Type:	Package
Title:	Installing and Loading R Packages for Reproducible Workflows
Description:	A single key function, 'Require' that makes rerun-tolerant versions of 'install.packages' and 'require' for CRAN packages, packages no longer on CRAN (i.e., archived), specific versions of packages, and GitHub packages. This approach is developed to create reproducible workflows that are flexible and fast enough to use while in development stages, while able to build snapshots once a stable package collection is found. As with other functions in a reproducible workflow, this package emphasizes functions that return the same result whether it is the first or subsequent times running the function, with subsequent times being sufficiently fast that they can be run every time without undue waiting burden on the user or developer.
URL:	https://Require.predictiveecology.org, https://github.com/PredictiveEcology/Require
Date:	2024-08-06
Version:	1.0.1
Depends:	R (≥ 4.0)
Imports:	data.table (≥ 1.10.4), methods, sys, tools, utils
Suggests:	covr, curl, diffobj, fpCompare, gitcreds, httr, pak, parallel, rematch2, rmarkdown, knitr, rlang, roxygen2, rprojroot, testthat (≥ 3.0.0), tibble, waldo, withr
Encoding:	UTF-8
Language:	en-CA
License:	GPL-3
VignetteBuilder:	knitr, rmarkdown
BugReports:	https://github.com/PredictiveEcology/Require/issues
ByteCompile:	yes
RoxygenNote:	7.3.2
Collate:	'CRAN.R' 'Require-helpers.R' 'Require-package.R' 'messages.R' 'Require2.R' 'RequireOptions.R' 'envs.R' 'extract.R' 'helpers.R' 'pak.R' 'pkgDep.R' 'pkgDep3.R' 'pkgSnapshot.R' 'setLibPaths.R' 'setup.R' 'zzz.R'
Config/testthat/edition:	3
NeedsCompilation:	no
Packaged:	2024-08-17 15:33:45 UTC; emcintir
Author:	Eliot J B McIntire [aut, cre], Alex M Chubaty [ctb], Her Majesty the Queen in Right of Canada, as represented by the Minister of Natural Resources Canada [cph]
Maintainer:	Eliot J B McIntire <eliot.mcintire@canada.ca>
Repository:	CRAN
Date/Publication:	2024-08-17 22:20:09 UTC

Require: Installing and Loading R Packages for Reproducible Workflows

Description

A single key function, 'Require' that makes rerun-tolerant versions of 'install.packages' and 'require' for CRAN packages, packages no longer on CRAN (i.e., archived), specific versions of packages, and GitHub packages. This approach is developed to create reproducible workflows that are flexible and fast enough to use while in development stages, while able to build snapshots once a stable package collection is found. As with other functions in a reproducible workflow, this package emphasizes functions that return the same result whether it is the first or subsequent times running the function, with subsequent times being sufficiently fast that they can be run every time without undue waiting burden on the user or developer.

This is an "all in one" function that will run install.packages for CRAN and GitHub https://github.com/ packages and will install specific versions of each package if versions are specified either via an (in)equality (e.g., "glue (>=1.6.2)" or "glue (==1.6.2)" for an exact version) or with a packageVersionFile. If require = TRUE, the default, the function will then run require on all named packages that satisfy their version requirements. If packages are already installed (packages supplied), and their optional version numbers are satisfied, then the "install" component will be skipped.

Usage

Require(
  packages,
  packageVersionFile,
  libPaths,
  install_githubArgs = list(),
  install.packagesArgs = list(INSTALL_opts = "--no-multiarch"),
  standAlone = getOption("Require.standAlone", FALSE),
  install = getOption("Require.install", TRUE),
  require = getOption("Require.require", TRUE),
  repos = getOption("repos"),
  purge = getOption("Require.purge", FALSE),
  verbose = getOption("Require.verbose", FALSE),
  type = getOption("pkgType"),
  upgrade = FALSE,
  returnDetails = FALSE,
  ...
)

Install(
  packages,
  packageVersionFile,
  libPaths,
  install_githubArgs = list(),
  install.packagesArgs = list(INSTALL_opts = "--no-multiarch"),
  standAlone = getOption("Require.standAlone", FALSE),
  install = TRUE,
  repos = getOption("repos"),
  purge = getOption("Require.purge", FALSE),
  verbose = getOption("Require.verbose", FALSE),
  type = getOption("pkgType"),
  upgrade = FALSE,
  ...
)

Arguments

Details

Install is the same as Require(..., require = FALSE), for convenience.

Value

Require is intended to replace base::require, thus it returns a logical, named vector indicating whether the named packages have been loaded. Because Require also has the ability to install packages, a return value of FALSE does not mean that it did not install correctly; rather, it means it did not attach with require, which could be because it did not install correctly, or also because e.g., require = FALSE.

standAlone will either put the Required packages and their dependencies all within the libPaths (if TRUE) or if FALSE will only install packages and their dependencies that are otherwise not installed in .libPaths()[1], i.e., the current active R package directory. Any packages or dependencies that are not yet installed will be installed in libPaths.

GitHub Package

Follows remotes::install_github standard. As with remotes::install_github, it is not possible to specify a past version of a GitHub package unless that version is a tag or the user passes the SHA that had that package version. Similarly, if a developer does a local install e.g., via pkgload::install, of an active project, this package will not be able know of the GitHub state, and thus pkgSnapshot will not be able to recover this state as there is no SHA associated with a local installation. Use Require (or remotes::install_github) to create a record of the GitHub state.

