Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
htslib: Autospec creation for update from version 1.16 to version 1.19
Andrew Whitwham (12):
Uodate copyright for winter release.
January 2023 NEWS update.
Keeping the NEWS file up-to-date.
More additions and improvements.
Added htscodecs update to v1.4.0
More space.
Switched back to openssl for Alpine.
Amalgamate multiple CIGAR ops into single entry. (#1607)
Stop the overwriting of the end value.
Summer 2023 copyright update.
Speed up removal of lines in large headers.
Winter News 2023 (PR #1703)
Bergur Ragnarsson (3):
draft fix
fix memory leak
exit early on error
David Seifert (1):
Use POSIX `grep`
Fabian Klötzl (1):
improve parsing performance
Fangrui Song (1):
Apply the packed attribute to uint*_u types for Clang
James Bonfield (86):
Update man SEE ALSO sections from .BR to .IR so the website uses URLs
bgzip text compression mode
Make the bgzip -g option less opaque.
Make tabix support CSI indices with large positions.
Prevent crash when only FASTA entry has no sequence.
Add an fai_line_length function.
Check for invalid BC tags in fastq output.
Warn if ref file is given but it doesn't contain the refs we need.
Fix buffer read-overrun in bam_plp_insertion_mod.
Fix ref fix from c91804c
Make it easier to modify shared library permissions during install
Add CRAM SQ/M5 header checking when specifying a fasta file. (PR #1522)
Speed up load_ref_portion.
Expand CRAM API a bit to cope with new samtools cram_size command. (PR #1546)
Merges neighbouring I and D ops into one op within pileup. (PR #1552)
Improve API docs for bgzf_mt
FIx a bug in the codec learning algorithm for TOKA
Fix a bug with multi-threading and embed_ref=2 on name sorted data
Use non-ref mode when all else fails for CRAM encoding
Add some documentation on cram encoder code structure
Fix Cram compression container substitution matrix generation.
Tweak the CRAM_SUBST_MATRIX table.
Prevent spurious and random system errors from test_bgzf.c
Fix cram_index_query_last function
Avoid deeply nested containment list on old CRAM indices.
Permit fastq output to create empty FASTQ records for seq "*".
Fix a couple small VCF auto-indexing bugs.
Backport attractivechaos/klib#78 to htslib.
Slightly speed up various cram decoding functions (#1580)
Remove CRAM 3.1 warning.
Trivial fix to expr, removing "^".
Add MZ:i tag as a check for base modification validity. (#1590)
Fix typo in kh_int_hash_func2 macro.
Rename aux tag MZ to MN.
Protect against overly large containers.
Don't create overly large CRAM blocks.
Add a missing break statement in cram_codec_to_id. (#1614)
Fix fd_seek on pipes on modern MinGW releases.
Change bounds checking in probaln_glocal
Fix a containment bug in cram_index_last.
Migrate base modification code out of sam.c
Correct base modification implicit / explicit status when mixed together.
Add a bam_mods_queryi interface.
Add bam_parse_basemod2 API with additional flags argument.
Add more internal sam_mods.c documentation
Update bam_next_basemod too to cope with HTS_MOD_REPORT_UNCHECKED.
Fix decompress_peek_gz to cope with files starting on empty gzip blocks.
Fix to 2e672f33 decompress_peek_gz change.
Add fai_thread_pool interface.
NEWS updates for pending release
Makes bam_parse_cigar able to modify existing BAM records rather than
Fix cut and paste errors in bam_aux2f documentation
The first stage of vcf_parse_format speed improvements.
Further VCF reading speeds optimisations.
Revert most of the vcf_parse_info improvements.
Add an hclen SAM filter function.
Fix a minor memory leak in malformed CRAM EXTERNAL blocks. (#1671)
Enable auto-vectorisation in CRAM 3.1 codecs.
Cache key header lengths.
Allow vcf_format to work on packed data, plus bcf_fmt_array improvements.
Speed up kputd by approx 130%.
