v0.4.1

This is a recommended update for all users. This update particularly affects users working on the test MN.

NEW

• The package now connects to test.arcticdata.io instead of dev.nceas.ucsb.edu when using the "developmoent" environment
• Added a set_file_name helper function that automates the process of changing the fileName property on an object.
• Added a get_all_versions function which gets all the versions of a particular object
• get_package now can get the parent package as well if the optional argument parente is set to TRUE

CHANGED

• set_access_and_rights now defaults to setting read, write, and changePermission permissions so there's no longer any need to set that argument manually if you want the default set of permissions.
• publish_object can now guess the format ID of the object you're publishing

FIXED

• Updated documentation in a number of places around the package
• Updates via publish_update no longer overwrite the fileName property with "science_metadata" and instead use the existing value

v0.4

This release contains a number of new, high-level functions and is a recommended update for all users.

BREAKING

• The arguments to publish_object and publish_update have been changed.

NEW

• The package checks its version against the latest GitHub release and notifies the user if their version is outdated.
• Added get_package function which returns a structured list of PIDs describing a Data Package (metadata, data, resource map, etc). This should be very helpful, though it should be used with caution because it uses the Solr index to get the information which isn't exactly the same thing as a Data Package. The better way to do this would be to parse the RDF/XML of the Resource Map, which I have yet to write up a method to do.
• Added set_access function which allows setting a set of access policy statements.
• Started adding help functions for filling in DataONE format IDs. New functions include format_iso and format_eml which return the DataONE format ID string. These are useful because I can never remember the full format string and, by making them functions, I can use autocomplete to fill them in.
• publish_update and update_resource_map now have an optional argument check_first which enabled/disables checking for the existence of all PIDs passed in arguments prior to performing the update. This speeds things up when the user is sure the PIDs exist.
• Added a few helper functions to the EML module: eml_creator, eml_contact, eml_individual_name, eml_project
• Acquiring the MNode when running env_load can now be skipped with argument skip_mn=FALSE
• Added a new high-level function for inserting a package from files/folders on disk: create_from_folder

CHANGED

• The README now contains mini-vignettes which describe a number of common use cases
• update_rights_holder is now set_rights_holder
• Moved various functions related to modifying System Metadata out into their own file, sysmeta.R
  FIXED
• Exported a number of functions I forgot to export last time around.
• Cleaned up some places where cat was used and replaced them with log_message.

REMOVED

• get_related_pids is now deprecated in favor of get_package (See NEW)

v0.3

Small release to deal with some critical stuff. Also includes some new features.

NEW

• You can now create an EML creator instance with the creator() function
• You can now pass a vector of PIDs to get_related_pids
• You can now create an EML methods section with the package

FIXED

• publish_update was fixed to correctly fill in EML otherEntity elements. There was a bug with providing online distribution URLs using PIDs rather than doc IDs
• get_related_pids now uses dataone::query so it now properly passes in an auth token which helps a lot!

v0.2

This release cleans up the package a lot to where is almost passes R CMD CHECK muster. It adds some crucial features to methods in editing.R so it is a recommended update. The workflow for editing existing packages is much better overall.

General changes:

• Dependencies dataone and datapack now reference the CRAN release of both packages (woohoo!) rathern than the HEAD of the respective repos on GitHub.
  • Drafted Roxygen docstrings for every function, however minimal
  • Functions in editing.R now return lists with new PIDs and these lists follow a similar structure.
  • Added a new set of helper functions in eml.R which handles creating new EML otherEntity subtrees for data files.
  • env_load is totally rewritten, see API changes

API changes:

• The pid argument to update_rights_holder is now vectorized
  • publish_object now checks that the arguments you pass into it exist prior to doing anything. Note: This means it takes longer to run.
  • publish_object now sets the eml 'system' attribute to 'https://arcticdata.io
  • publish_object now adds EML otherEntity subtrees for each element of data_pids you pass. It overwrites anything already set on the old EML record.
  • env_load now takes as its first argument the name of the environment you want to load (default is development). This simplifies getting an environment and MNode connection set up greatly.

Bug fixes:

• publish_object now sets rights holder correctly when doing its work. Previously it was leaving some objects with the wrong rights holder.

Miscellaneous changes (non-breaking):

• Removed old (never used) Redis module

Initial release

v0.1

Move update_rights_holder to access.R