Package Snapshots

To build a snapshot of the desired packages and their versions, first run Require with all packages, then pkgSnapshot. If a libPaths is used, it must be used in both functions.

Mutual Dependencies

This function works best if all required packages are called within one Require call, as all dependencies can be identified together, and all package versions will be addressed (if there are no conflicts), allowing a call to pkgSnapshot() to take a snapshot or "record" of the current collection of packages and versions.

Local Cache of Packages

When installing new packages, Require will put all source and binary files in an R-version specific subfolder of getOption("Require.cachePkgDir") whose default is RPackageCache(), meaning cache packages locally in a project-independent location, and will reuse them if needed. To turn off this feature, set options("Require.cachePkgDir" = FALSE).

Note

For advanced use and diagnosis, the user can set verbose = TRUE or 1 or 2 (or via options("Require.verbose")). This will attach an attribute attr(obj, "Require") to the output of this function.

Author(s)

Maintainer: Eliot J B McIntire eliot.mcintire@canada.ca (ORCID)

Other contributors:

• Alex M Chubaty achubaty@for-cast.ca (ORCID) [contributor]

• Her Majesty the Queen in Right of Canada, as represented by the Minister of Natural Resources Canada [copyright holder]

See Also

Useful links:

• https://Require.predictiveecology.org

• https://github.com/PredictiveEcology/Require

• Report bugs at https://github.com/PredictiveEcology/Require/issues

Examples

## Not run: 
opts <- Require:::.setupExample()

library(Require)
getCRANrepos(ind = 1)
Require("utils") # analogous to require(stats), but it checks for
#   pkg dependencies, and installs them, if missing

# unquoted version
Require(c(tools, utils))

if (Require:::.runLongExamples()) {
  # Install in a new local library (libPaths)
  tempPkgFolder <- file.path(tempdir(), "Require/Packages")
  # use standAlone, means it will put it in libPaths, even if it already exists
  #   in another local library (e.g., personal library)
  Install("crayon", libPaths = tempPkgFolder, standAlone = TRUE)

  # Mutual dependencies, only installs once -- e.g., cli
  tempPkgFolder <- file.path(tempdir(), "Require/Packages")
  Install(c("cli", "R6"), libPaths = tempPkgFolder, standAlone = TRUE)

  # Mutual dependencies, only installs once -- e.g., rlang
  tempPkgFolder <- file.path(tempdir(), "Require/Packages")
  Install(c("rlang", "ellipsis"), libPaths = tempPkgFolder, standAlone = TRUE)

  #####################################################################################
  # Isolated projects -- Use a project folder and pass to libPaths or set .libPaths() #
  #####################################################################################
  # GitHub packages
  if (requireNamespace("gitcreds", quietly = TRUE)) {
    #if (is(try(gitcreds::gitcreds_get(), silent = TRUE), "gitcreds")) {
      ProjectPackageFolder <- file.path(tempdir(), "Require/ProjectA")
      if (requireNamespace("curl")) {
        Require("PredictiveEcology/fpCompare@development",
          libPaths = ProjectPackageFolder,
        )
      }

      # No install because it is there already
      Install("PredictiveEcology/fpCompare@development",
        libPaths = ProjectPackageFolder,
      ) # the latest version on GitHub

      ############################################################################
      # Mixing and matching GitHub, CRAN, with and without version numbering
      ############################################################################
      pkgs <- c(
        "remotes (<=2.4.1)", # old version
        "digest (>= 0.6.28)", # recent version
        "PredictiveEcology/fpCompare@a0260b8476b06628bba0ae73af3430cce9620ca0" # exact version
      )
      Require::Require(pkgs, libPaths = ProjectPackageFolder)
    #}
  }
  Require:::.cleanup(opts)
}

## End(Not run)

GITHUB_PAT-aware and main-master-aware download from GitHub

Description

Equivalent to utils::download.file, but taking the GITHUB_PAT environment variable and using it to access the Github url.

Usage

.downloadFileMasterMainAuth(
  url,
  destfile,
  need = "HEAD",
  verbose = getOption("Require.verbose"),
  verboseLevel = 2
)

Arguments

Value

This is called for its side effect, namely, the same as utils::download.file, but using a GITHUB_PAT, it if is in the environment, and trying both master and main if the actual url specifies either master or main and it does not exist.

Partial alternative (faster) to installed.packages

Description

This reads the DESCRIPTION files only, so can only access fields that are available in the DESCRIPTION file. This is different than installed.packages which has many other fields, like "Built", "NeedsCompilation" etc. If those fields are needed, then this function will return an empty column in the returned character matrix.

Usage

.installed.pkgs(
  lib.loc = .libPaths(),
  which = c("Depends", "Imports", "LinkingTo"),
  other = NULL,
  purge = getOption("Require.purge", FALSE),
  packages = NULL,
  collapse = FALSE
)

Arguments

GitHub package tools

Description

A series of helpers to access and deal with GitHub packages

Usage

DESCRIPTIONFileVersionV(file, purge = getOption("Require.purge", FALSE))

DESCRIPTIONFileOtherV(file, other = "RemoteSha")

dlGitHubDESCRIPTION(
  pkg,
  purge = getOption("Require.purge", FALSE),
  verbose = getOption("Require.verbose")
)

Arguments

Details

dlGitHubDESCRIPTION retrieves the DESCRIPTION file from GitHub.com

This environment variable "R_TESTS" is set during testing, and it points to a file called Startup.Rs that is placed in the .libPaths(). If the .libPaths() is changed during the testing, then that file will not be found, and install.packages will fail to install a package with an error of source file not found. See: https://github.com/HenrikBengtsson/startup/issues/19.

