Dockerize R package build

.github/workflows/master.yml

@@ -137,6 +137,10 @@ jobs:
         Rscript -e 'covr::codecov()' || :
       env:
         COVR_RUNNING: true
+    - name: Test package build
+      working-directory: mlflow/R/mlflow
+      run: |
+        ./build-package.sh
     - name: Show 00check.log on failure
       if: ${{ failure() }}
       run: |

mlflow/R/mlflow/.Rbuildignore

@@ -19,3 +19,7 @@ Reference_Manual_mlflow.md
 ^logs/
 ^depends\.Rds
 ^R-version
+^\.utils\.R$
+^\.build-package\.R$
+^build-package\.sh$
+^Dockerfile\.build$

mlflow/R/mlflow/.build-package.R

@@ -0,0 +1,16 @@
+source(".utils.R")
+
+# Increase the timeout length for `utils::download.file` because the default value (60 seconds)
+# could be too short to download large packages such as h2o.
+options(timeout=300)
+# Install dependencies required for the submission check.
+devtools::install_deps(".", dependencies = TRUE)
+# Bundle up the package into a .tar.gz file. This file will be submitted to CRAN.
+package_path <- devtools::build(".")
+# Run the submission check against the built package.
+devtools::check_built(
+    package_path,
+    remote = should_enable_cran_incoming_checks(),
+    error_on = "note",
+    args = c("--no-tests", "--as-cran"),
+)

mlflow/R/mlflow/.gitignore

@@ -19,3 +19,4 @@ Reference_Manual_mlflow.md
 *.Rcheck/
 depends.Rds
 R-version
+*.tar.gz

mlflow/R/mlflow/.run-tests.R

@@ -1,24 +1,15 @@
+source("../.utils.R")
+
 parent_dir <- dir("../", full.names = TRUE)
 package <- parent_dir[grepl("mlflow_", parent_dir)]
 
 library(reticulate)
 use_condaenv(mlflow:::mlflow_conda_env_name())
 
-# Disable CRAN incoming feasibility check within a week after the latest release because it fails.
-#
-# Relevant code:
-# https://github.com/wch/r-source/blob/4561aea946a75425ddcc8869cdb129ed5e27af97/src/library/tools/R/QC.R#L8005-L8008
-install.packages(c("xml2", "rvest"))
-library(xml2)
-library(rvest)
-
-url <- "https://cran.r-project.org/web/packages/mlflow/index.html"
-html <- read_html(url)
-xpath <- '//td[text()="Published:"]/following-sibling::td[1]/text()'
-published_date <- as.Date(html_text(html_nodes(html, xpath=xpath)))
-today <- Sys.Date()
-days_since_last_release <- difftime(today, published_date, units="days")
-remote <- as.numeric(days_since_last_release) > 7
-
-devtools::check_built(path = package, remote = remote, error_on = "note", args = "--no-tests")
+devtools::check_built(
+    path = package,
+    remote = should_enable_cran_incoming_checks(),
+    error_on = "note",
+    args = "--no-tests"
+)
 source("testthat.R")

mlflow/R/mlflow/.utils.R

@@ -0,0 +1,18 @@
+
+# This script defines utility functions only used during development.
+
+should_enable_cran_incoming_checks <- function() {
+    # The CRAN incoming feasibility check performs a package recency check (this is undocumented)
+    # that examines the number of days since the last release and raises a NOTE if it's < 7.
+    # Relevant code:
+    # https://github.com/wch/r-source/blob/4561aea946a75425ddcc8869cdb129ed5e27af97/src/library/tools/R/QC.R#L8005-L8008
+    # This check needs to be disabled for a week after releasing a new version.
+    desc_url <- url("https://cran.r-project.org/web/packages/mlflow/DESCRIPTION")
+    field <- "Date/Publication"
+    desc <- read.dcf(desc_url, fields = c(field))
+    close(desc_url)
+    publication_date <- as.Date(unname(desc[1, field]))
+    today <- Sys.Date()
+    days_since_last_release <- as.numeric(difftime(today, publication_date, units="days"))
+    days_since_last_release > 7
+}

mlflow/R/mlflow/Dockerfile.build

@@ -0,0 +1,5 @@
+FROM rocker/r-ver:4.1.2
+
+RUN apt-get update -y
+RUN apt-get install pandoc -y
+RUN Rscript -e 'install.packages("devtools", dependencies = TRUE)'

mlflow/R/mlflow/build-package.sh

@@ -0,0 +1,5 @@
+#!/usr/bin/env bash
+set -ex
+
+docker build -f Dockerfile.build -t r-build-package .
+docker run --rm --workdir /app -v $(pwd):/app r-build-package Rscript -e 'source(".build-package.R", echo = TRUE)'