Changelog

Table of Contents

• 3.0.1
  • New features
  • API changes
  • Notable Bug-fixes
• 3.0.0 ("Escala")

This changelog contains a top-level entry for each release with sections on new features, API changes and notable bug-fixes (not all bug-fixes will be listed).

Get to know SeqAn3 with our tutorials.

Please see the release announcement: https://www.seqan.de/announcing-seqan3/

See the porting guide for some help on porting: http://docs.seqan.de/seqan/3-master-user/howto_porting.html

See the documentation on API stability to learn about when API changes are allowed.

3.0.1

Note that 3.1.0 will be the first API stable release and interfaces in this release might still change.

New features

Alphabet

• Added seqan3::semialphabet_any, a semi-alphabet that type erases all other semi-alphabets of the same size (#981).
• Added seqan3::dna3bs, an alphabet that mimics a bisulfite-treated dna4 sequence (#1191).

Alignment

• The score type used in the alignment score matrix and the result type is configurable through a template argument of the seqan3::align_cfg::result configuration (#1340).
• The functionseqan3::align_pairwise can be parallelised using theseqan3::align_cfg::parallel configuration (#1379, #1444).

Argument parser

• Simplified reading file extensions from formatted files with theseqan3::input_file_validator and seqan3::output_file_validator (#863).
• The seqan3::value_list_validator is now constructible from a range or a parameter pack (#1298).
• Enable subcommand argument parsing, see How-to for an example (#1185).
• The seqan3::argument_parser::add_option (and add_positional_option) calls allow enum types when using the seqan3::enumeration_names customisation point (#1196).

Build system

• find_package(SeqAn3) is now case-insensitive and always populates SEQAN3_* variables in all upper-case (#1427).

Core

• Added seqan3::lzcnt, seqan3::tzcnt, and seqan3::popcount for bit manipulation (#1141).
• Added traits for "metaprogramming" with seqan3::type_list and type packs (#1204, #1214, #1273).
• Added SIMD functions seqan3::upcast and seqan3::upcast_signed (#1190).

I/O

• We increased our input performance using a faster iterator on the stream buffer (#1030).
• Support of padded alignments in the SAM/BAM format was added (#1173).
• Reading seqan3::field::cigar into a vector over seqan3::cigar is supported via seqan3::alignment_file_input (#1192).
• Writing seqan3::field::cigar into a vector over seqan3::cigar is supported via seqan3::alignment_file_output (#1192).
• Asynchronous input (background file reading) supported via seqan3::view::async_input_buffer (#1205).

Range

• Added seqan3::views::kmer_hash, a view that computes hash values of an alphabet sequence given a seqan3::shape (#946).
• Added seqan3::views::to, a view that returns a container created from a range by copying all elements (#1033).
• Added seqan3::dynamic_bitset, a container that stores single bits and has a dynamic size (#1153).
• Added seqan3::views::translate_join, analogue to seqan3::views::translate but returns a flattened range (#1171).
• Added seqan3::views::to_simd, a view that transforms a range of ranges into chunks of seqan3::simd vectors (#1190).
• Added seqan3::views::as_const, a view that provides only const & to elements of the underlying range (#1410).
• Added seqan3::views::move, a view that turns lvalue-references into rvalue-references (#1410).
• Renamed seqan3::views::all to seqan3::views::type_reduce (#1410).

Search

• The memory footprint of FM-indices over text collections was reduced (#1363).

Std

• We provide a std::to_chars overload for floating point data types in our seqan3/std/from_chars header (#1160).

API changes

• The required version of the ranges-v3 library has increased: We now support the versions >= 0.10.0 and < 0.11.0, increasing the previous requirement of >= 0.5.0 and < 0.6.0 (#1471).
• Customising for third party types has changes slightly: You are only affected if you added types to seqan3::custom::. Please see About Customisation (#1225).
• All our concepts are named in the snake_case style (e.g. seqan3::WritableAlphabet -> seqan3::writable_alphabet)! This change was motivated by the decision of the ISO C++ committee to also use snake case everywhere (#1235).

Alphabet

• The seqan3::cigar alphabet is not an seqan3::alphabet anymore but only a seqan3::semialphabet (#1285).

