- update Globus SDK to globus-sdk==3.62.0
- update to Python 3.11.13
- run black on all service and test files, fix most linting complaints (pylint, sonarlint)
- Added a new KBaseAuth client
- Modified the
bulk_specificationendpoint to parse DTSmanifest.jsonfiles with adtsflag. - Added a DTS file watcher that moves a user's files from the DTS dropoff point to the user's subdirectory.
- Added
AUTH_TOKEN,DTS_MANIFEST_SCHEMA, andDTS_STAGING_DIRconfig values.- These are all required.
AUTH_TOKENis a service token used by the service to verify user existence.
- Added a separate StagingServiceConfig object that manages configurations.
- Modified the
AUTH_URLconfig to point to the root of the auth service.
- Fixed a bug that would cause NaN and Inf values in xSV to be returned as JSON barewords, which could cause some JSON parsers to fail. They are now returned as strings.
- Changed the Excel parser to not consider NaN and Inf as missing values to maintain consistency with the xSV parsers
- Fixed a bug that under some circumstances could cause incomplete file metadata to be returned.
- Alter the behavior of the bulk specification file writers to return an error if the input
typesparameter is empty. - Fixed a bug in the csv/tsv bulk specification parser that would case a failure if the first header of a file had trailing separators. This occurs if a csv/tsv file is opened and saved by Excel.
- Fixed a bug in the csv/tsv bulk specification parser that would include an empty entry for each empty line in the file.
- Add
write_bulk_specificationendpoint for writing bulk specifications - Add
import_filetypesendpoint for getting datatype -> filetype -> extension mappings - Fixed a bug in the csv/tsv bulk specification parser that would throw an error on any empty lines in the file, even at the end. The parser now ignores empty lines the same way the Excel parser does.
- Fixed a bug in the
bulk_specificationendpoint where a missing header item in a*.?svfile would cause it to be replaced with a strange name. - Fixed a bug in the
bulk_specificationendpoint where a missing header item in an Excel file would cause a duplicate header error rather than a missing header error. - As part of the two fixes above, some error message text has changed due to the rewrite of the parsers.
- added the
fileskey to the returned data from thebulk_specificationendpoint.
- Update to Python 3.9
- Add
bulk_specificationendpoint for parsing import specifications
- BACKWARDS INCOMPATIBILITY: remove the unused
appskey from the importer mappings endpoint. - added a
fileinfofield to the return of the importer mappings endpoint that includes the file prefix, suffix, and file type(s), if any. - reverted change to expose dotfiles in the api by default
- attempting to upload a dotfile will now cause an error
- Added support for Genbank
*.gband*.gbffextensions - Added support for gzipped Reads, Assemblies, Genbank Files, and GFF files.
- Added new endpoint
importer-mappings/for getting a mapping of importers for file names - Updated endpoint to use GET and query_string
- Ran black
- BUGFIX: Change head/tail functionality to return 1024 chars to avoid bad cases with really large one line files
- Update FASTQ/SRA to use their own APP Uis
- Add a file name check to void user uploading files with name starting with space
- Update list endpoint so that it only hide .globus_id file by default
- Add a add-acl-concierge endpoint
- Added configs for that endpoint
- Added options to dockerfile/docker-compose
- Added capability to check 'kbase_session_backup' cookie
- Added a
add-aclandremove-aclendpoint for globus endpoint access - Change logging to STDOUT
- Added a
downloadendpoint for files