- add new function
czso_filter_catalogue()
which provides an ergonomic search of the catalogue, searching relevant fields of the catalogue for a union of the search terms. - related to above,
czso_get_catalogue()
has a newsearch_terms
parameter, which filters the catalogue inline.
- move to new CZSO API
- remove deprecated functions (all exported functions not starting with
czso_
)
- fix link with 302 code for CRAN checks
- fix documentation to avoid CRAN NOTEs
- handle invalid JSON in old CZSO API
- hard-deprecate old functions; they will be removed in the next version
- update Roxygen2 version and rebuild documentation to handle CRAN validation of HTML manual
- removed usethis dependency; messages now done by {cli}
- removed stringr dependency
- readded stringi dependency
- On MacOS Monterey, native
curl
cannot reach https://www.czso.cz/, making it impossible for R to reach it. So on Monterey machines with the defaultcurl
configuration, all functions accessing the server will fail with a message instructing the user to set an option in .Renviron to resolve this issue. This is a temporary situation until the bug is fixed in MacOS or a fix is found in R that does not require the user to set environment variables.
- remove the unnecessary {stringi} dependency
- update tests to fix CRAN failure due to unreachable provider server
- fixed minor bug in how multi-file archives are handled
NOTE: there has been an unexpected change to the National Open Data Catalogue which caused problems with the dataset IDs returned by czso_get_catalogue()
. This release provides a temporary fix based on a patch to the API kindly provided by the Czech Statistical Office.
Future releases will handle the change in a more robust way once the final form of the catalogue API is determined by the CZSO.
- fix SPARQL query to return all datasets, incl. 2 that were not showing up compared to the official data catalogue web interface
- return label, rather than IRI, for temporal resolution, in the output of
czso_get_catalogue()
- minor fixes for CRAN checks
- update SPARQL query to reflect new version of provider's data catalogue
- fix bug in
czso_get_table()
(@jlacko) - add link in README to Awesome Official statistics list
- minor README edit for CRAN resubmission
- minor README edits for CRAN resubmission
- codelists (číselníky) can now be retrieved with
czso_get_codelist()
. This includes hierarchies between codelist items, and English versions where available
- new
dest_dir
parameter inczso_get_[table|codelist]()
allows you to store downloaded data files in a custom location. This location can be set at script or user (.Rprofile) level by setting theczso.dest_dir
option. - improved UI: more informative and better formatted messages, hints and reminders about common mishaps
- documentation added in
czso_get_table()
on where to get definitions of indicators and variables used by CZSO. - more date-type columns are parsed and typed correctly
- deprecated functions no longer used internally
- encoding guess no longer breaks on Linux (@jlacko)
- functions named
get_czso_*
are now deprecated and will emit a warning if used. Useczso_*()
instead.
- Description field of DESCRIPTION now contains a link to the data provider
- documentation in all functions now provide more detail on what the functions return
czso_get_table_schema()
now throws error for non-JSON schema and includes URL in error messageczso_get_table()
now writes files it cannot read with the right extensionczso_get_dataset_doc()
returns URL only invisibly ifaction
is"open"
- add cran-comments.md
- fixed dplyr-related CHECK NOTE
- updated LICENSE for CRAN
- update URL in README
- all functions accessing CZSO data now return helpful error if the dataset cannot be found
- fixed deprecation warnings to display correct package name
- added helpful error message for no access to the internet
- spelling corrections
- better error output when CZSO server returns error
- all user-facing functions are now
czso_*
to avoid conflicts and aid discovery via auto-complete. Original functions are soft-deprecated.
- improvements to documentation
- added code of conduct and contributing guide
- new
get_dataset_doc()
function for accessing documentation - new
get_table_schema()
function for retrieving table schema - exported
get_dataset_metadata()
function for accessing detailed metadata
- get_czso_catalogue() is now much faster as it uses the open data catalogue's API instead of downloading a huge CSV list of all datasets. It is less flexible as it does not allow direct filtering.
- handle encoding of some older datasets, which may not be UTF-8
- relaxed stringi version requirement to make Win build work
- both exported functions are renamed to
get_czso_catalogue()
andget_czso_table()
to avoid clashes with other packages; original functions are soft-deprecated and will be removed in future versions.
- fix bug in
get_czso_catalogue()
where an empty tibble was returned because the source CSV was read incorrectly (due tovroom
handling of newlines inside fields)
- add per-session caching to
get_catalogue()
andget_table()
, incl. newforce_redownload
parameter
- fixed error when loading zipped files in
get_table()
- first version with functioning core workflow
- Added a
NEWS.md
file to track changes to the package.