cran_results() will now error if the package fails to retrieve results from CRAN. (#53, suggested by @eddelbuettel)cran_incoming() function does not include the packages found the archive folder by default. (#49, suggested by @krlmlr)cran_incoming() gains a new argument. The results are sorted by date (in decreasing order) by default. Using sort_by_date = FALSE to get the old behavior (#49, suggested by @krlmlr).httr2 package is now used internally to download files and data from the CRAN servers.foghorn to error. The version number for such packages is now NA instead of 0.0.0. (#45, reported by @bastistician)size column for the object returned by cran_incoming() is now of type character.winbuilder_queue() was always empty following a change to the content returned by the Win-builder FTP server (#43, reported by @bbolker).cran_incoming() and winbuilder_queue() the version numbers in the tibbles are of class package_version (suggested by @krlmlr).cran_incoming() output includes the size of the tarball archive.cran_incoming() returns a zero-row tibble instead of NULL when the inspected folder is empty.progress was not documented (and not implemented properly) for cran_results() and cran_details().n_cran_flavors reads the table on the CRAN website that lists the number of flavors, caches it, and returns this number. This is a more robust way to ensure that the number of flavors (used in the package) is always accurate. There is also the possibility of setting the number of flavors (and disabling caching) using options, see the help for the n_cran_flavors() function for more information.as.tibble with as_tibblecran_incoming() is improved (#35 by @bbolker).waiting in the CRAN submission queue wasn’t documented and could not be inspected by using the argument folders in the function cran_incoming() (#35 by @bbolker and #38, reported by @HenrikBengtsson).v.names in the function stats::reshape wasn’t fully spelled (#37 by @jennybc).cran_incoming now includes the date/time (contribution by @bbolker, #30)foghorn respects the CRAN mirror set by the users instead of using https://cran.r-project.org.foghorn will retry up to three times.foghorn would, in some cases, return that a valid package name published on CRAN did not exit (#29, @zkamvar).newbies folder found on the CRAN FTP incoming server has been added to the list of places to check packages in CRAN’s submission queue (#32).noemail folder has been removed from the CRAN FTP incoming server. cran_incoming() has been updated to reflect this change.foghorn has an hex logo. Thanks to Maëlle Salmon for feedback on initial design.Add cran_incoming() to retrieve the list of packages currently in the CRAN incoming queue. Feature requested by @krlmlr, #24.
foghorn is now compatible with R >= 3.1. Suggested by @jimhester, #26.
check_cran_results() and show_cran_results() are deprecated in favor of cran_results() and summary_cran_details() respectively.Add initial support for https://cranchecks.info as a data source (not exported/tested at this stage).
Take into consideration issues other than memtest when parsing the HTML CRAN check page.
Add argument print_ok to summary() method for cran_results that can optionally print an “all clear” message when all CRAN checks return “OK” for a package.
A progress bar is displayed when the RDS files are being downloaded from CRAN if the progress package is installed (#17).
The version number of the packages are displayed in cran_details().
All the functions return tibbles, with column names in lower case.
Fix bug #14 reported by @hadley, no issues are now represented as 0 instead of NA.
Fix bug that would display a number instead of the package name in some situations.