Changes

Unreleased

Release 2.51.0 (2024-10-25)
 - Banish more Sequels

Release 2.50.1 (2024-10-04)
 - Remove sequel specific references to Sequel from the code
 - Following a release on 07/09/2024, see https://metacpan.org/dist/App-perlbrew/changes
   the checksum of the script served by https://install.perlbrew.pl had changed.
   https://install.perlbrew.pl is a redirect to raw
   https://github.com/gugod/App-perlbrew/blob/master/perlbrew-install, so
   the change originates from GitHub and can be trusted. Our CI flow compares
   the checksum of the downloaded script to the expected value. We now store
   an updated expected checksum value, which corresponds to the latest release.

Release 2.50.0 (2024-08-07)
 - SMRT Link IsoSeq analysis has changed substantially in SMRT Link v13
   so old code and tests related to this was removed - further work is
   planned in this area to remove the Sequel subdirectory entirely. As
   it is an evolving area in sporadic use the loading code for IsoSeq
   analysis was moved to a separate new IsoSeqPublisher module.
     - added new WTSI::NPG::HTS::PacBio::Sequel::IsoSeqPublisher module
       which handles loading of analysis output from SMRT Link
       pb_segment_reads_and_isoseq jobs
     - old code for IsoSeq loading completely removed from
       WTSI::NPG::HTS::PacBio::Sequel::AnalysisPublisher

Release 2.49.2
 - Prevent access to a live mlwh database from tests.

Release 2.49.1
 - Removing Tidyp dependency from CI
 - Add iRODS 4.3.2, remove 4.3-nightly

Release 2.49.0
 - PacBio iRODS data - set QC state metadata when it is safe to do so.
 - Allow non deplexed Revio cells to publish now
 - Analysis loading changes post SMRT Link v13
   Remove very old test data and update some tests. Add new test data and minor
   code change to support loading of data from deplexing jobs in SMRT Link v13+
   (very rare as only done when deplexing on instrument is incorrect).

Release 2.48.0
 - Add sequencing_control.subreads.bam and index file to PacBio run
   archiving. These files can be used to regenerate QC metrics.
 - Add new (SMRT Link v13+) reports.zip file to PacBio Revio data 
   archiving.

Release 2.47.1

  - Amended policy on archival of index files. No longer skipping
    index files if there are no reads.

Release 2.47.0

  - Fixed some outdated POD
  - 'alt_target' iRODS metadata attribute value is now computed in the
    same way as the 'target' metadata attribute value. To be consistent
    with prior practice, the 'alt_target' value is assigned only when it
    is true. This resolves the issue of tag zero files, which are
    produced by the alternative process, previously having 'alt_target'
    metadata attribute value of 1.
  - Following introduction of ss_<STUDY_ID>_human iRODS groups to
    provide limited access to split-out human data and consolidation
    of logic about assignment of study_related iRODS groups in
    https://github.com/wtsi-npg/perl-irods-wrap/pull/297, the following
    changes were made:
      - contains_nonconsented_human and update_group_permissions methods
        were dropped from WTSI::NPG::HTS::Illumina::AlnDataObject
      - ss_<STUDY_ID>_human iRODS groups were added to test fixtures

Release 2.46.0

  - Add iRODS 4.3.1 Ubuntu 22.04 as a optional test target
  - Minor change to PacBio AnalysisPublisher
  - Update test matrix iRODS versions
      Remove the combination of 4.2.11 clients with a 4.2.7 server
      Add the combination of 4.2.7 clients with a 4.2.7 server
      Add the combination of 4.3-nightly clients with a 4.3-nightly server

Release 2.45.0

  - Add link to metadata repository to README
  - Refactor to use common iquest wrapper from perl-irods-wrap
  - Add Perl 5.34 to CI matrix
  - Fixed a bug in the PacBio metadata updater.
      Added a missing plate number when calling the find_pacbio_runs()
      method that retrieves up-to-date metadata from mlwh.
  - Refactor by hoisting environment variables
  - Install iRODS clients outside of working directory

Release 2.44.0

  - Changes to PacBio iRODS loading to support multi plate runs
    (requires npg_id_generation v4.0.0)
  - Unify location file writing for Illumina and PacBio products
  
Release 2.43.0

  - Change PacBio min deplex default as all data now HiFi
  - Remove dependency on npg_ml_warehouse for generating PacBio product IDs
  - Add Singularity container support for iRODS clients, remove Conda-based
    clients

Release 2.42.0

  - Add location file writer and implement in pacbio run/analysis publisher.
  - Change docker image repository used by CI

Release 2.41.0
  - Add --logconf option to npg_publish-tree.pl
  - Add id_product metadata to PacBio data objects.
  - Remove WTSI::NPG::HTS::PacBio::Sequel::APIClient (moved to
    wtsi-npg/perl-dnap-utilities).
  - Add sequence diagrams of metadata flow when publishing.