Minor tweaks to bcf_fmt_array and bcf_str_missing usage.
Fix a cram decode hang from block_resize.
Always do the CRAM mutex lock/destroys.
Add C++ casts for external headers.
Enable optimisation level -O3 for SAM QUAL+33 formatting.
Avoid a NULL pointer dereference while building CRAM embedded ref.
Avoid a NULL pointer deref when erroring writing CRAM to stdout
Make CRAM internal data structures use hts_pos_t.
Prevent CRAM 3 from attempting to write out out-of-bounds ref positions
Remove memory leak when cram_encode_container fails during a close.
Fix out by one error on extend_ref memory allocation.
Don't call cram_ref_decr on consensus-based references.
Check for 64-bit values in BETA codec initialiser.
Protect against CRAM slice end going beyond end of reference.
Prevent extend_ref from making huge mallocs on very sparse data.
Improve the fuzzer to write BAM/CRAM and BCF too.
Fix memory leaks on failed CRAM encode.
Permit embed_ref=2 mode to be reenabled after using no_ref.
Tighten memory constraints for cram_decode.
Further rewrite of the fuzz test harness
Reduce maxmimum BCF header len in fuzzer.
Fix buffer read overrun in cram_encode_aux.
Disable hts_set_fai_filename call in hts_open_fuzzer.
Avoid undefined behaviour integer overflow in extend_ref
Fix integer overflow in cram_compress_block2
John Marshall (19):
Add bam_aux_first()/bam_aux_next() tagged aux field iterator API
Document that bam_aux_del()'s `s` parameter must be non-NULL (& reformat)
Add symbol versioning to the ELF shared-object file
Mention in INSTALL that using plugins may need -rdynamic
Set _XOPEN_SOURCE in configure if it's not already set
Add missing Makefile dependencies [minor]
Make last_in a pointer to const [minor]
Add "uncompressed" in hts_format_description() where appropriate
Add hclose()-doesn't-close-fd option and use it for hopen("-")
Take advantage of shared hopen("-") in htsfile.c
Explicitly fclose(stdout) in test/test_view.c too
Fix hfile_libcurl small seek bug
Apply memchr optimisation in the general delimiter case too
Add test case compiling the public headers as C++
Remove remnants of HTS_HAVE_NEON, unused since PR #1587
Remove NUMERIC_VERSION, unused since PR #1226
Document primarily MM/ML and fix base-mod-related typos
Compare as off_t (not size_t) and fix printf specifiers
Install annot-tsv.1 man page and alphabetise annot-tsv rules
Lilian Janin (2):
Fix error code 0 returned by bcftools after error
Make bcftools return an error code != 0 after [E::bgzf_read_block] Invalid BGZF header at offset xxx
Petr Danecek (15):
Remove variable redeclaration warnings from perl test script
Make bcf_hdr_seqnames() work with gapped chromosome ids
Make bcf_hdr_idinfo_exists() more robust
Check if VCF POS column could be fully parsed
Allow repeated calls of bcf_sr_set_regions (PR #1624)
An attempt to parse malformatted region such as {1:1}-2 should fail
Add new annot-tsv program
Output full diff on failing tests
Output full diff on failing tests
Make qsort in regidx order-reproducible across platforms
Provide a nill (dot) value when the field is empty
Use EXIT_SUCCESS with -h and EXIT_FAILURE on errors
Address various comments
VCF parsing fix
Allow renaming of the default -a annotations (#1709)
Rob Davies (65):
Fix n-squared complexity in sample line with many adjacent tabs
Switch to building libdeflate with cmake
Add faidx_seq_len64() and fai_adjust_region() interfaces
Rework / add new faidx tests
Fix build on ancient versions of gcc
Ensure strings in config_vars.h are escaped correctly
Switch MacOS CI tests to an ARM-based image
Cut down the number of embed_ref=2 tests that get run
Cap bgzf_getline return value to INT_MAX
Make tbx_parse1 work for lines longer than 2Gbytes
Use correct type for ret in vcf_write()
Don't error when making an iterator on a tid not in the index
Happy New Year
Catch errors from bgzf_getline() in hts_readlist, hts_readlines
