• Implement a cache for API calls (including file download) when dataset version is specified. The functions will reload from the cache automatically the second time. (#112, #135, by @mtmorgan)
  • Improve recommendation for rdata loading (#107, #127)
  • get_file_by_*() can now return the download URL to be used in external functions or programs, useful for large files (#128, implemented in #129 @JBGruber and @kuriwaki)
  • Removes remote resource from vignette and move them to ghactions (#131)
  • Update expired token (#123)
  • Allow for downloading from draft datasets (#115, @Danny-dK)
  • Fix add dataset bug introduced in a recent version (#116)
  • Add progress bar for all downloads (#108)
  • Minor documentation improvements (#64, #107)
  • Faster method for detecting ingest (#113) and robustness to ingested files without a metadata file due to errors (#80)
  • No longer relies on foreign (#34)

Core Features

Testing

  • Doesn’t rely on CRAN for daily tests of API functions (#96)
  • Upgrade testthat to 3.0 edition (#97)
  • Move to GitHub Actions, and away from Travis-CI. (#98) Three GitHub Actions are defined for testing:
    • thorough testing related to PRs and the main branch,
    • straight-forward tests for each push to the dev branch, and
    • straight-forward tests run daily at 3am
  • Avoids tests in R Check if the dataverse server is not available (#77)
  • Avoids potential mixup of dataset ordering within a retrieved dataverse (#83)

New Methods

  • Add new get_dataframe_*() methods (#48, #66)

Small updates

  • Export initiate_sword_dataset(). (h/t Justin de Benedictis-Kessner)
  • Pass key, server, and ... arguments to internal get_dataverse() and get_dataset() calls.
  • Tests now run with an explicit empty API key.
  • Fixed a bug in internal function parse_dataset(), related to capitalization. (#17)
  • Vignette uses ‘remotes’ package in place of the archived ‘ghit’ package. (#24 @wibeasley)
  • Updated config for Travis-CI, such as switch to xenial Ubuntu release, specify repo’s org, and specify covr parameters. (#25 @wibeasley)
  • Added a provisional add_dataset_file() function. (#10)
  • Reorganized some code.
  • Noted that user-related functions are not implemented (yet). (#1)
  • Change vignette workflow so that vignettes are pre-built. (#1)
  • Removed XML dependency, updating all code to xml2 instead.
  • Removed urltools dependency.
  • Finished the “Data Archiving” vignette. (#1)
  • Fixed some bugs in dataverse_search()
  • get_file() now unzips its results when multiple files are requested and returns them as a raw vector.
  • Finished the “Data Retrieval” vignette. (#1)
  • Document dataverse_search() in a vignette. (#1)
  • Update README.
  • Update roxygen.
  • Add print.dataverse_file() method. (#12)
  • Added a dataverse_id.character() method. (#12)
  • Fixed a bug in api_url() related to parsing of the Dataverse server URL that was affected by an API change in urltools. (h/t John Little)
  • Initial commit