D#/add pandas pyarrow verified source

sources/arrow_pandas/__init__.py

@@ -0,0 +1,3 @@
+"""
+This source provides a simple starting point for loading data from arrow tables or pandas dataframes
+"""

sources/arrow_pandas/example_resources.py

@@ -0,0 +1,50 @@
+import dlt
+
+from typing import Generator
+
+from dlt.common import pendulum
+import pandas as pd
+import pyarrow as pa
+
+
+@dlt.resource(name="orders", write_disposition="append")
+def orders() -> Generator[pd.DataFrame, None, None]:
+    # this is example data, you will get this from somewhere on your resource function
+    EXAMPLE_ORDERS_DATA_FRAME = pd.DataFrame(
+        data={
+            "order_id": [1, 2, 3],
+            "customer_id": [1, 2, 3],
+            "ordered_at": [
+                pendulum.DateTime(2021, 1, 1, 4, 5, 6),
+                pendulum.DateTime(2021, 1, 3, 4, 5, 6),
+                pendulum.DateTime(2021, 1, 6, 4, 5, 6),
+            ],
+            "order_amount": [100.0, 200.0, 300.0],
+        }
+    )
+
+    # we can yield dataframes here, you will usually read them from a file or
+    # receive them from another library
+    yield EXAMPLE_ORDERS_DATA_FRAME
+
+
+@dlt.resource(
+    name="customers",
+    write_disposition="merge",
+    primary_key="customer_id",
+    merge_key="customer_id",
+)
+def customers() -> Generator[pd.DataFrame, None, None]:
+    # we can yield arrow tables here, you will usually read them from a file or
+    # receive them from another library
+    EXAMPLE_CUSTOMERS_DATA_FRAME = pd.DataFrame(
+        data={
+            "customer_id": [1, 2, 3],
+            "name": ["Alice", "Bob", "Charlie"],
+            "age": [25, 30, 35],
+        }
+    )
+
+    # here we convert our dataframe to an arrow table, usually you would just yield the
+    # dataframe if you have it, this is for demonstration purposes
+    yield pa.Table.from_pandas(EXAMPLE_CUSTOMERS_DATA_FRAME)

sources/arrow_pandas/requirements.txt

@@ -0,0 +1,2 @@
+pandas>=2.0.0
+dlt>=0.4.0

sources/arrow_pandas_pipeline.py

@@ -0,0 +1,14 @@
+"""Very simple pipeline, to be used as a starting point for pandas or arrow pipelines.
+
+"""
+
+import dlt
+from arrow_pandas.example_resources import orders, customers
+
+
+if __name__ == "__main__":
+    pipeline = dlt.pipeline(
+        "orders_pipeline", destination="duckdb", dataset_name="orders_dataset"
+    )
+    # run both resources
+    pipeline.run([orders, customers])

tests/arrow_pandas/test_arrow_pandas_source.py

@@ -0,0 +1,20 @@
+import dlt
+import pytest
+
+from sources.arrow_pandas.example_resources import orders, customers
+from tests.utils import ALL_DESTINATIONS, assert_load_info, load_table_counts
+
+
+@pytest.mark.parametrize("destination_name", ALL_DESTINATIONS)
+def test_example_resources(destination_name: str) -> None:
+    """Simple test for the example resources."""
+
+    p = dlt.pipeline("orders_pipeline", destination=destination_name, full_refresh=True)
+    orders.apply_hints(incremental=dlt.sources.incremental("ordered_at"))
+
+    # run pipeline
+    info = p.run([orders(), customers()])
+
+    # check that the data was loaded
+    assert_load_info(info)
+    assert load_table_counts(p, "orders", "customers") == {"orders": 3, "customers": 3}