In ErdaradunGaztea/woodendesc: Find Packages and Their Dependencies in Any 'R' Repository

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "90%"
)
library(woodendesc)

woodendesc

{woodendesc} strives to provide an unified API to query any R repository for the following data:

• list of available packages,
• version code of any selected package,
• direct dependencies of any selected package.

The package tries to cache as much as possible and reuse existing cache wherever possible to minimize internet usage. Many repositories (CRAN, Bioconductor, R-universe, and local libraries) have custom implementations of their handlers to make use of additional possibilities they present (e.g. previous versions on CRAN or old releases of Bioconductor).

Installation

# Install from CRAN
install.packages("woodendesc")
# Install the development version from GitHub:
# install.packages("devtools")
devtools::install_github("turtletopia/woodendesc")

How to use woodendesc?

Each repository has its dedicated functions for listing available packages, package version, and package dependencies.

A list of available packages is returned as a character vector. An example call returning a list of CRAN packages:

library(woodendesc)
# Too many packages to list them all
all_cran_pkgs <- wood_cran_packages()
print(all_cran_pkgs, max = 10)

An available package version is returned as a single code. An example call checking the current version of Biobase package on Bioconductor:

wood_bioc_version("Biobase")

Dependencies of a package are returned as a data frame with the same set of columns for all calls. An example call for locally installed digest package (dependency of woodendesc):

wood_local_dependencies("digest")

Note that some repositories store more than just one version of a package (e.g. CRAN and Github). These cases can be distinguished by pluralized _versions function name component and they return all version codes instead; a character vector of any length. This is how it works for gglgbtq on Github:

wood_github_versions("gglgbtq", "turtletopia")

If a single version code is needed, there's a counterpart that selects the latest version only. The current version of aurrera on Github is:

wood_github_latest("aurrera", "turtletopia")

Last, but not least, there are three functions for searching through multiple repositories and collecting the results. Read the documentation for details on how to specify repos parameter.

wood_packages() collects available packages from the specified repositories and returns unique elements:

wood_packages(repos = c("core", "https://colinfay.me"))

wood_versions() returns all package versions available on at least one of the specified repositories; may take more than one package:

wood_versions(
  c("aurrera", "woodendesc"),
  repos = c("local#all", "runiverse@turtletopia", "cran")
)

wood_dependencies() returns a list of package dependencies, a data frame for each package specified; the first working repository is used for each package, as the dependencies are not supposed to be summable:

wood_dependencies(
  c("versionsort", "gglgbtq"),
  repos = c("local#all", "runiverse@turtletopia", "cran")
)

And if you ever need to clear cache, simply call:

wood_clear_cache()

End notes

I believe in equal rights and treatment for everybody, regardless of their sexuality, gender identity, skin tone, nationality, and other features beyond human control. Thus, I do not allow woodendesc to be used in any project that promotes hate based on the aforementioned factors.

ErdaradunGaztea/woodendesc documentation built on Feb. 1, 2025, 2:26 a.m.