arrow/csv/writer.h

0001 // Licensed to the Apache Software Foundation (ASF) under one
0002 // or more contributor license agreements.  See the NOTICE file
0003 // distributed with this work for additional information
0004 // regarding copyright ownership.  The ASF licenses this file
0005 // to you under the Apache License, Version 2.0 (the
0006 // "License"); you may not use this file except in compliance
0007 // with the License.  You may obtain a copy of the License at
0008 //
0009 //   http://www.apache.org/licenses/LICENSE-2.0
0010 //
0011 // Unless required by applicable law or agreed to in writing,
0012 // software distributed under the License is distributed on an
0013 // "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
0014 // KIND, either express or implied.  See the License for the
0015 // specific language governing permissions and limitations
0016 // under the License.
0017
0018 #pragma once
0019
0020 #include <memory>
0021
0022 #include "arrow/csv/options.h"
0023 #include "arrow/io/interfaces.h"
0024 #include "arrow/ipc/type_fwd.h"
0025 #include "arrow/record_batch.h"
0026 #include "arrow/table.h"
0027
0028 namespace arrow {
0029 namespace csv {
0030
0031 // Functionality for converting Arrow data to Comma separated value text.
0032 // This library supports all primitive types that can be cast to a StringArray or
0033 // a LargeStringArray.
0034 // It applies to following formatting rules:
0035 //  - For non-binary types no quotes surround values.  Nulls are represented as the empty
0036 //  string.
0037 //  - For binary types all non-null data is quoted (and quotes within data are escaped
0038 //  with an additional quote).
0039 //    Null values are empty and unquoted.
0040
0041 /// \defgroup csv-write-functions High-level functions for writing CSV files
0042 /// @{
0043
0044 /// \brief Convert table to CSV and write the result to output.
0045 /// Experimental
0046 ARROW_EXPORT Status WriteCSV(const Table& table, const WriteOptions& options,
0047                              arrow::io::OutputStream* output);
0048 /// \brief Convert batch to CSV and write the result to output.
0049 /// Experimental
0050 ARROW_EXPORT Status WriteCSV(const RecordBatch& batch, const WriteOptions& options,
0051                              arrow::io::OutputStream* output);
0052 /// \brief Convert batches read through a RecordBatchReader
0053 /// to CSV and write the results to output.
0054 /// Experimental
0055 ARROW_EXPORT Status WriteCSV(const std::shared_ptr<RecordBatchReader>& reader,
0056                              const WriteOptions& options,
0057                              arrow::io::OutputStream* output);
0058
0059 /// @}
0060
0061 /// \defgroup csv-writer-factories Functions for creating an incremental CSV writer
0062 /// @{
0063
0064 /// \brief Create a new CSV writer. User is responsible for closing the
0065 /// actual OutputStream.
0066 ///
0067 /// \param[in] sink output stream to write to
0068 /// \param[in] schema the schema of the record batches to be written
0069 /// \param[in] options options for serialization
0070 /// \return Result<std::shared_ptr<RecordBatchWriter>>
0071 ARROW_EXPORT
0072 Result<std::shared_ptr<ipc::RecordBatchWriter>> MakeCSVWriter(
0073     std::shared_ptr<io::OutputStream> sink, const std::shared_ptr<Schema>& schema,
0074     const WriteOptions& options = WriteOptions::Defaults());
0075
0076 /// \brief Create a new CSV writer.
0077 ///
0078 /// \param[in] sink output stream to write to (does not take ownership)
0079 /// \param[in] schema the schema of the record batches to be written
0080 /// \param[in] options options for serialization
0081 /// \return Result<std::shared_ptr<RecordBatchWriter>>
0082 ARROW_EXPORT
0083 Result<std::shared_ptr<ipc::RecordBatchWriter>> MakeCSVWriter(
0084     io::OutputStream* sink, const std::shared_ptr<Schema>& schema,
0085     const WriteOptions& options = WriteOptions::Defaults());
0086
0087 /// @}
0088
0089 }  // namespace csv
0090 }  // namespace arrow