SeqAn3 3.4.0-rc.4
The Modern C++ library for sequence analysis.
Loading...
Searching...
No Matches
io/sequence_file/output.hpp
Go to the documentation of this file.
1// SPDX-FileCopyrightText: 2006-2025 Knut Reinert & Freie Universität Berlin
2// SPDX-FileCopyrightText: 2016-2025 Knut Reinert & MPI für molekulare Genetik
3// SPDX-License-Identifier: BSD-3-Clause
4
10#pragma once
11
12#include <cassert>
13#include <filesystem>
14#include <fstream>
15#include <ranges>
16#include <string>
17#include <variant>
18#include <vector>
19
23#include <seqan3/io/detail/record.hpp>
26#include <seqan3/io/record.hpp>
40
41namespace seqan3
42{
43
44// ----------------------------------------------------------------------------
45// sequence_file_output
46// ----------------------------------------------------------------------------
47
62template <detail::fields_specialisation selected_field_ids_ = fields<field::seq, field::id, field::qual>,
63 detail::type_list_of_sequence_file_output_formats valid_formats_ =
64 type_list<format_embl, format_fasta, format_fastq, format_genbank, format_sam>>
66{
67public:
73 using selected_field_ids = selected_field_ids_;
75 using valid_formats = valid_formats_;
77 using stream_char_type = char;
79
82
83 static_assert(
84 []() constexpr
85 {
86 for (field f : selected_field_ids::as_array)
87 if (!field_ids::contains(f))
88 return false;
89 return true;
90 }(),
91 "You selected a field that is not valid for sequence files, please refer to the documentation "
92 "of sequence_file_output::field_ids for the accepted values.");
93
100 using value_type = void;
102 using reference = void;
104 using const_reference = void;
106 using size_type = void;
110 using iterator = detail::out_file_iterator<sequence_file_output>;
112 using const_iterator = void;
114 using sentinel = std::default_sentinel_t;
116
132
154 selected_field_ids const & SEQAN3_DOXYGEN_ONLY(fields_tag) = selected_field_ids{},
155 std::ios_base::openmode openmode = std::ios_base::out | std::ios::binary) :
156 primary_stream{new std::ofstream{}, stream_deleter_default}
157 {
158 primary_stream->rdbuf()->pubsetbuf(stream_buffer.data(), stream_buffer.size());
159 static_cast<std::basic_ofstream<char> *>(primary_stream.get())
160 ->open(filename, openmode | std::ios_base::out | std::ios::binary);
161
162 if (!primary_stream->good())
163 throw file_open_error{"Could not open file " + filename.string() + " for writing."};
164
165 // possibly add intermediate compression stream
166 secondary_stream = detail::make_secondary_ostream(*primary_stream, filename);
167
168 // initialise format handler or throw if format is not found
169 detail::set_format(format, filename);
170 }
171
187 template <output_stream stream_t, sequence_file_output_format file_format>
188 requires std::same_as<typename std::remove_reference_t<stream_t>::char_type, stream_char_type>
189 sequence_file_output(stream_t & stream,
190 file_format const & SEQAN3_DOXYGEN_ONLY(format_tag),
191 selected_field_ids const & SEQAN3_DOXYGEN_ONLY(fields_tag) = selected_field_ids{}) :
192 primary_stream{&stream, stream_deleter_noop},
193 secondary_stream{&stream, stream_deleter_noop},
194 format{detail::sequence_file_output_format_exposer<file_format>{}}
195 {
196 static_assert(list_traits::contains<file_format, valid_formats>,
197 "You selected a format that is not in the valid_formats of this file.");
198 }
199
201 template <output_stream stream_t, sequence_file_output_format file_format>
202 requires std::same_as<typename std::remove_reference_t<stream_t>::char_type, stream_char_type>
203 sequence_file_output(stream_t && stream,
204 file_format const & SEQAN3_DOXYGEN_ONLY(format_tag),
205 selected_field_ids const & SEQAN3_DOXYGEN_ONLY(fields_tag) = selected_field_ids{}) :
206 primary_stream{new stream_t{std::move(stream)}, stream_deleter_default},
207 secondary_stream{&*primary_stream, stream_deleter_noop},
208 format{detail::sequence_file_output_format_exposer<file_format>{}}
209 {
210 static_assert(list_traits::contains<file_format, valid_formats>,
211 "You selected a format that is not in the valid_formats of this file.");
212 }
214
236 iterator begin() noexcept
237 {
238 return {*this};
239 }
240
255 sentinel end() noexcept
256 {
257 return {};
258 }
259
278 template <typename record_t>
279 void push_back(record_t && r)
280 requires detail::record_like<record_t>
281 {
282 write_record(detail::get_or_ignore<field::seq>(r),