Description

The environment variable is set here: ⁠https://github.com/wch/r-source/blob/8b6429feb661b02e2b2b6df1757b31cf1250a33e/src/library/tools/R/testing.R#L472-Lundefined⁠

Usage

R_TESTSomit()

Require options

Description

These provide top-level, powerful settings for a comprehensive reproducible workflow. See Details below.

Usage

RequireOptions()

getRequireOptions()

Details

RequireOptions()

prints the default values of package options set at startup, which may have been changed (e.g., by the user) during the current session.

getRequireOptions()

prints the current values of package options.

Below are options that can be set with options("Require.xxx" = newValue), where xxx is one of the values below, and newValue is a new value to give the option. Sometimes these options can be placed in the user's .Rprofile file so they persist between sessions.

The following options are likely of interest to most users:

install

Default: TRUE. This is the default argument to Require, but does not affect Install. If this is FALSE, then no installations will be attempted, and missing packages will result in an error.

RPackageCache

Default: cacheGetOptionCachePkgDir(), which must be either a path or a logical. To turn off package caching, set this to FALSE. This can be set using an environment variable e.g. Sys.setenv(R_REQUIRE_PKG_CACHE = "somePath"), or Sys.setenv(R_REQUIRE_PKG_CACHE = "TRUE"); if that is not set, then an either a path or logical option (options(Require.cachePkgDir = "somePath") or options(Require.cachePkgDir = TRUE)). If TRUE, the default folder location cachePkgDir() will be used. If this is TRUE or a path is provided, then binary and source packages will be cached here. Subsequent downloads of same package will use local copy. Default is to have packages not be cached locally so each install of the same version will be from the original source, e.g., CRAN, GitHub.

otherPkgs

Default: A character vector of packages that are generally more successful if installed from Source on Unix-alikes. Since there are repositories that offer binary packages builds for Linux (e.g., RStudio Package Manager), the vector of package names indicated here will default to a standard CRAN repository, forcing a source install. See also spatialPkgs option, which does the same for spatial packages.

purge

Default: FALSE. If set to (almost) all internal caches used by Require will be deleted and rebuilt. This should not generally be necessary as it will automatically be deleted after (by default) 1 hour (set via R_AVAILABLE_PACKAGES_CACHE_CONTROL_MAX_AGE environment variable in seconds)

spatialPkgs

Default: A character vector of packages that are generally more successful if installed from Source on Unix-alikes. Since there are repositories that offer binary packages builds for Linux (e.g., RStudio Package Manager), the vector of package names indicated here will default to a standard CRAN repository, forcing a source install. See also otherPkgs option, which does the same for non-spatial packages.

useCranCache

Default: FALSE. A user can optionally use the locally cached packages that are available due to a user's use of the crancache package.

verbose

Default: 1. See ?Require.

Create a custom "available.packages" object

Description

This is the mechanism by which install.packages determines which packages should be installed from where. With this override, we can indicate arbitrary repos, Package, File for each individual package.

Usage

availablePackagesOverride(
  toInstall,
  repos,
  purge,
  type = getOption("pkgType"),
  verbose = getOption("Require.verbose")
)

Arguments

Needs VersionOnRepos, versionSpec and inequality columns

Description

Needs VersionOnRepos, versionSpec and inequality columns

Usage

availableVersionOK(pkgDT)

Arguments

Clear Require Cache elements

Description

Clear Require Cache elements

Usage

cacheClearPackages(
  packages,
  ask = interactive(),
  Rversion = versionMajorMinor(),
  clearCranCache = FALSE,
  verbose = getOption("Require.verbose")
)

clearRequirePackageCache(
  packages,
  ask = interactive(),
  Rversion = versionMajorMinor(),
  clearCranCache = FALSE,
  verbose = getOption("Require.verbose")
)

Arguments

The default cache directory for Require Cache

Description

A wrapper around tools::R_user_dir("Require", which = "cache") that creates the directory, if it does not exist.

Usage

cacheDefaultDir()

Value

The default cache directory

Path to (package) cache directory

Description

Sets (if create = TRUE) or gets the cache directory associated with the Require package.

Usage

cacheDir(create, verbose = getOption("Require.verbose"))

cachePkgDir(create)

Arguments

Details

To set a different directory than the default, set the system variable: R_USER_CACHE_DIR = "somePath" and/or R_REQUIRE_PKG_CACHE = "somePath" e.g., in .Renviron file or Sys.setenv(). See Note below.

Value

If !is.null(cacheGetOptionCachePkgDir()), i.e., a cache path exists, the cache directory will be created, with a README placed in the folder. Otherwise, this function will just return the path of what the cache directory would be.

Note

Currently, there are 2 different Cache directories used by Require: cacheDir and cachePkgDir. The cachePkgDir is intended to be a sub-directory of the cacheDir. If you set Sys.setenv("R_USER_CACHE_DIR" = "somedir"), then both the package cache and cache dirs will be set, with the package cache a sub-directory. You can, however, set them independently, if you set "R_USER_CACHE_DIR" and "R_REQUIRE_PKG_CACHE" environment variable. The package cache can also be set with options("Require.cachePkgDir" = "somedir").