Release 2.40.0
  - PacBio - added support niche run folder processing special case
    to support an ongoing PhD process (not likely to be required once
    the project is finished)
  - Update action versions in CI

Release 2.39.0
  - PacBio - support iRODS loading for changed deplexing off instrument
    files & directory structure in SMRT Link v11.
  - Update to ubuntu-latest in actions

Release 2.38.1

  - Update baton version in github actions workflow to 4.0.0

Release 2.38.0

  - PacBio - add new files from new run configurations in SMRT Link v11.0
    to RunPublisher
  - Update baton version in github actions workflow to 3.3.0
  - Remove iRODS 4.2.10 from github actions workflow
  - archive metrics files produced by SamHaplotag
  - PacBio - extend runauditor to cope with fixing permissions on
    runfolder subdirectories created by deplexing on board
  - Demote error to warning when EBI and iRODS checksums are different

Release 2.37.1

  - Bugfix: Object limit for the iRODS single replica object-finding query

Release 2.37.0

  - Bugfix: iquest parsing for the iRODS single replica object-finding query
  - Change the field used to determine IsCCS for PacBio
  - Archive the substitution metrics qc JSON file for Illumina

Release 2.36.0

  - Add ML warehouse JSON file generation to npg_publish_tree.pl
  - Add standard input option for providing metadata to npg_publish_tree.pl
  - Add --limit CLI option to npg_update_single_replica_metadata.pl

  - Bug fix: Avoid creating duplicate entries in the ML warehouse JSON file
    generated by the Illumina run publisher

Release 2.35.0

  - Tweak Sequel AnalysisPublisher for SMRT Link 10.2 to allow xml
    in entry-points subdir.
  - Correct propagation of the ml warehouse db connection to st::api::lims
    object. The publisher was previously written to cache the connection
    created by LIMSFactory, but failed to do so, leading to creation of
    multiple connections. 

Release 2.34.0

  - PacBio - prep and load fasta file output from IsoSeq analysis.
  - Change baton version in actions to 3.2.0 for compatibility
  - Add Haplotagging logs to log publisher
  
Release 2.33.0

  - Tweaks to ebi meta updater to account for files with no md5 in
    the subtrack db (now common due to special projects).
  - Also avoid nulls in ebi_sub_acc field for ebi meta updates.
  - Tests simplified.

Release 2.32.0
 
  - Archive raw pulldown metrics (gatk_collecthsmetrics) txt file from
    the qc directory.
  - Disable the samtools CRAM reference cache during tests.

Release 2.31.1

  - substitution_analyis filename fix

Release 2.31.0

 - PacBio - ensure runfolders are writable for md5 files prior to iRODS
   loading. 
 - Allow another BAM file prefix in PacBio analysis loading (IsoSeq).
 - PacBio code to check and change unix permissions on runfolders.
 - GitHub Actions: verbose testing and update baton clients

Release 2.30.0

 - Add substitution_analyis and substition_metrics ancillary files.
 - Add PacBio API query (query_datasets) to support mlwarehouse loading.
 - Add some extra checks before loading PacBio secondary analysis data.
 - Update alignment file header parser for minimap2.
 - Install NPG and CPAN Perl modules separately in CI, to avoid caching
   NPG modules.
 - Move from Travis CI to GitHub Actions.

Release 2.29.0

 - Additional PacBio API queries to support mlwarehouse loading.

Release 2.28.0

 - Support loading of Pacbio Sequel OnInstrument data.

