tscottcoombes1 commented on code in PR #1534: URL: https://github.com/apache/iceberg-python/pull/1534#discussion_r1947207894
########## pyiceberg/table/upsert_util.py: ########## @@ -0,0 +1,158 @@ + +# 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 pyarrow import Table as pyarrow_table +import pyarrow as pa +from pyarrow import compute as pc +from pyiceberg import table as pyiceberg_table + +from pyiceberg.expressions import ( + BooleanExpression, + And, + EqualTo, + Or, + In, +) + +def get_filter_list(df: pyarrow_table, join_cols: list) -> BooleanExpression: + + unique_keys = df.select(join_cols).group_by(join_cols).aggregate([]) + + pred = None + + if len(join_cols) == 1: + pred = In(join_cols[0], unique_keys[0].to_pylist()) + else: + pred = Or(*[ + And(*[ + EqualTo(col, row[col]) + for col in join_cols + ]) + for row in unique_keys.to_pylist() + ]) + + return pred + +def dups_check_in_source(df: pyarrow_table, join_cols: list) -> bool: + """ + This function checks if there are duplicate rows in the source table based on the join columns. + It returns True if there are duplicate rows in the source table, otherwise it returns False. + """ + # Check for duplicates in the source table + source_dup_count = len( + df.select(join_cols) + .group_by(join_cols) + .aggregate([([], "count_all")]) + .filter(pc.field("count_all") > 1) + ) + + return source_dup_count > 0 + +def get_rows_to_update(source_table: pa.Table, target_table: pa.Table, join_cols: list) -> pa.Table: + + """ + This function takes the source_table, trims it down to rows that match in both source and target. + It then does a scan for the non-key columns to see if any are mis-aligned before returning the final row set to update + """ + + all_columns = set(source_table.column_names) + join_cols_set = set(join_cols) + + non_key_cols = list(all_columns - join_cols_set) + + + match_expr = None + + for col in join_cols: + target_values = target_table.column(col).to_pylist() + expr = pc.field(col).isin(target_values) + + if match_expr is None: + match_expr = expr + else: + match_expr = match_expr & expr Review Comment: sorry not clear. So these two functions: `get_rows_to_update` & `get_rows_to_insert` both share some functionality. They look for columns that match. then for the insert you then flip it, and find all the columns that don't match. so instead of calculating the matches twice, you could: * have a `get_matches` function which returns the mask * pass it to both of these functions * and then for the inserts do `pa.compute.invert(mask)` ########## pyiceberg/table/upsert_util.py: ########## @@ -0,0 +1,158 @@ + +# 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 pyarrow import Table as pyarrow_table +import pyarrow as pa +from pyarrow import compute as pc +from pyiceberg import table as pyiceberg_table + +from pyiceberg.expressions import ( + BooleanExpression, + And, + EqualTo, + Or, + In, +) + +def get_filter_list(df: pyarrow_table, join_cols: list) -> BooleanExpression: + + unique_keys = df.select(join_cols).group_by(join_cols).aggregate([]) + + pred = None + + if len(join_cols) == 1: + pred = In(join_cols[0], unique_keys[0].to_pylist()) + else: + pred = Or(*[ + And(*[ + EqualTo(col, row[col]) + for col in join_cols + ]) + for row in unique_keys.to_pylist() + ]) + + return pred + +def dups_check_in_source(df: pyarrow_table, join_cols: list) -> bool: + """ + This function checks if there are duplicate rows in the source table based on the join columns. + It returns True if there are duplicate rows in the source table, otherwise it returns False. + """ + # Check for duplicates in the source table + source_dup_count = len( + df.select(join_cols) + .group_by(join_cols) + .aggregate([([], "count_all")]) + .filter(pc.field("count_all") > 1) + ) + + return source_dup_count > 0 + +def get_rows_to_update(source_table: pa.Table, target_table: pa.Table, join_cols: list) -> pa.Table: + + """ + This function takes the source_table, trims it down to rows that match in both source and target. + It then does a scan for the non-key columns to see if any are mis-aligned before returning the final row set to update + """ + + all_columns = set(source_table.column_names) + join_cols_set = set(join_cols) + + non_key_cols = list(all_columns - join_cols_set) + + + match_expr = None + + for col in join_cols: + target_values = target_table.column(col).to_pylist() + expr = pc.field(col).isin(target_values) + + if match_expr is None: + match_expr = expr + else: + match_expr = match_expr & expr + + + matching_source_rows = source_table.filter(match_expr) + + rows_to_update = [] + + for index in range(matching_source_rows.num_rows): + + source_row = matching_source_rows.slice(index, 1) + + + target_filter = None + + for col in join_cols: + target_value = source_row.column(col)[0].as_py() + if target_filter is None: + target_filter = pc.field(col) == target_value + else: + target_filter = target_filter & (pc.field(col) == target_value) + + matching_target_row = target_table.filter(target_filter) + + if matching_target_row.num_rows > 0: + needs_update = False + + for non_key_col in non_key_cols: + source_value = source_row.column(non_key_col)[0].as_py() + target_value = matching_target_row.column(non_key_col)[0].as_py() + + if source_value != target_value: + needs_update = True + break + + if needs_update: + rows_to_update.append(source_row) + + if rows_to_update: + rows_to_update_table = pa.concat_tables(rows_to_update) + else: + rows_to_update_table = pa.Table.from_arrays([], names=source_table.column_names) + + common_columns = set(source_table.column_names).intersection(set(target_table.column_names)) + rows_to_update_table = rows_to_update_table.select(list(common_columns)) + + return rows_to_update_table + +def get_rows_to_insert(source_table: pa.Table, target_table: pa.Table, join_cols: list) -> pa.Table: + + source_filter_expr = None + + for col in join_cols: + + target_values = target_table.column(col).to_pylist() + expr = pc.field(col).isin(target_values) + + if source_filter_expr is None: + source_filter_expr = expr + else: + source_filter_expr = source_filter_expr & expr Review Comment: I got it to work like this: ``` import pyarrow as pa import pandas as pd df_1 = pd.DataFrame({'name': ["tom", "matt", "nonmatch"], 'age': [10, 20, 100], "third": [1,2,3]}) tbl_1 = pa.Table.from_pandas(df_1) df_2 = pd.DataFrame({'name': ["tom", "harry", "nonmatch2"], 'age': [10, 20, 0], "third": [1,2,1]}) tbl_2 = pa.Table.from_pandas(df_2) masks = [ pa.compute.is_in(tbl_1[col], value_set=tbl_2[col]) for col in ["name","age","third"]] mask = [pa.compute.all(x) for x in zip(*masks)] tbl_1.filter(mask) ``` -- 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
