mattmartin14 commented on code in PR #1534: URL: https://github.com/apache/iceberg-python/pull/1534#discussion_r1941309081
########## pyproject.toml: ########## @@ -842,6 +851,5026 @@ ignore_missing_imports = true module = "daft.*" ignore_missing_imports = true +[[tool.mypy.overrides]] +module = "pyiceberg_core.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "pyparsing.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "pyspark.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "strictyaml.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "sortedcontainers.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "sqlalchemy.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "Cython.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "setuptools.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "tenacity.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "pyarrow.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "pandas.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "snappy.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "zstandard.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "pydantic.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "pydantic_core.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "pytest.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "fastavro.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "mmh3.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "hive_metastore.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "thrift.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "requests_mock.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "click.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "rich.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "fsspec.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "s3fs.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "azure.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "adlfs.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "gcsfs.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "packaging.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "tests.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "boto3" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "botocore.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "mypy_boto3_glue.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "moto" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "aiobotocore.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "aiohttp.*" +ignore_missing_imports = true + +[[tool.mypy.overrides]] +module = "duckdb.*" Review Comment: I'm not sure what to tell you on this; my org has a strict way of accessing our internal python package repo; i was able to finally get poetry to work with it and ran the lock file. Not sure if this is harming anything or not ########## tests/table/test_merge_rows.py: ########## @@ -0,0 +1,397 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +from pyiceberg.catalog.sql import SqlCatalog +import os +import shutil + +_TEST_NAMESPACE = "test_ns" + +try: + from datafusion import SessionContext +except ModuleNotFoundError as e: + raise ModuleNotFoundError("For merge_rows, DataFusion needs to be installed") from e + +def get_test_warehouse_path(): + curr_dir = os.path.dirname(os.path.abspath(__file__)) + return f"{curr_dir}/warehouse" + +def get_sql_catalog(namespace: str) -> SqlCatalog: + warehouse_path = get_test_warehouse_path() + catalog = SqlCatalog( + "default", + **{ + "uri": f"sqlite:///:memory:", + "warehouse": f"file://{warehouse_path}", + }, + ) + + catalog.create_namespace(namespace=namespace) + return catalog + +def purge_warehouse(): + warehouse_path = get_test_warehouse_path() + + if os.path.exists(warehouse_path): + shutil.rmtree(warehouse_path) + +def show_iceberg_table(table, ctx: SessionContext): + import pyarrow.dataset as ds + table_name = "target" + if ctx.table_exist(table_name): + ctx.deregister_table(table_name) + ctx.register_dataset(table_name, ds.dataset(table.scan().to_arrow())) + ctx.sql(f"SELECT * FROM {table_name} limit 5").show() + +def show_df(df, ctx: SessionContext): + import pyarrow.dataset as ds + ctx.register_dataset("df", ds.dataset(df)) + ctx.sql("select * from df limit 10").show() + +def gen_source_dataset(start_row: int, end_row: int, composite_key: bool, add_dup: bool, ctx: SessionContext): + + additional_columns = ", t.order_id + 1000 as order_line_id" if composite_key else "" + + dup_row = f""" + UNION ALL + ( + SELECT t.order_id {additional_columns} + , date '2021-01-01' as order_date, 'B' as order_type + from t + limit 1 + ) + """ if add_dup else "" + + + sql = f""" + with t as (SELECT unnest(range({start_row},{end_row+1})) as order_id) + SELECT t.order_id {additional_columns} + , date '2021-01-01' as order_date, 'B' as order_type + from t + {dup_row} + """ + + #print(sql) + + df = ctx.sql(sql).to_arrow_table() + + + return df + +def gen_target_iceberg_table(start_row: int, end_row: int, composite_key: bool, ctx: SessionContext): + + additional_columns = ", t.order_id + 1000 as order_line_id" if composite_key else "" + + df = ctx.sql(f""" + with t as (SELECT unnest(range({start_row},{end_row+1})) as order_id) + SELECT t.order_id {additional_columns} + , date '2021-01-01' as order_date, 'A' as order_type + from t + """).to_arrow_table() + + catalog = get_sql_catalog(_TEST_NAMESPACE) + table = catalog.create_table(f"{_TEST_NAMESPACE}.target", df.schema) + + table.append(df) + + return table + +def test_merge_scenario_1a_simple(): + + """ + tests a single insert and update + """ + + ctx = SessionContext() + + table = gen_target_iceberg_table(1, 2, False, ctx) + source_df = gen_source_dataset(2, 3, False, False, ctx) + + res = table.merge_rows(df=source_df, join_cols=["order_id"]) + + rows_updated_should_be = 1 + rows_inserted_should_be = 1 + + assert res['rows_updated'] == rows_updated_should_be, f"rows updated should be {rows_updated_should_be}, but got {res['rows_updated']}" + assert res['rows_inserted'] == rows_inserted_should_be, f"rows inserted should be {rows_inserted_should_be}, but got {res['rows_inserted']}" + + purge_warehouse() + print('merge rows: test scenario 1a pass') + +def test_merge_scenario_1b_simple(): + + """ + tests a single insert and update; skips a row that does not need to be updated + """ + + ctx = SessionContext() + + df = ctx.sql(f""" + select 1 as order_id, date '2021-01-01' as order_date, 'A' as order_type + union all + select 2 as order_id, date '2021-01-01' as order_date, 'A' as order_type + """).to_arrow_table() + + catalog = get_sql_catalog(_TEST_NAMESPACE) + table = catalog.create_table(f"{_TEST_NAMESPACE}.target", df.schema) + + table.append(df) + + source_df = ctx.sql(f""" + select 1 as order_id, date '2021-01-01' as order_date, 'A' as order_type + union all + select 2 as order_id, date '2021-01-01' as order_date, 'B' as order_type + union all + select 3 as order_id, date '2021-01-01' as order_date, 'A' as order_type + """).to_arrow_table() + + res = table.merge_rows(df=source_df, join_cols=["order_id"]) + + rows_updated_should_be = 1 + rows_inserted_should_be = 1 + + assert res['rows_updated'] == rows_updated_should_be, f"rows updated should be {rows_updated_should_be}, but got {res['rows_updated']}" + assert res['rows_inserted'] == rows_inserted_should_be, f"rows inserted should be {rows_inserted_should_be}, but got {res['rows_inserted']}" + + purge_warehouse() + print('merge rows: test scenario 1b (skip 1 row) pass') + + +def test_merge_scenario_1c_simple(): + + """ + tests a single insert and update; primary key is a date column + """ + + ctx = SessionContext() + + df = ctx.sql(f""" + select date '2021-01-01' as order_date, 'A' as order_type + union all + select date '2021-01-02' as order_date, 'A' as order_type + """).to_arrow_table() + + catalog = get_sql_catalog(_TEST_NAMESPACE) + table = catalog.create_table(f"{_TEST_NAMESPACE}.target", df.schema) + + table.append(df) + + source_df = ctx.sql(f""" + select date '2021-01-01' as order_date, 'A' as order_type + union all + select date '2021-01-02' as order_date, 'B' as order_type + union all + select date '2021-01-03' as order_date, 'A' as order_type + """).to_arrow_table() + + res = table.merge_rows(df=source_df, join_cols=["order_date"]) + + rows_updated_should_be = 1 + rows_inserted_should_be = 1 + + assert res['rows_updated'] == rows_updated_should_be, f"rows updated should be {rows_updated_should_be}, but got {res['rows_updated']}" + assert res['rows_inserted'] == rows_inserted_should_be, f"rows inserted should be {rows_inserted_should_be}, but got {res['rows_inserted']}" + + purge_warehouse() + print('merge rows: test scenario 1c (date as key column) pass') + +def test_merge_scenario_1d_simple(): + + """ + tests a single insert and update; primary key is a string column + """ + + ctx = SessionContext() + + df = ctx.sql(f""" + select 'abc' as order_id, 'A' as order_type + union all + select 'def' as order_id, 'A' as order_type + """).to_arrow_table() + + catalog = get_sql_catalog(_TEST_NAMESPACE) + table = catalog.create_table(f"{_TEST_NAMESPACE}.target", df.schema) + + table.append(df) + + source_df = ctx.sql(f""" + select 'abc' as order_id, 'A' as order_type + union all + select 'def' as order_id, 'B' as order_type + union all + select 'ghi' as order_id, 'A' as order_type + """).to_arrow_table() + + res = table.merge_rows(df=source_df, join_cols=["order_id"]) + + rows_updated_should_be = 1 + rows_inserted_should_be = 1 + + assert res['rows_updated'] == rows_updated_should_be, f"rows updated should be {rows_updated_should_be}, but got {res['rows_updated']}" + assert res['rows_inserted'] == rows_inserted_should_be, f"rows