Release 2.27.0

 - set_object_permissions in WTSI::NPG::Data::ConsentWithdrawn updated to
   use <user>#<zone> rather than <user> only format 
 - Pacbio RSII specific code clear out
 - Restore the gbs plex meta data

Release 2.26.0

 - Add iRODS 4.2.8 to tests

   iRODS 4.2.8 is marked as an expected failure because we have not
   yet built Conda packages of irods-icommands or libhts plugins for
   iRODS 4.2.8.

 - Pacbio Sequel merged analysis report file changes due to version update
 - Pacbio Sequel generate and publish archive of QC images and reports.

Release 2.25.0

 - Pacbio Sequel merged analysis report file changes due to version update

Release 2.24.1

 - Order Conda channels for tests so that NPG channels have highest priority
   and conda-forge the lowest.

Release 2.24.0

 - Create merged analysis report for Sequel analysis jobs
 - Remove iRODS 4.1.12 from tests

Release 2.23.0

 - Bugfix: handling of --exclude CLI arguments by npg_publish_tree.pl.
 - Make Treepublisher create an empty target directory, even when
   there are no files to publish.

Release 2.22.0

 - New script, npg_publish_tree.pl, to publish arbitrary directory trees of
   files to iRODS, set permissions and add metadata to the destination root
   collection.
 - Add filter parameter to TreePublisher::publish_tree.
 - Add named parameters to BatchPublisher::publish_file_batch.
 - Fix teardown of iRODS collection in TreePublishTest.

Release 2.21.0

 - New script, npg_process_consent_withdrawn.pl, to withdraw permissions on
   iRODS data for samples with concent withdrawn. Ported from the internal
   SVN repository and refactored to use WTSI::NPG::iRODS class for iRODS
   operations.
 - Add testing with iRODS v4.2.7 clients to Travis CI matrix.

Release 2.20.0

 - When archiving logs skip large .err files in archive/tmp_.. directories
 - Pacbio tweak to support traction barcode identifiers and loading older analysis.

Release 2.19.0

 - PacBio run loading add options as in other scripts to support loading
   of older runs.
 - PacBio tweak to analysis loading to support traction barcode identifiers.

Release 2.18.0

 - Add support for PacBio run deletion.
 - Change ML warehouse query used to find PacBio runs in order to support
   Traction LIMS (use pac_bio_run_name).
 - Change GD package name on Travis from libgd2-xpm-dev to libgd2-dev

Release 2.17.0

 - Uniquify pac_bio_run records from mlwarehouse - to handle unexpected
   duplicate entries and stop adding target = 1 to scraps BAM files. 
   Change secondary archiving to support SMRT Link v8.
 - Use samtools executable instead of samtools_irods.
 - Travis CI tests:
     use prod WSI Conda channel,
     use specific version of samtools, pin it to v.1.10.0,
     to enable samtools' access to iRODS, install HTS lib irods pligins
     since they do not come with the samtools v.1.10.0 we have in the
     channel
 - Test change to cope with samtools v.1.10.0 inserting PG lines into
   the header.
 - WTSI::NPG::HTS::PacBio::Sequel::AnalysisPublisher restrict filename 
   checks to just the filename
 - Publish new files generated by GATK.
 - Removed ONT publishing support (including inotify and HDF5 dependencies).

Release 2.16.0

 -  WTSI::NPG::HTS::Illumina::ResultSet - added pulldown_metrics to 
    qc_regex.

Release 2.15.0

 - remove usage of private method and prep for simplified path framework
 - avoid defining own builders for runfolder accessors by inheriting from
   runfolder object

Release 2.14.0
 
 - Add support for loading PacBio ccs BAM files and setting target = 1
   on relevant PacBio sequence files.

Release 2.13.0

 - WTSI::NPG::HTS::Illumina::ResultSet - added geno to genotype_regex.
 - WTSI::NPG::HTS::Illumina::ResultSet - added quant.zip to ancillary_regex.
 - WTSI::NPG::HTS::Illumina::ResultSet - added _target_autosome.stats to 
   ancillary_regex.
 - WTSI::NPG::HTS::PacBio::Sequel::AnalysisPublisher - archive tag zero file.
 - Control caching of st::api::lims objects to reduce memory use in highly
   plexed runs. WTSI::NPG::HTS::LIMSFactory now uses Cache::LRU to limit
   the number of cached st::api::lims objects to 100.
 - Added BioNano Saphyr run publisher and run publishing script

Release 2.12.1

 - WTSI::NPG::HTS::Illumina::AlnDataObject - corrected is_paired_read
   calculation and added missing test.
 - Illumina RunPublisher modified to use a TreePublisher backed - no
   change in behaviour is expected.

