # November 13, 1846 - Neptune
+        return full_set.take([0, 1, 2, 3, 4, 5, 7])
+    if end_date < datetime.datetime(1930, 3, 13):
+        # March 13, 1930 - Pluto discovered by Clyde William Tombaugh
+        return full_set.take([0, 1, 2, 3, 4, 5, 6, 7])
 
-    return full_set.filter(mask)
+    return full_set
 
 
 def schema():