Get the option for Require.cachePkgDir

Description

First checks if an environment variable Require.cachePkgDir is set and defines a path. If not set, checks whether the options("Require.cachePkgDir") is set. If a character string, then it returns that. If TRUE, then use cachePkgDir(). If FALSE then returns NULL.

Usage

cacheGetOptionCachePkgDir()

Purge everything in the Require cache

Description

Require uses caches for local Package saving, local caches of available.packages, local caches of GitHub (e.g., "DESCRIPTION") files, and some function calls that are cached. This function clears all of them.

Usage

cachePurge(packages = FALSE, repos = getOption("repos"))

purgeCache(packages = FALSE, repos = getOption("repos"))

Arguments

Value

Run for its side effect, namely, all cached objects are removed.

Creates the directories, and adds version number

Description

Creates the directories, and adds version number

Usage

checkLibPaths(libPaths, ifMissing, exact = FALSE, ...)

Arguments

Check directory path

Description

Checks the specified path to a directory for formatting consistencies, such as trailing slashes, etc.

Usage

checkPath(path, create)

## S4 method for signature 'character,logical'
checkPath(path, create)

## S4 method for signature 'character,missing'
checkPath(path)

## S4 method for signature 'NULL,ANY'
checkPath(path)

## S4 method for signature 'missing,ANY'
checkPath()

Arguments

Value

Character string denoting the cleaned up filepath.

Note

This will not work for paths to files. To check for existence of files, use file.exists(). To normalize a path to a file, use normPath() or normalizePath().

See Also

file.exists(), dir.create().

Examples

## normalize file paths
paths <- list("./aaa/zzz",
              "./aaa/zzz/",
              ".//aaa//zzz",
              ".//aaa//zzz/",
              ".\\\\aaa\\\\zzz",
              ".\\\\aaa\\\\zzz\\\\",
              file.path(".", "aaa", "zzz"))

checked <- normPath(paths)
length(unique(checked)) ## 1; all of the above are equivalent

## check to see if a path exists
tmpdir <- file.path(tempdir(), "example_checkPath")

dir.exists(tmpdir) ## FALSE
tryCatch(checkPath(tmpdir, create = FALSE), error = function(e) FALSE) ## FALSE

checkPath(tmpdir, create = TRUE)
dir.exists(tmpdir) ## TRUE

unlink(tmpdir, recursive = TRUE) # clean up

Pass through function for chooseCRANmirror

Description

This is here to allow mocking during unit testing related to chooseCRANmirror.

Usage

chooseCRANmirror2(...)

Arguments

Compare package versions

Description

Alternative to utils::compareVersion that is vectorized on version, versionSpec and/or inequality. This will also return an NA element in the returned vector if one of the arguments has NA for that element.

Usage

compareVersion2(version, versionSpec, inequality)

Arguments

Value

a logical vector of the length of the longest of the 3 arguments.

Only checks for deprecated libPath argument (singular)

Description

Only checks for deprecated libPath argument (singular)

Usage

dealWithMissingLibPaths(
  libPaths,
  standAlone = getOption("Require.standAlone", FALSE),
  ...
)

Arguments

Detach and unload all packages

Description

This uses pkgDepTopoSort internally so that the package dependency tree is determined, and then packages are unloaded in the reverse order. Some packages don't unload successfully for a variety of reasons. Several known packages that have this problem are identified internally and not unloaded. Currently, these are glue, rlang, ps, ellipsis, and, processx.

Usage

detachAll(
  pkgs,
  dontTry = NULL,
  doSort = TRUE,
  verbose = getOption("Require.verbose")
)

Arguments

Value

A numeric named vector, with names of the packages that were attempted. 2 means the package was successfully unloaded, 1 it was tried, but failed, 3 it was not loaded, so was not unloaded.

Available and archived versions

Description

These are wrappers around available.packages and also get the archived versions available on CRAN.

Usage

dlArchiveVersionsAvailable(
  package,
  repos = getOption("repos"),
  verbose = getOption("Require.verbose")
)

available.packagesCached(
  repos,
  purge,
  verbose = getOption("Require.verbose"),
  returnDataTable = TRUE,
  type
)

Arguments

Details

dlArchiveVersionsAvailable searches CRAN Archives for available versions. It has been borrowed from a sub-set of the code in a non-exported function: remotes:::download_version_url

Deals with missing libPaths arg, and takes first

Description

Deals with missing libPaths arg, and takes first

Usage

doLibPaths(libPaths, standAlone = FALSE)

Arguments

1st level –> create the .pkgEnv object in Require

Description

1st level –> create the .pkgEnv object in Require

Usage

envPkgCreate(parentEnv = asNamespace("Require"))

Arguments

3rd level for DESCRIPTIONFile

Description

3rd level for DESCRIPTIONFile

Usage

envPkgDepDESCFileCreate()

3rd level for deps #############################################

Description

3rd level for deps #############################################

Usage

envPkgDepDepsCreate()

Extract info from package character strings

Description

Cleans a character vector of non-package name related information (e.g., version)

Usage

extractPkgName(pkgs, filenames)

extractVersionNumber(pkgs, filenames)

extractInequality(pkgs)

extractPkgGitHub(pkgs)

Arguments

Value

Just the package names without extraneous info.