Release 2.12.0

 - WTSI::NPG::HTS::LIMSFactory - for performance, cache st::api::lims
   objects.
 - Add id_product metadata attribute to primary metadata, compute the
   attribute's value as a digest of composition JSON string.
 - Metadata updater for Illumina sequencing data - search for a run
   collection in multiple locations.
 - Support for minor Pacbio API changes.

Release 2.11.0

 - Only add run-, lane- or plex-level metadata (id_run=x, lane=y,
   tag_index=z) on merged data when all the constituents are from the
   same run, lane or plex. E.g.

   run 1000, lane 1; run 1000, lane 2 gives id_run=1000

   run 1000, lane 1; run 1001, lane 1 gives lane=1
   
   run 1000, lane 1, plex 2; run 1000, lane 2, plex 2;
   run 1000, lane 3, plex 2 gives id_run=1000, tag_index=2

Release 2.10.0

 - Make the RunPublisher include paths relative to the source
   directory. This generalises the behaviour for QC files to other
   file types.
 - Bug fix: restored correct creation and consumption of restart
   files.

Release 2.9.2
 - Register RunParameters.xml file for Illumina run publishing.
 - When publishing cram files, only publish crai (not bai).

Release 2.9.1
 - Illumina::AlnDataObject
     Call super() in update_group_permissions to trigger inherited
     before and after methods which manage setting the 'public' data
     access group.

Release 2.9
 - Archive output from PacBio auto deplexing jobs.
 - Use tears -d to ensure the default iRODS server is used (avoids
   a HEIRARCHY_ERROR when using getHostForGet cross-zone).
 - Improvements and bug fixes for publishing from the GridION.

Release 2.8.1
 - Use tears 1.2.4 (-w flag now required for write to iRODS)

Release 2.8
 - Switched to disposable-irods 1.3 (iRODS packages from Sanger S3,
   replaced RENCI FTP site).
 - Added single-server option.
 - Added support for data files to change during the tar process
   and to detect and archive those changes.
 - Individually check the checksum of files local to the gridion
   where they are loaded and compare with the iRODS checksum.
 - Updated the samtools and htslib versions to 1.7.
 - Added tar file auditor.
 - Use manifest checksums to confirm file contents.
 - Support loading PacBio RSII files created using out of date reagents.
 - Allow a user-supplied checksum to be used in TarPublisher and TarStream
   (so GridIONRunPublisher can allow the checksum of uncompressed data to be used).
 - Publish fastq files during the catchup phase.
 - Allow auditing of older GridION tar files which used a relative path for
   tarred files.
 - Archiving changes for illumina sequencing genotype files.
 - Fix for the regex used when choosing to ignore files.
 - Minor fix to DataObjectFactory.pm.

Release 2.7
 - Moved quant file tests to illumina run publisher tests
 - Added compression extentions to illumina ancillary file regular expression
 - Added tag hops files to illumina run publisher tests

Release 2.6

 - Extended log publisher find command to include STAR aligner log files.
 - Add 'STAR' as a valid aligner to the header parser.
 - Make file formats 'tab' and 'zip' have restricted access.
 - Add compress suffixes to ancillary lane/plex file patterns.

Release 2.5

 - Refactored BioNano publication, with metadata from MLWH.
 - Add file type option in PacBio meta updater.
 - Added a configurable local output directory.
 - GridION: ancillary files and metadata; configurable TMPDIR.
 - Added Star's tab and Salmon's quant.zip to the list of file
   formats with restricted access.

Release 2.4

 - Added GridION publishing and primary metadata support.
 - Requires perl-irods-wrap >= 3.*

Release 2.3

 - Made tests requiring h5repack TODO until we work out why it fails
   intermittently.
 - MinIONRunPublisher
     use /tmp instead of /dev/shm.
     added boolean attributes to control compression.
 - Support streaming MinION data to iRODS;
     this adds dependencies on tears, GNU tar and h5 tools (h5repack).
 - Update as well as add PacBio legacy meta data.
 - Test against htslib 1.5, samtools 1.5.

