- An upstream bug fix in
httr
for progress bars means that some work-around can be removed - Some minor tweaks for the latest version of
httr
box_search
(and convenience functionsbox_search_files
,box_search_folders
,box_search_trash
) allow use of box.com's superb searching facilities to find files and folders. Various parameters can be set to search filenames, file contents, and to limit searches to certain file types. Results are summarised by default, can be coerced viaas.data.frame
. They can also be passed in place of a file_id to other boxr functions (e.g.box_dl
) to perform an operation on the first result.box_read_excel
Finally support for everyone else's favorite way to store data!box_write
A convenience function to convert R objects to common file types (e.g. .json, .csv, .tsv, .xlsx, etc.), and upload to box. Using the default write function (rio::export), the file type can be determined automatically from the filename provided.box_add_description
A simple way to add a description to a file on box.com. These are a useful way to decribe the contents of a file, and can also be used like commit messages on GitHub, to describe recent changes made.box_fresh_auth
A convenience function for users having trouble authenticating -- it will delete existing tokens (which are by default hidden files) for a fresh authentication process- S3 classes for files, folders, and object lists
- Objects of class
boxr_object_list
can be passed directly to functions in place of a file_id string - Objects of class
boxr_object_list
now have anas.data.frame
method
- Example of usage with
magrittr
pipes is added to the README - box file/folder id's are now validated locally before requests are sent
- Filenames are now validated locally, with helpful/informative error mesages
box_read
now accepts a user specified read function, which is now by defaultrio::import
box_read
will now try and do the right thing for files without an extension by considering the MIME type of the API response- S3 classes and methods have been consolidated into three basic types, and functions now use them in a consistent manner. Now documented for those interested (
?boxr_S3_classes
) - Print methods: Prettier and more informative
box_getwd
no longer logs an uninformative message- Documentation / collaboration improvements (improved function documentation, variable naming conventions, and a code of conduct)
- Improved tests
options(boxr.progress = TRUE)
is now respected consistently- Fix for spurious warnings comming from the latest version of
httr
(see jeroen/curl#30 and r-lib/httr#252) - Fix for weird reporting for certain valid API queries, which return exactly 0 results
Note: Skipped a version increment for CRAN iterations
- Namespace stuff for the latest Rbuild under Windows
Now up on CRAN:
install.packages("boxr")
- Fixed broken link for image assets in vignette
- The default OAuth2.0 token cache is now
~/.boxr-oauth
This release contains small changes to help new users get started (especially those who don't get to see the GitHub README).
- The package vignette (
vignette("boxr")
) is now more oriented towards getting new users up and running (fc931cd), providing much of the same information as at?box_auth
, but with screenshots and friendlier formatting - Upon loading boxr, users are now provided with the GitHub issues URL for bug reports. If it looks like the user hasn't used boxr before (
!file.exists("~/.boxr_oauth")
), the code to generate the Getting Started vignette is presented (e18864d) - R help 'homepage':
?boxr
now takes the user to a summary of the package, which links back to the Getting Started vignette (#31) - Continuous integration now covers Windows via appveyor (08925e5), and for useRs coming from GitHub, there's a few more visual aids in the README to help them evaluate/grok the package (1c574c2, 593101d, 670487b)
boxr v0.2.6 is on CRAN, the following will do the trick:
install.packages("boxr")
- json is now parsed consistently by
box_read()
andbox_read_json()
box_getwd()
andbox_setwd()
now write to the correct options vars (e.g., work properly again) (#27)
- General CRAN policy stuff
- Improved tests (f6cc01c, 55d6581)
- Several new wrapper functions for
box_read
have been added (#25):box_read_csv
box_read_tsv
box_read_json
box_previous_versions
allows you to query previous versions of a filebox_dl
can now download specific versions of a file, via either theversion_no
orversion_id
parameter (#21)
- URLs for the set-up process now point to https://www.box.com/developers/services, as opposed to https://www.box.com/developers/services (#20)
- Tiny changes to make the package more amenable to CRAN (not there yet!)
- Much more thorough documentation (#1, #3, #16)
- box_dir_diff now has it's own S3 class (#16)