Argument parser

• The seqan3::value_list_validator is not constructible from a std::initialiser_list any more (e.g. seqan3::value_list_validator{{1,2,3}} does not work, use seqan3::value_list_validator{1,2,3} instead) (#1298).
• Changed class signature of input/output file validators: Most user code will be unaffected; to fix possible compiler errors you need to add an empty template parameter list to the respective instances (e.g. change input_file_validator to input_file_validator<>) (#863).
• The member type that denotes which arguments a validator can validate has been renamed from value_type to option_value_type (#1394).

Build system

• find_package accepts minimum versions (e.g. find_package(SEQAN3 3.0.1) requires at least SeqAn3 with a version of >= 3.0.1 and < 4.0.0) (#1425).
• The variable SEQAN3_VERSION_STRING defined by find_package(SEQAN3) was renamed to SEQAN3_VERSION (#1425).

Core

• The type_list header has moved: If you included <seqan3/core/type_list.hpp> you need to change the path to <seqan3/core/type_list/type_list.hpp> (#1204).

I/O

• Removed the field-based in- and output interface for sequence and structure files through std::get and std::tie: Output can instead be achieved with seqan3::views:zip(), for input we will implement unzip() in the future (#1398 #1412).
• The seqan3::field::flag of SAM/BAM input and output is now an enum instead of an integer, see seqan3::sam_flag (#1390).
• Uppercase seqan3::field names are deprecated. Use the lower case field names instead. You can easily find and replace all occurrences by the following regex: find field::([A-Z_]+) replace field::\L$1 (#1421).
• Removed the char type from the input and output files: Most user code will be unaffected; however, if you have fully specified all templates of any of the input or output files in your code, you need to remove the template parameter to select the char type of the stream, e.g. change seqan3::sequence_file_input<traits_t, fields_t, formats_t, char> to seqan3::sequence_file_input<traits_t, fields_t, formats_t>. Before this change, setting the char type gave the impression that also streams over wide characters are supported which is not the case yet (#1400).

Range

• The seqan3::concatenated_sequences::data() function has been deprecated: Use seqan3::concatenated_sequences::raw_data() instead (#1208).
• seqan3::to_char must always return a built-in character type (#1285).
• seqan3/range/view has be renamed to seqan3/range/views (#1251).
• namespace seqan3::view has been renamed to seqan3::views (#1251).

Search

• Changed class signature of (bi_)fm_index: All code that relies on automatic template deduction will be unaffected. In case you specified the template parameters of a seqan3::fm_index or seqan3::bi_fm_index you will need to add the alphabet type as first parameter and pass a seqan3::text_layout instead of a bool to indicate the text layout (single, collection). For example, fm_index<false> index{text} where text is of type dna4_vector needs to be changed to fm_index<dna4, text_layout::single> index{text} (#1222).
• The construct() method of the (bi_)fm_index is now private: Use the constructor seqan3::fm_index::fm_index(text_t && text) or seqan3::bi_fm_index::bi_fm_index(text_t && text) instead (#1222).
• The seqan3::fm_index::char_type member was renamed to seqan3::fm_index::alphabet_type The same applies for the seqan3::bi_fm_index (#1433).
• The seqan3::fm_index_cursor::index_char_type member was renamed to seqan3::fm_index_cursor::index_alphabet_type The same applies for the seqan3::bi_fm_index_cursor (#1433).

Notable Bug-fixes

• All our headers are self contained (#1085).
• The alignment algorithm with edit distance returns the correct back coordinate (#1093).
• Inserting or deleting gaps into an empty seqan3::gap_decorator does not cause assert anymore (#1109).
• Some fixes to edge cases in BAM file writing (#1110).
• The application name of the seqan3::argument_parser is restricted to alpha-numeric characters and _ and - (#1133).
• Copying and moving the seqan3::fm_index and seqan3::bi_fm_index now work properly (#1144).
• Searching in the seqan3::fm_index and seqan3::bi_fm_index constructed from a text collection containing a single text now returns the correct result (#1316).
• The view seqan3::views::take is sized if the underlying range is sized (#1146).
• The detection of the pthread library works correctly on linux based systems (#1200).
• The translation table for nucleotide to amino acid translation was corrected (#1485).
• The amino acid score matrices were corrected (#1455).

3.0.0 ("Escala")

This is the initial release of SeqAn3. It is an entirely new library so there is no changelog that covers the differences to SeqAn2.

Note that 3.1.0 will be the first API stable release and interfaces in this release might still change.