283 detail::get_or_ignore<field::id>(r),
284 detail::get_or_ignore<field::qual>(r));
285 }
286
308 template <typename tuple_t>
309 void push_back(tuple_t && t)
310 requires tuple_like<tuple_t> && (!detail::record_like<tuple_t>)
311 {
312 // index_of might return npos, but this will be handled well by get_or_ignore (and just return ignore)
313 write_record(detail::get_or_ignore<selected_field_ids::index_of(field::seq)>(t),
314 detail::get_or_ignore<selected_field_ids::index_of(field::id)>(t),
315 detail::get_or_ignore<selected_field_ids::index_of(field::qual)>(t));
316 }
317
341 template <typename arg_t, typename... arg_types>
342 void emplace_back(arg_t && arg, arg_types &&... args)
343 {
344 push_back(std::tie(arg, args...));
345 }
346
368 template <std::ranges::input_range rng_t>
371 {
372 for (auto && record : range)
373 push_back(std::forward<decltype(record)>(record));
374 return *this;
375 }
376
404 template <std::ranges::input_range rng_t>
407 {
408 f = range;
409 return f;
410 }
411
413 template <std::ranges::input_range rng_t>
416 {
417 f = range;
418 return std::move(f);
419 }
421
424
429 {
430 return *secondary_stream;
431 }
433
434protected:
437 std::vector<char> stream_buffer{std::vector<char>(1'000'000)};
438
446 static void stream_deleter_noop(std::basic_ostream<stream_char_type> *)
447 {}
449 static void stream_deleter_default(std::basic_ostream<stream_char_type> * ptr)
450 {
451 delete ptr;
452 }
453
455 stream_ptr_t primary_stream{nullptr, stream_deleter_noop};
457 stream_ptr_t secondary_stream{nullptr, stream_deleter_noop};
458
460 using format_type =
461 typename detail::variant_from_tags<valid_formats, detail::sequence_file_output_format_exposer>::type;
463 format_type format;
465
467 template <typename seq_t, typename id_t, typename qual_t>
468 void write_record(seq_t && seq, id_t && id, qual_t && qual)
469 {
470 assert(!format.valueless_by_exception());
472 [&](auto & f)
473 {
474 {
475 f.write_sequence_record(*secondary_stream, options, seq, id, qual);
476 }
477 },
478 format);
479 }
480
482 friend iterator;
483};
484
491template <output_stream stream_t, sequence_file_output_format file_format>
493 file_format const &)
496
498template <output_stream stream_t, sequence_file_output_format file_format>
500 file_format const &)
503
505template <output_stream stream_t,
506 sequence_file_output_format file_format,
507 detail::fields_specialisation selected_field_ids>
509 file_format const &,
511
513template <output_stream stream_t,
514 sequence_file_output_format file_format,
515 detail::fields_specialisation selected_field_ids>
517 file_format const &,
520} // namespace seqan3
The generic concept for sequence file out formats.
Definition sequence_file/output_format_concept.hpp:88
A class for writing sequence files, e.g. FASTA, FASTQ ...
Definition io/sequence_file/output.hpp:66
sequence_file_output & operator=(sequence_file_output const &)=delete
Copy assignment is explicitly deleted, because you can't have multiple access to the same file.
sequence_file_output(stream_t &, file_format const &, selected_field_ids const &) -> sequence_file_output< selected_field_ids, type_list< file_format > >
This is an overloaded member function, provided for convenience. It differs from the above function o...
sequence_file_output(stream_t &&, file_format const &) -> sequence_file_output< typename sequence_file_output<>::selected_field_ids, type_list< file_format > >
This is an overloaded member function, provided for convenience. It differs from the above function o...
void const_iterator
The const iterator type is void, because files are not const-iterable.
Definition io/sequence_file/output.hpp:112
char stream_char_type
Character type of the stream(s).
Definition io/sequence_file/output.hpp:77
std::default_sentinel_t sentinel
The type returned by end().
Definition io/sequence_file/output.hpp:114
sequence_file_output(stream_t &&, file_format const &, selected_field_ids const &) -> sequence_file_output< selected_field_ids, type_list< file_format > >
Deduction guide for given stream, file format and field ids.
void push_back(tuple_t &&t)
Write a record in form of a std::tuple to the file.
Definition io/sequence_file/output.hpp:309
sequence_file_output(stream_t &stream, file_format const &format_tag, selected_field_ids const &fields_tag=selected_field_ids{})
Construct from an existing stream and with specified format.