See Also

trimVersionNumber()

Examples

extractPkgName("Require (>=0.0.1)")
extractVersionNumber(c(
  "Require (<=0.0.1)",
  "PredictiveEcology/Require@development (<=0.0.4)"
))
extractInequality("Require (<=0.0.1)")
extractPkgGitHub("PredictiveEcology/Require")

A helper function to get or set CRAN repos

Description

This will get the current option in getOption('repos'), and if that is not set to a url, then it will prompt the user to select a mirror, unless ind is set, in which case, it will use that mirror (in chooseCRANmirror())

Usage

getCRANrepos(repos = NULL, ind)

Arguments

The packages argument may have up to 4 pieces of information for GitHub packages: name, repository, branch, version. For CRAN-alikes, it will only be 2 pieces: name, version. There can also be an inequality or equality, if there is a version.

Description

The packages argument may have up to 4 pieces of information for GitHub packages: name, repository, branch, version. For CRAN-alikes, it will only be 2 pieces: name, version. There can also be an inequality or equality, if there is a version.

Usage

getDeps(pkgDT, which, recursive, type = type, repos, libPaths, verbose)

Arguments

Details

If version is not supplied, it will take the local, installed version, if it exists. Otherwise, it is assumed that the HEAD is desired. The function will find it in the ap or on github.com. For github packages, this is obviously a slow step, which can be accelerated if user supplies a sha or a version e.g., getDeps("PredictiveEcology/LandR@development (==1.0.2)")

Value

A (named) vector of SaveNames, which is a concatenation of the 2 or 4 elements above, plus the which and the recursive.

Invert a 2-level list

Description

This is a simple version of purrr::transpose, only for lists with 2 levels.

Usage

invertList(l)

Arguments

Value

A list with 2 levels deep, inverted from l

Examples

# create a 2-deep, 2 levels in first, 3 levels in second
a <- list(a = list(d = 1, e = 2:3, f = 4:6), b = list(d = 5, e = 55))
invertList(a) # creates 2-deep, now 3 levels outer --> 2 levels inner

Join a data.table with a Package column to available.packages

Description

Will join available.packages() with pkgDT, if pkgDT does not already have a column named Depends, which would be an indicator that this had already happened.

Usage

joinToAvailablePackages(pkgDT, repos, type, which, verbose)

Arguments

Value

The returned data.table will have most of the columns from available.packages appended to the pkgDT, including Depends, Imports, Suggests. It will change the column name that is normally returned from available.packages as Version to VersionOnRepos.

Create link to file, falling back to making a copy if linking fails.

Description

First try to create a hardlink to the file. If that fails, try a symbolic link (symlink) before falling back to copying the file. "File" here can mean a file or a directory.

Usage

linkOrCopy(from, to, allowSymlink = FALSE)

fileRenameOrMove(from, to)

Arguments

This converts master or main to HEAD for a git repo

Description

This will also convert a git repo with nothing after the @ to @HEAD

Usage

masterMainToHead(gitRepo)

Arguments

Value

The git repository with @HEAD if it had @master, @main or no @.

Use message to print a clean square data structure

Description

Sends to message, but in a structured way so that a data.frame-like can be cleanly sent to messaging.

This will only show a message if the value of verbose is greater than the verboseLevel. This is mostly useful for developers of code who want to give users of their code easy access to how verbose their code will be. A developer of a function will place this messageVerbose internally, setting the verboseLevel according to how advanced they may want the message to be. 1 is a reasonable default for standard use, 0 would be for "a very important message for all users", 2 or above would be increasing levels of details for e.g., advanced use. If a user sets to -1 with this numeric approach, they can avoid all messaging.

Usage

messageDF(df, round, verbose = getOption("Require.verbose"), verboseLevel = 1)

messageVerbose(..., verbose = getOption("Require.verbose"), verboseLevel = 1)

messageVerboseCounter(
  pre = "",
  post = "",
  verbose = getOption("Require.verbose"),
  verboseLevel = 1,
  counter = 1,
  total = 1,
  minCounter = 1
)

Arguments

Value

Used for side effects, namely messaging that can be turned on or off with different numeric values of verboseLevel. A user sets the verboseLevel for a particular message.

modifyList for multiple lists

Description

This calls utils::modifyList iteratively using base::Reduce, so it can handle >2 lists. The subsequent list elements that share a name will override previous list elements with that same name. It also will handle the case where any list is a NULL. Note: default keep.null = TRUE, which is different than modifyList

Usage

modifyList2(..., keep.null = FALSE)

modifyList3(..., keep.null = TRUE)

Arguments

Details

More or less a convenience around Reduce(modifyList, list(...)), with some checks, and the addition of keep.null = TRUE by default.

Note

modifyList3 retains the original behaviour of modifyList2 (prior to Oct 2022); however, it cannot retain NULL values in lists.

Examples

modifyList2(list(a = 1), list(a = 2, b = 2))
modifyList2(list(a = 1), NULL, list(a = 2, b = 2))
modifyList2(
  list(a = 1), list(x = NULL), list(a = 2, b = 2),
  list(a = 3, c = list(1:10))
)

Normalize filepath

Description

Checks the specified filepath for formatting consistencies:

1. use slash instead of backslash;

2. do tilde etc. expansion;

3. remove trailing slash.