Release 2.2

 - PacBio
    Sequel monitor to use new completedAt date.
    Add warning for Sequel R&D run entry.
    Archive Sequel adapters.fasta.
    Restrict access for relevant PacBio files.

Release 2.1

 - Added --restart-file CLI option to the Illumina run publisher to enable
   job-specific file naming.

 - Added a local cache of loaded file names. A restarted job will use
   this to determine which files remain to be loaded.

 - Added a CLI option to abort a loading run after a user-specified
   maximum number of errors.

 - Use WTSI::NPG::iRODS::Publisher, removed deprecated
   WTSI::NPG::HTS::Publisher.

 - Added monitor for PacBio Sequel and changed metadata files after v4
   upgrade.

 - Added library_name meta data for PacBio.

 - Added support for PacBio Sequel.


Release 2.0

 - API change: removed WTSI::NPG::HTS::Annotator in favour of
   WTSI::NPG::iRODS::Annotator.

 - PacBio: no longer require run ids to be specified whe updating
   metadata.

 - PacBio: remove dependancy on multi value user defined fields.
   Change run_uuid field to optional to support R&D runs/wells.

Release 1.6

 - BioNano: add command line script for publication.

 - Added a script to update PacBio metadata in iRODS.

 - Added support for PacBio legacy metadata.

 - Added PacBio run monitor CLI script.

 - Archival of bam_stats files.

Release 1.5

 - Bug fix: Propagate a failure to read a CRAM header

 - Bug fix: Failure to parse a JSON read count cache file is captured
   and added to the error count. It is no longer immediately fatal,
   but will cause a no-zero exit of the loading script.

 - Bug fix: Failure to make an MD5 cache file is reduced from a fatal
   error to a warning.

 - Warnings now come through the logger, rather than raw carping,
   so they gain a WARN tag.

 - The default log level has been lowered from ERROR to WARN.

 - Added WTSI_NPG_BUILD_BRANCH environment variable to permit overriding
   of the default build branch.

 - Added negation to CLI file category selection.

 - Added support for the samplesheet lims driver.

 - WTSI::NPG::HTS::Publisher now supports multi-value AVUs. Previously
   it retained only the last value processed for a particular attribute.

Release 1.4

 - Added library_type metadata.

 - Added tgz as recognised file suffix for metadata.

 - Improved logging; configurable per-class, reduced verbosity in metadata
   updater, increased default verbosity in scripts.

 - The metadata updater will handle an id_run of 0.

 - Added a CLI option to specify an id_run for cases where it can't be
   detected automatically.

 - Avoid loading index files for empty alignment files.

 - Avoid loading the JSON-wrapped samtools stats created in more recent
   runs.

 - Added a PacBio run publisher and monitor.

 - Added log file publisher.

 - Ensure test dependencies are installed.

Release 1.3

 - Defer checksums on (re)loading files until after upload. Assume
   that iRODS checksums are in a good state prior to upload.

 - Bug fix: avoid calling $obj->str on a string, triggered when remote
   path is a collection.

 - Count errors during group permission removal and re-throw.

 - Added strict_groups parameter to overridden update_group_permissions
   method.

 - Use the return values of metadata-setting methods to inform the caller
   of any failures while each operation remains in a try-catch.

 - Add the ability to load InterOp files.

 - Added Illumina namespace.

Release 1.2

 - Added --alt-process and --archive-path command line options to
   publish_illumina_run.

 - Added options aliases, e.g. position/lanes, to publish_illumina_run.

 - Initial seqchksum digest metadata support for publish_illumina_run.

 - Restrict the types of secondary metadata on ancillary files i.e 
   Restrict JSON file secondary metadata to study_id.

 - Change run option to id_run (or id-run).

 - Filter data objects by lane and tag index (without recourse to 
   metadata).

Release 1.1

 - Added support for alternative ML warehouse drivers.

 - The default samtools is now samtools_irods.

 - The Publisher now avoids creating MD5 cache files for small files
   and tests for stale cache files.

 - Bug fix: Corrected handling of nonconsented human.

 - Bug fix: Corrected caching of file lists in RunPublisher.

 - Bug fix: Publish run-level XML files.

 - Bug fix: Obtain num reads value from the correct flagstats JSON
   file for alignment subsets.

Release 1.0