Definition io/sequence_file/output.hpp:189
void push_back(record_t &&r)
Write a seqan3::record to the file.
Definition io/sequence_file/output.hpp:279
sequence_file_output(sequence_file_output &&)=default
Move construction is defaulted.
sequence_file_output & operator=(sequence_file_output &&)=default
Move assignment is defaulted.
sequence_file_output(sequence_file_output const &)=delete
Copy construction is explicitly deleted, because you can't have multiple access to the same file.
sequence_file_output(stream_t &, file_format const &) -> sequence_file_output< typename sequence_file_output<>::selected_field_ids, type_list< file_format > >
Deduction guide for given stream and file format.
valid_formats_ valid_formats
A seqan3::type_list with the possible formats.
Definition io/sequence_file/output.hpp:75
sequence_file_output(std::filesystem::path filename, selected_field_ids const &fields_tag=selected_field_ids{}, std::ios_base::openmode openmode=std::ios_base::out|std::ios::binary)
Construct from filename.
Definition io/sequence_file/output.hpp:153
selected_field_ids_ selected_field_ids
A seqan3::fields list with the fields selected for the record.
Definition io/sequence_file/output.hpp:73
sequence_file_output(stream_t &&stream, file_format const &format_tag, selected_field_ids const &fields_tag=selected_field_ids{})
This is an overloaded member function, provided for convenience. It differs from the above function o...
Definition io/sequence_file/output.hpp:203
friend sequence_file_output operator|(rng_t &&range, sequence_file_output &&f)
This is an overloaded member function, provided for convenience. It differs from the above function o...
Definition io/sequence_file/output.hpp:414
~sequence_file_output()=default
Destructor is defaulted.
void emplace_back(arg_t &&arg, arg_types &&... args)
Write a record to the file by passing individual fields.
Definition io/sequence_file/output.hpp:342
sequence_file_output & operator=(rng_t &&range)
Write a range of records (or tuples) to the file.
Definition io/sequence_file/output.hpp:369
void value_type
The value type (void).
Definition io/sequence_file/output.hpp:100
sentinel end() noexcept
Returns a sentinel for comparison with iterator.
Definition io/sequence_file/output.hpp:255
void reference
The reference type (void).
Definition io/sequence_file/output.hpp:102
friend sequence_file_output & operator|(rng_t &&range, sequence_file_output &f)
Write a range of records (or tuples) to the file.
Definition io/sequence_file/output.hpp:405
void const_reference
The const reference type (void).
Definition io/sequence_file/output.hpp:104
sequence_file_output()=delete
Default constructor is explicitly deleted, you need to give a stream or file name.
detail::out_file_iterator< sequence_file_output > iterator
The iterator type of this view (an output iterator).
Definition io/sequence_file/output.hpp:110
iterator begin() noexcept
Returns an iterator to current position in the file.
Definition io/sequence_file/output.hpp:236
void size_type
The size type (void).
Definition io/sequence_file/output.hpp:106
T data(T... args)
Provides seqan3::views::elements.
Provides the seqan3::sequence_file_format_genbank class.
Provides the seqan3::format_sam.
T format(T... args)
T forward(T... args)
T get(T... args)
field
An enumerator for the fields used in file formats.
Definition record.hpp:60
Whether a type behaves like a tuple.
Provides various utility functions.
Provides exceptions used in the I/O module.
Stream concepts.
Provides various utility functions required only for output.
The main SeqAn3 namespace.
Definition aligned_sequence_concept.hpp:26
SeqAn specific customisations in the standard namespace.
Provides the seqan3::detail::out_file_iterator class template.
Provides the seqan3::record template and the seqan3::field enum.
Provides seqan3::detail::record_like.
Provides seqan3::sequence_file_output_format and auxiliary classes.
Provides seqan3::sequence_file_output_options.
T size(T... args)
A class template that holds a choice of seqan3::field.
Definition record.hpp:125
Thrown if there is an unspecified filesystem or stream error while opening, e.g. permission problem.
Definition io/exception.hpp:36
The class template that file records are based on; behaves like a std::tuple.
Definition record.hpp:190
The options type defines various option members that influence the behaviour of all or some formats.
Definition sequence_file/output_options.hpp:23
Type that contains multiple types.
Definition type_list.hpp:26
T tie(T... args)
Provides traits for seqan3::type_list.
Provides seqan3::tuple_like.
Provides seqan3::views::convert.
T visit(T... args)
Provides seqan3::views::zip.
Hide me