Usage

normPath(path)

## S4 method for signature 'character'
normPath(path)

## S4 method for signature 'list'
normPath(path)

## S4 method for signature 'NULL'
normPath(path)

## S4 method for signature 'missing'
normPath()

## S4 method for signature 'logical'
normPath(path)

Arguments

Value

Character vector of cleaned up filepaths.

Examples

## normalize file paths
paths <- list("./aaa/zzz",
              "./aaa/zzz/",
              ".//aaa//zzz",
              ".//aaa//zzz/",
              ".\\\\aaa\\\\zzz",
              ".\\\\aaa\\\\zzz\\\\",
              file.path(".", "aaa", "zzz"))

checked <- normPath(paths)
length(unique(checked)) ## 1; all of the above are equivalent

## check to see if a path exists
tmpdir <- file.path(tempdir(), "example_checkPath")

dir.exists(tmpdir) ## FALSE
tryCatch(checkPath(tmpdir, create = FALSE), error = function(e) FALSE) ## FALSE

checkPath(tmpdir, create = TRUE)
dir.exists(tmpdir) ## TRUE

unlink(tmpdir, recursive = TRUE) # clean up

Convert numeric to character with padding

Description

This will pad floating point numbers, right or left. For integers, either class integer or functionally integer (e.g., 1.0), it will not pad right of the decimal. For more specific control or to get exact padding right and left of decimal, try the stringi package. It will also not do any rounding. See examples.

Usage

paddedFloatToChar(x, padL = ceiling(log10(x + 1)), padR = 3, pad = "0")

Arguments

Value

Character string representing the filename.

Author(s)

Eliot McIntire and Alex Chubaty

Examples

paddedFloatToChar(1.25)
paddedFloatToChar(1.25, padL = 3, padR = 5)
paddedFloatToChar(1.25, padL = 3, padR = 1) # no rounding, so keeps 2 right of decimal

2nd level

Description

2nd level

Usage

pakEnv()

Parse a github package specification

Description

This converts a specification like PredictiveEcology/Require@development into separate columns, "Account", "Repo", "Branch", "GitSubFolder" (if there is one)

Usage

parseGitHub(pkgDT, verbose = getOption("Require.verbose"))

Arguments

Details

parseGitHub turns the single character string representation into 3 or 4: Account, Repo, Branch, SubFolder.

Value

parseGitHub returns a data.table with added columns.

2nd level

Description

2nd level

Usage

pkgDepEnv()

Package dependencies when one or more packages removed

Description

This is primarily for package developers. It allows the testing of what the recursive dependencies would be if a package was removed from the immediate dependencies.

Usage

pkgDepIfDepRemoved(
  pkg = character(),
  depsRemoved = character(),
  verbose = getOption()
)

Arguments

Value

A list with 3 named lists Direct, Recursive and IfRemoved. Direct will show the top level direct dependencies, either Remaining or Removed. Recursive will show the full recursive dependencies, either Remaining or Removed. IfRemoved returns all package dependencies that are removed for each top level dependency. If a top level dependency is not listed in this final list, then it means that it is also a recursive dependency elsewhere, so its removal has no effect.

Examples

## Not run: 
if (Require:::.runLongExamples()) {
  opts <- Require:::.setupExample()

  pkgDepIfDepRemoved("reproducible", "data.table")

  Require:::.cleanup(opts)
}

## End(Not run)

Reverse package depends

Description

This is a wrapper around tools::dependsOnPkgs, but with the added option of topoSort, which will sort them such that the packages at the top will have the least number of dependencies that are in pkgs. This is essentially a topological sort, but it is done heuristically. This can be used to e.g., detach or unloadNamespace packages in order so that they each of their dependencies are detached or unloaded first.

pkgDep2 is a convenience wrapper of pkgDep that "goes one level in", i.e., the first order dependencies, and runs the pkgDep on those.

This will first look in local filesystem (in .libPaths()) and will use a local package to find its dependencies. If the package does not exist locally, including whether it is the correct version, then it will look in (currently) CRAN and its archives (if the current CRAN version is not the desired version to check). It will also look on GitHub if the package description is of the form of a GitHub package with format account/repo@branch or account/repo@commit. For this, it will attempt to get package dependencies from the GitHub ‘DESCRIPTION’ file. This is intended to replace tools::package_dependencies or pkgDep in the miniCRAN package, but with modifications to allow multiple sources to be searched in the same function call.

Usage

pkgDepTopoSort(
  pkgs,
  deps,
  reverse = FALSE,
  topoSort = TRUE,
  libPaths,
  useAllInSearch = FALSE,
  returnFull = TRUE,
  recursive = TRUE,
  purge = getOption("Require.purge", FALSE),
  which = c("Depends", "Imports", "LinkingTo"),
  type = getOption("pkgType"),
  verbose = getOption("Require.verbose"),
  ...
)

pkgDep2(...)

pkgDep(
  packages,
  libPaths,
  which = c("Depends", "Imports", "LinkingTo"),
  recursive = TRUE,
  depends,
  imports,
  suggests,
  linkingTo,
  repos = getOption("repos"),
  keepVersionNumber = TRUE,
  includeBase = FALSE,
  includeSelf = TRUE,
  sort = TRUE,
  simplify = TRUE,
  purge = getOption("Require.purge", FALSE),
  verbose = getOption("Require.verbose"),
  type = getOption("pkgType"),
  Additional_repositories = FALSE,
  ...
)

Arguments

Value

A possibly ordered, named (with packages as names) list where list elements are either full reverse depends.