Add configure (enable|disable)-versioned-symbols options and tests
Add Makefile rule to update the symbol version file
Strip out symbol versions from shlib-exports-so.txt
Update to htscodecs v1.4.0
Minor NEWS adjustment and additonal item
Switch to CURLINFO_CONTENT_LENGTH_DOWNLOAD_T for newer libcurl
Fix crypt4gh redirection
Remove use of sprintf() from HTSlib source
Make SIMD tests work when building multiarch binaries
Make MacOS tests build a multiarch version of the library
Fix bug where bin number could overflow when looking for max_off
Make reg2bins faster on whole-chromosome queries
Make reg2intervals() faster on whole-chromosome queries
Update to latest htscodecs
Don't set _POSIX_C_SOURCE for htscodecs tests
Fix trailing space in config.h made by configure
Ignore generated config_vars.h file in copyright check
Switch to `/usr/bin/env perl` for all perl scripts
Adjust comments in probaln_glocal()
Expand test-bcf-sr.c capabilities
Add synced reader region tests, and move no-index tests
Fix possible double frees in bcf_hdr_add_hrec() error handling
Prevent dangling hrec pointer after bcf_hdr_add_hrec() failure
Remove items from hdict in bcf_hdr_remove()
Ensure number of modifications is always set in bam_parse_basemod2()
Ensure simple_test_driver.sh cleans up its temporary files
Ensure base mod test result is noticed by the Makefile
Improve test/test_mod.c
Switch to htscodecs 1.5.1
Skip CRC checks when fuzzing
Prevent out-of-memory reports when fuzzing
Add missing dependency on libhts.a for hts_open_fuzzer
Fix virtual offset adjustment when indexing on-the-fly
Fix BCF/VCF on-the-fly indexing issues
Fix crypt4gh redirection
Update htscodecs to v1.5.2 (2aca18b3)
Simplify run_test function
Fix test/test.pl -F dependencies
Simplify test/test.pl
Move all annot_tsv tests into their own function
Fix up some annot-tsv error checks / reports
Reformulate man page for house style
Make compiler flag detection work with zig cc
Fix a couple of unused value warnings when built with NDEBUG
Fix @pg linking when records make a loop
Fix possible shift of negative value in cram_encode_aux()
Move typecast to the right place
Ensure no-stored-sequence reads are counted in container size
Make output on unrecognised options go to stderr
Update to latest htscodecs
Update htscodecs to v1.6.0
kojix2 (2):
Fix a typo in sam.h documentation
Fix example in docs for sam_hdr_add_line
pd3 (4):
Remove a bottleneck in VCF header processing
Add support for non-standard chromosome names containing [:-] characters
Clarify usage, include output example
Temporary workaround when excessive memory is required by FORMAT fields
vasudeva8 (10):
Adds bcf_strerror method (PR #1510)
Ensure NUL termination of Z/H data in sam_format_aux1; fix base mod state reuse
Changes to avoid segfault with uncompressed bam (PR #1632)
Demonstration of htslib/sam api usage.
formatting update
bgzf_useek fails when offset is above block limits
Add support for multiple files to bgzip
Limited usage text to 80 chars per line
man update for -a option (PR #1716)
Fix example column names and explain core column renaming
Étienne Mollier (2):
cram/cram_external.c: fix external htscodecs include
htslib-s3-plugin.7: fix whatis entry
Noteworthy changes in release 1.19 (12th December 2023)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Updates
-------
* A temporary work-around has been put in the VCF parser so that it is
less likely to fail on rows with a large number of ALT alleles,
where Number=G tags like PL can expand beyond the 2Gb limit enforced
by HTSlib. For now, where this happens the offending tag will be dropped
so the data can be processed, albeit without the likelihood data.
In future work, the library will instead convert such tags into their
local alternatives (see samtools/hts-specs#434).
(NEWS truncated at 15 lines)- Loading branch information
