add_dockerfile {pracpac}

Add a Dockerfile to the docker directory

Description

Adds a Dockerfile to the docker directory created by create_docker_dir. Allows for specification of several preset use cases, whether or not use use renv to manage dependencies, and optional overriding the base image.

Usage

add_dockerfile(
  pkg_path = ".",
  img_path = NULL,
  use_renv = TRUE,
  use_case = "default",
  base_image = NULL,
  repos = NULL
)

Arguments

pkg_path

Path to the package directory. Default is "." for the current working directory, which assumes developer is working in R package root. However, this can be set to another path as needed.

img_path

Path to the write the docker image definition contents. The default NULL will use ⁠docker/⁠ as a subdirectory of the pkg_path.

use_renv

Logical; use renv? Defaults to TRUE. If FALSE, package dependencies are scraped from the DESCRIPTION file and the most recent versions will be installed in the image.

use_case

Name of the use case. Defaults to "default", which only uses the base boilerplate. See vignette("use-cases", package="pracpac") for other use cases (e.g., shiny, rstudio, pipeline).

base_image

Name of the base image to start FROM. Default is NULL and the base image will be derived based on use_case. Optionally override this by setting the name of the base image (including tag if desired).

repos

Option to override the repos used for installing packages with renv by passing name of repository. Only used if use_renv = TRUE. Default is NULL meaning that the repos specified in renv lockfile will remain as-is and not be overridden.

Details

This function is run as part of use_docker but can be used on its own.

See vignette("use-cases", package="pracpac") for details on use cases.

Value

Invisibly returns a list of package info returned by pkg_info. Primarily called for side-effect to create Dockerfile.

Examples

## Not run: 

# Specify path to example package source and copy to tempdir()
# Note that in practice you do not need to copy to a tempdir()
# And in fact it may be easiest to use pracpac relative to your package directory root
ex_pkg_src <- system.file("hellow", package = "pracpac", mustWork = TRUE)
file.copy(from = ex_pkg_src, to = tempdir(), recursive = TRUE)

# Default: FROM rocker/r-ver:latest with no additional template
# By default add_dockerfile requires you either to specify use_renv = FALSE
# Or run renv_deps() prior to add_dockerfile()
# The use_docker() wrapper runs these sequentially, and is recommended for most usage
add_dockerfile(pkg_path = file.path(tempdir(), "hellow"), use_renv = FALSE)
# Specify tidyverse base image
renv_deps(pkg_path = file.path(tempdir(), "hellow"))
add_dockerfile(pkg_path = file.path(tempdir(), "hellow"), base_image="rocker/tidyverse:4.2.2")
# Specify different default repo
add_dockerfile(pkg_path = file.path(tempdir(), "hellow"), repos="https://cran.wustl.edu/")
# RStudio template
add_dockerfile(pkg_path = file.path(tempdir(), "hellow"), use_case="rstudio")
# Shiny template
add_dockerfile(pkg_path = file.path(tempdir(), "hellow"), use_case = "shiny")
# Pipeline template
add_dockerfile(pkg_path = file.path(tempdir(), "hellow"), use_case="pipeline")

## End(Not run)