Note

tools::package_dependencies and pkgDep will differ under the following circumstances:

1. GitHub packages are not detected using tools::package_dependencies;

2. tools::package_dependencies does not detect the dependencies of base packages among themselves, e.g., methods depends on stats and graphics.

Examples

## Not run: 
if (Require:::.runLongExamples()) {
  opts <- Require:::.setupExample()

  pkgDepTopoSort(c("Require", "data.table"), reverse = TRUE)

  Require:::.cleanup(opts)
}

## End(Not run)

## Not run: 
if (Require:::.runLongExamples()) {
  opts <- Require:::.setupExample()

  pkgDep2("reproducible")
  # much bigger one
  pkgDep2("tidyverse")

  Require:::.cleanup(opts)
}

## End(Not run)
## Not run: 
if (Require:::.runLongExamples()) {
  opts <- Require:::.setupExample()

  pkgDep("tidyverse", recursive = TRUE)

  # GitHub, local, and CRAN packages
  pkgDep(c("PredictiveEcology/reproducible", "Require", "plyr"))

  Require:::.cleanup(opts)
}

## End(Not run)

Take a snapshot of all the packages and version numbers

Description

This can be used later by Require to install or re-install the correct versions. See examples.

Usage

pkgSnapshot(
  packageVersionFile = getOption("Require.packageVersionFile"),
  libPaths = .libPaths(),
  standAlone = FALSE,
  purge = getOption("Require.purge", FALSE),
  exact = TRUE,
  includeBase = FALSE,
  verbose = getOption("Require.verbose")
)

pkgSnapshot2(
  packageVersionFile = getOption("Require.packageVersionFile"),
  libPaths,
  standAlone = FALSE,
  purge = getOption("Require.purge", FALSE),
  exact = TRUE,
  includeBase = FALSE,
  verbose = getOption("Require.verbose")
)

Arguments

Details

A file is written with the package names and versions of all packages within libPaths. This can later be passed to Require.

pkgSnapshot2 returns a vector of package names and versions, with no file output. See examples.

Value

Will both write a file, and (invisibly) return a vector of packages with the version numbers. This vector can be used directly in Require, though it should likely be used with require = FALSE to prevent attaching all the packages.

Examples

## Not run: 
if (Require:::.runLongExamples()) {
  opts <- Require:::.setupExample()

  # install one archived version so that below does something interesting
  libForThisEx <- tempdir2("Example")
  Require("crayon (==1.5.1)", libPaths = libForThisEx, require = FALSE)
  # Normal use -- using the libForThisEx for example;
  #    normally libPaths would be omitted to get all
  #    packages in user or project library
  tf <- tempfile()

  # writes to getOption("Require.packageVersionFile")
  # within project; also returns a vector
  # of packages with version
  pkgs <- pkgSnapshot(
    packageVersionFile = tf,
    libPaths = libForThisEx, standAlone = TRUE # only this library
  )

  # Now move this file to another computer e.g. by committing in git,
  #   emailing, googledrive
  #   on next computer/project
  Require(packageVersionFile = tf, libPaths = libForThisEx)

  # Using pkgSnapshot2 to get the vector of packages and versions
  pkgs <- pkgSnapshot2(
    libPaths = libForThisEx, standAlone = TRUE
  )
  Install(pkgs) # will install packages from previous line

  Require:::.cleanup(opts)
  unlink(getOption("Require.packageVersionFile"))
}

## End(Not run)

Recursive function to remove .basePkgs

Description

Recursive function to remove .basePkgs

Usage

rmBase(includeBase = formals(pkgDep)[["includeBase"]], deps)

Arguments

R versions

Description

Reference table of R versions and their release dates (2018 and later).

Usage

rversions

Format

An object of class data.frame with 21 rows and 2 columns.

Details

Update this as needed using rversions::r_versions():

⁠ # install.packages("rversions") v = rversions::r_versions() keep = which(as.Date(v$date, format = " as.Date("2018-01-01", format = " dput(v[keep, c("version", "date")]) ⁠

Set .libPaths

Description

This will set the .libPaths() by either adding a new path to it if standAlone = FALSE, or will concatenate c(libPath, tail(.libPaths(), 1)) if standAlone = TRUE. Currently, the default is to make this new .libPaths() "sticky", meaning it becomes associated with the current directory even through a restart of R. It does this by adding and/updating the ‘.Rprofile’ file in the current directory. If this current directory is a project, then the project will have the new .libPaths() associated with it, even through an R restart.

Usage

setLibPaths(
  libPaths,
  standAlone = TRUE,
  updateRprofile = getOption("Require.updateRprofile", FALSE),
  exact = FALSE,
  verbose = getOption("Require.verbose")
)

Arguments

Details

This details of this code were modified from https://github.com/milesmcbain. A different, likely non-approved by CRAN approach that also works is here: https://stackoverflow.com/a/36873741/3890027.

Value

The main point of this function is to set .libPaths(), which will be changed as a side effect of this function. As when setting options, this will return the previous state of .libPaths() allowing the user to reset easily.