inserted should be {rows_inserted_should_be}, but got {res['rows_inserted']}" + + purge_warehouse() + print('merge rows: test scenario 1d (string as key column) pass') + +def test_merge_scenario_2_10k_rows(): + + """ + tests merging 10000 rows on a single key to simulate larger workload + """ + + ctx = SessionContext() + + table = gen_target_iceberg_table(1, 10000, False, ctx) + source_df = gen_source_dataset(5001, 15000, False, False, ctx) + + + res = table.merge_rows(df=source_df, join_cols=["order_id"]) + + rows_updated_should_be = 5000 + rows_inserted_should_be = 5000 + + assert res['rows_updated'] == rows_updated_should_be, f"rows updated should be {rows_updated_should_be}, but got {res['rows_updated']}" + assert res['rows_inserted'] == rows_inserted_should_be, f"rows inserted should be {rows_inserted_should_be}, but got {res['rows_inserted']}" + + purge_warehouse() + print('merge rows: test scenario 2 pass') + +def test_merge_scenario_3_composite_key(): + + """ + tests merging 200 rows with a composite key + """ + + ctx = SessionContext() + + table = gen_target_iceberg_table(1, 200, True, ctx) + source_df = gen_source_dataset(101, 300, True, False, ctx) + + + res = table.merge_rows(df=source_df, join_cols=["order_id", "order_line_id"]) + + rows_updated_should_be = 100 + rows_inserted_should_be = 100 + + assert res['rows_updated'] == rows_updated_should_be, f"rows updated should be {rows_updated_should_be}, but got {res['rows_updated']}" + assert res['rows_inserted'] == rows_inserted_should_be, f"rows inserted should be {rows_inserted_should_be}, but got {res['rows_inserted']}" + + purge_warehouse() + print('merge rows: composite keys test pass') + +def test_merge_update_only(): + + """ + tests explicit merge options to do update only + """ + + ctx = SessionContext() + + table = gen_target_iceberg_table(1, 1000, False, ctx) + source_df = gen_source_dataset(501, 1500, False, False, ctx) + + res = table.merge_rows(df=source_df, join_cols=["order_id"], when_matched_update_all=True, when_not_matched_insert_all=False) + + rows_updated_should_be = 500 + rows_inserted_should_be = 0 + + assert res['rows_updated'] == rows_updated_should_be, f"rows updated should be {rows_updated_should_be}, but got {res['rows_updated']}" + assert res['rows_inserted'] == rows_inserted_should_be, f"rows inserted should be {rows_inserted_should_be}, but got {res['rows_inserted']}" + + purge_warehouse() + print('merge rows: update only pass') + +def test_merge_insert_only(): + """ + tests explicit merge options to do insert only + """ + + ctx = SessionContext() + + table = gen_target_iceberg_table(1, 1000, False, ctx) + source_df = gen_source_dataset(501, 1500, False, False, ctx) + + res = table.merge_rows(df=source_df, join_cols=["order_id"], when_matched_update_all=False, when_not_matched_insert_all=True) + + rows_updated_should_be = 0 + rows_inserted_should_be = 500 + + assert res['rows_updated'] == rows_updated_should_be, f"rows updated should be {rows_updated_should_be}, but got {res['rows_updated']}" + assert res['rows_inserted'] == rows_inserted_should_be, f"rows inserted should be {rows_inserted_should_be}, but got {res['rows_inserted']}" + + purge_warehouse() + print('merge rows: insert only pass') + +def test_merge_source_dups(): + + """ + tests duplicate rows in source + """ + + ctx = SessionContext() + + table = gen_target_iceberg_table(1, 10, False, ctx) + source_df = gen_source_dataset(5, 15, False, True, ctx) + + res = table.merge_rows(df=source_df, join_cols=["order_id"]) + + error_msgs = res['error_msgs'] + + assert 'Duplicate rows found in source dataset' in error_msgs, f"error message should contain 'Duplicate rows found in source dataset', but got {error_msgs}" + + purge_warehouse() + print('merge rows: source dups test pass') + +def test_key_cols_misaligned(): + + """ + tests join columns missing from one of the tables + """ + + ctx = SessionContext() + + #generate dummy target iceberg table + df = ctx.sql("select 1 as order_id, date '2021-01-01' as order_date, 'A' as order_type").to_arrow_table() + + catalog = get_sql_catalog(_TEST_NAMESPACE) + table = catalog.create_table(f"{_TEST_NAMESPACE}.target", df.schema) + + table.append(df) + + df_src = ctx.sql("select 1 as item_id, date '2021-05-01' as order_date, 'B' as order_type").to_arrow_table() + + res = table.merge_rows(df=df_src, join_cols=['order_id']) + error_msgs = res['error_msgs'] + + assert 'Join columns missing in tables' in error_msgs, f"error message should contain 'Join columns missing in tables', but got {error_msgs}" + + purge_warehouse() + + print('merge rows: key cols misaligned test pass') + +if __name__ == "__main__": Review Comment: removed if __name__ == "__main__" -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org For additional commands, e-mail: issues-h...@iceberg.apache.org
