leekeiabstraction commented on code in PR #169: URL: https://github.com/apache/fluss-rust/pull/169#discussion_r2704504257
########## crates/fluss/src/client/table/upsert.rs: ########## @@ -0,0 +1,372 @@ +// 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. + +use crate::client::table::writer::{TableWriter, UpsertWriter}; +use crate::client::{WriteFormat, WriteRecord, WriterClient}; +use crate::error::Error::IllegalArgument; +use crate::error::Result; +use crate::metadata::{KvFormat, RowType, TableInfo, TablePath}; +use crate::row::compacted::CompactedRow; +use crate::row::encode::{KeyEncoder, KeyEncoderFactory, RowEncoder, RowEncoderFactory}; +use crate::row::field_getter::FieldGetter; +use crate::row::{BinaryRow, InternalRow}; +use std::sync::Arc; + +use bitvec::prelude::bitvec; +use bytes::Bytes; + +#[allow(dead_code)] +pub struct TableUpsert { + table_path: TablePath, + table_info: TableInfo, + writer_client: Arc<WriterClient>, + target_columns: Option<Arc<Vec<usize>>>, +} + +#[allow(dead_code)] +impl TableUpsert { + pub fn new( + table_path: TablePath, + table_info: TableInfo, + writer_client: Arc<WriterClient>, + ) -> Self { + Self { + table_path, + table_info, + writer_client, + target_columns: None, + } + } + + pub fn partial_update(&self, target_columns: Option<Vec<usize>>) -> Result<Self> { + if let Some(columns) = &target_columns { + let num_columns = self.table_info.row_type().fields().len(); + + if let Some(&invalid_column) = columns.iter().find(|&&col| col >= num_columns) { + return Err(IllegalArgument { + message: format!( + "Invalid target column index: {invalid_column} for table {}. The table only has {num_columns} columns.", + self.table_path + ), + }); + } + } + + Ok(Self { + table_path: self.table_path.clone(), + table_info: self.table_info.clone(), + writer_client: self.writer_client.clone(), + target_columns: target_columns.map(Arc::new), + }) + } + + pub fn partial_update_with_column_names(&self, target_column_names: &[&str]) -> Result<Self> { + let row_type = self.table_info.row_type(); + let col_indices: Vec<(&str, Option<usize>)> = target_column_names + .iter() + .map(|col_name| (*col_name, row_type.get_field_index(col_name))) + .collect(); + + if let Some((missing_name, _)) = col_indices.iter().find(|(_, ix)| ix.is_none()) { + return Err(IllegalArgument { + message: format!( + "Cannot find target column `{}` for table {}.", + missing_name, self.table_path + ), + }); + } + + let valid_col_indices: Vec<usize> = col_indices + .into_iter() + .map(|(_, index)| index.unwrap()) + .collect(); + + self.partial_update(Some(valid_col_indices)) + } + + pub fn create_writer(&self) -> Result<impl UpsertWriter> { + UpsertWriterFactory::create( + Arc::new(self.table_path.clone()), + Arc::new(self.table_info.clone()), + self.target_columns.clone(), + Arc::clone(&self.writer_client), + ) + } +} + +#[allow(dead_code)] +struct UpsertWriterImpl<RE> +where + RE: RowEncoder, +{ + table_path: Arc<TablePath>, + writer_client: Arc<WriterClient>, + // TODO: Partitioning + // partition_field_getter: Option<Box<dyn KeyEncoder>>, + primary_key_encoder: Box<dyn KeyEncoder>, + target_columns: Option<Arc<Vec<usize>>>, + bucket_key_encoder: Option<Box<dyn KeyEncoder>>, + kv_format: KvFormat, + write_format: WriteFormat, + row_encoder: RE, + field_getters: Box<[FieldGetter]>, + table_info: Arc<TableInfo>, +} + +#[allow(dead_code)] +struct UpsertWriterFactory; + +#[allow(dead_code)] +impl UpsertWriterFactory { + pub fn create( + table_path: Arc<TablePath>, + table_info: Arc<TableInfo>, + partial_update_columns: Option<Arc<Vec<usize>>>, + writer_client: Arc<WriterClient>, + ) -> Result<impl UpsertWriter> { + let data_lake_format = &table_info.table_config.get_datalake_format()?; + let row_type = table_info.row_type(); + let physical_pks = table_info.get_physical_primary_keys(); + + let names = table_info.get_schema().auto_increment_col_names(); + + Self::sanity_check( + row_type, + &table_info.primary_keys, + names, + &partial_update_columns, + )?; + + let primary_key_encoder = KeyEncoderFactory::of(row_type, physical_pks, data_lake_format)?; + let bucket_key_encoder = if !table_info.is_default_bucket_key() { + Some(KeyEncoderFactory::of( + row_type, + table_info.get_bucket_keys(), + data_lake_format, + )?) + } else { + None + }; + + let kv_format = table_info.get_table_config().get_kv_format()?; + let write_format = WriteFormat::from_kv_format(&kv_format)?; + + let field_getters = FieldGetter::create_field_getters(row_type); + + Ok(UpsertWriterImpl { + table_path, + writer_client, + primary_key_encoder, + target_columns: partial_update_columns, + bucket_key_encoder, + kv_format: kv_format.clone(), + write_format, + row_encoder: RowEncoderFactory::create(kv_format, row_type.clone())?, Review Comment: Nice catch. I started with InternalRow, similar to Java side but later opted for accepting CompactedRow directly because the trait and struct signatures were getting unwieldily big. I'll have another stab. -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