Examples

## Not run: 
if (Require:::.runLongExamples()) {
  opts <- Require:::.setupExample()
  origDir <- setwd(tempdir())
  td <- tempdir()
  setLibPaths(td) # set a new R package library locally
  setLibPaths() # reset it to original
  setwd(origDir)
  # Using standAlone = FALSE means that newly installed packages
  #   will be installed
  #   in the new package library, but loading packages can come
  #   from any of the ones listed in .libPaths()

  # will have 2 or more paths
  otherLib <- file.path(td, "newProjectLib")
  setLibPaths(otherLib, standAlone = FALSE)
  # Can restart R, and changes will stay

  # remove the custom .libPaths()
  setLibPaths() # reset to previous; remove from .Rprofile
  # because libPath arg is empty

  Require:::.cleanup(opts)
  unlink(otherLib, recursive = TRUE)
}

## End(Not run)

Setup for binary Linux repositories

Description

Enable use of binary package builds for Linux from the RStudio Package Manager repo. This will set the repos option, affecting the current R session. It will put this binaryLinux in the first position. If the getOption("repos") is NULL, it will put backupCRAN in second position.

Usage

setLinuxBinaryRepo(
  binaryLinux = urlForArchivedPkgs,
  backupCRAN = srcPackageURLOnCRAN
)

Arguments

Like setdiff, but takes into account names

Description

This will identify the elements in l1 that are not in l2. If missingFill is provided, then elements that are in l2, but not in l1 will be returned, assigning missingFill to their values. This might be NULL or "", i.e., some sort of empty value. This function will work on named lists, named vectors and likely on other named classes.

Usage

setdiffNamed(l1, l2, missingFill)

Arguments

Details

There are 3 types of differences that might occur with named elements: 1. a new named element, 2. an removed named element, and 3. a modified named element. This function captures all of these. In the case of unnamed elements, e.g., setdiff, the first two are not seen as differences, if the values are not different.

Value

A vector or list of the elements in l1 that are not in l2, and optionally the elements of l2 that are not in l1, with values set to missingFill

Setup a project library, cache, options

Description

setup and setupOff are currently deprecated. These may be re-created in a future version. In its place, a user can simply put .libPaths(libs, include.site = FALSE) in their .Rprofile file, where libs is the directory where the packages should be installed and should be a folder with the R version number, e.g., derived by using checkLibPaths(libs).

Usage

setup(
  newLibPaths,
  RPackageFolders,
  RPackageCache = cacheGetOptionCachePkgDir(),
  standAlone = getOption("Require.standAlone", TRUE),
  verbose = getOption("Require.verbose")
)

setupOff(removePackages = FALSE, verbose = getOption("Require.verbose"))

Arguments

A list of R packages that should likely be installed from Source, not Binary

Description

The list of R packages that Require installs from source on Linux, even if the getOptions("repos") is a binary repository. This list can be updated by the user by modifying the options Require.spatialPkgs or Require.otherPkgs. Default "force source only packages" are visible with RequireOptions().

Usage

sourcePkgs(additional = NULL, spatialPkgs = NULL, otherPkgs = NULL)

Arguments

Value

A sorted concatenation of the 3 input parameters.

split for a data.table that keeps integrity of a column of lists of data.table objects

Description

data.table::split does 2 bad things:

1. reorders if using f

2. destroys the integrity of a column that is a list of data.tables, when using by So, to keep order, need by, but to keep integrity, need f. This function

Usage

splitKeepOrderAndDTIntegrity(pkgDT, splitOn)

Arguments

Value

A list of data.table objects of length(unique(splitOn)).

download.files or install.packages in a separate process

Description

This uses sys package so that messaging can be controlled. This also provides the option to parallelize by spawning multiple background process to allow parallel e.g., downloads. Noting that if libcurl is installed (and detected using capabilities("libcurl")), then no explicit parallelism will be allowed, instead method = "libcurl" will be passed enabling parallel downloads.

Usage

sysInstallAndDownload(
  args,
  splitOn = "pkgs",
  doLine = "outfiles <- do.call(download.packages, args)",
  returnOutfile = FALSE,
  doLineVectorized = TRUE,
  tmpdir,
  libPaths,
  verbose
)

Arguments

Value

Mostly for side effects, namely installed packages or downloaded packages or files. However, in the case of returnOutfile = TRUE, then a list of filenames will be returned with any outputs from the doLine.

Make a temporary (sub-)directory

Description

Create a temporary subdirectory in .RequireTempPath(), or a temporary file in that temporary subdirectory.

Usage

tempdir2(
  sub = "",
  tempdir = getOption("Require.tempPath", .RequireTempPath()),
  create = TRUE
)

Arguments

See Also

tempfile2()

Make a temporary subfile in a temporary (sub-)directory

Description

Make a temporary subfile in a temporary (sub-)directory

Usage

tempfile2(
  sub = "",
  tempdir = getOption("Require.tempPath", .RequireTempPath()),
  ...
)

Arguments

See Also

tempdir2()

Trim version number off a compound package name

Description

The resulting string(s) will have only name (including github.com repository if it exists).

Usage

trimVersionNumber(pkgs)

Arguments

See Also

extractPkgName()

Examples

trimVersionNumber("PredictiveEcology/Require (<=0.0.1)")

Update installed packages with latest available versions

Description

Similar to update.packages, but works for archived, non-archived, and Github packages.

Usage

updatePackages(
  libPaths = .libPaths()[1],
  purge = FALSE,
  verbose = getOption("Require.verbose")
)

Arguments

Value

Run for its side effect, namely, updating installed packages to their latest possible state, whether they are on CRAN currently, archived, or on GitHub.