## [0.4] - 2024-18-12
### Breaking changes
### Added
- Manage bigwig files as input tracks
- Optional 'category' key for extra files. Will replace the path ('extra_file_xxx') in the download page
- New option `on_job_fail` in `beauris.yml`: if set to `rerun`, will retry previously failed jobs
- Add an option called `check_files_exist` that allows loading the file without automatically verifying all paths.
- Verify that yaml files use correct extension (.yml)
- Add tags for entities and a method `get_tags()`.
- Add methods `load_organisms()` that return list of all organisms and `load_organisms_by_tags()` that returns a list of all organisms filtered by tags.
- Add a check in check_fasta to verify maximum number of sequences (to detect fragmented assemblies)
- Add a config option to disable auto snp generation for bam files in jbrowse
- Added support for the use of remote url in yml files for input data
- Added support for compressed input data as `.gz`
- Added support for checksum verification of input data with sha256
### Changed
- ogs_check: Added an error message for mRNA having some UTR without CDS subfeatures
- stricter validation of versions
- Send notification even after merge commits (to the closed MR)
- Added resiliency for drmaa runner
- Should manage entity unicity differently now (all entities must be unique, except CDS that can have the same ID in a specific gene as a discontinous feature)
- OGS_check should manage 'exotic entities' -> All non-coding entities will now be written into an optional 'exotic.gff' file. This file will only be used in jbrowse, and not for functional annotation or GNB.
- Added a git pull before commiting locked files to avoid an error in some cases
### Fixed
- Manage case where the 'Name' key is set but empty in gff files
- Set asset folder and its content to 755/644 permissions
- Fixed task .state file has duplicate files in it (#128)
- Job exit_code not collected when ci job starts after remote job ending (#148)