From cf440786ca3f863365cd3954688f59e3d31b3d78 Mon Sep 17 00:00:00 2001
From: pharmaverse-bot <113703390+pharmaverse-bot@users.noreply.github.com>
Date: Thu, 7 Sep 2023 17:48:24 +0200
Subject: [PATCH] exclude files and add auto-doc

---
 .github/workflows/check-templates.yml | 78 ++++++++++++++++++++++-----
 1 file changed, 65 insertions(+), 13 deletions(-)

diff --git a/.github/workflows/check-templates.yml b/.github/workflows/check-templates.yml
index de5e9aac..e1992c23 100644
--- a/.github/workflows/check-templates.yml
+++ b/.github/workflows/check-templates.yml
@@ -7,6 +7,11 @@ on:
         default: '4.1'
         required: false
         type: string
+      exclude-templates-data:
+        description: 'List of data generated by templates to exclude (templates script will run, but results won't be push to pharmaverseadam repo) - comma seperated list'
+        default: 'adrs_basic_onco'
+        required: false
+        type: 'string'
   push:
     branches:
       - main
@@ -189,7 +194,7 @@ jobs:
           }
         shell: Rscript {0}
 
-      - name: Add other data formats and suffixes
+      - name: Add other data formats and suffixes, exclude data
         run: |
           library(readxl)
           library(zip)
@@ -202,34 +207,80 @@ jobs:
           folder_path <- file.path(Sys.getenv("GITHUB_WORKSPACE"), "tmp")
           rda_files <- list.files(path = folder_path, pattern = "\\.rda$")
 
+          ex_files <- unlist(strsplit("${{ inputs.exclude-templates-data }}", ",")) # files to exclude
+
           for (rda_file in rda_files) {
-            print(sprintf("converting file %s", rda_file))
-            data <- loadRData(file.path(folder_path, rda_file))
-            suffix <- gsub("pharmaverse/admiral", "", "${{ github.repository }}")
-            if (nchar(suffix) > 0) {
-              rda_file_w_suffix <- gsub("\\.rda$", sprintf("_%s.rda", suffix), rda_file)
-              csv_file <- gsub("\\.rda$", sprintf("_%s.csv", suffix), rda_file)
-              file.rename(file.path(folder_path, rda_file), file.path(folder_path, rda_file_w_suffix))
+            # delete file if part of inputs.exclude-templates-data
+            if (rda_file %in% ex_files) {
+              file.remove(file.path(folder_path, rda_file))
+              cat("Deleted:", rda_file, "\n")
             }
             else {
-              csv_file <- gsub("\\.rda$", ".csv", rda_file)
+              print(sprintf("converting file %s", rda_file))
+              data <- loadRData(file.path(folder_path, rda_file))
+              suffix <- gsub("pharmaverse/admiral", "", "${{ github.repository }}")
+              if (nchar(suffix) > 0) {
+                rda_file_renamed <- gsub("\\.rda$", sprintf("_%s.rda", suffix), rda_file)
+                csv_file <- gsub("\\.rda$", sprintf("_%s.csv", suffix), rda_file)
+                file.rename(file.path(folder_path, rda_file), file.path(folder_path, rda_file_renamed))
+              }
+              else {
+                csv_file <- gsub("\\.rda$", ".csv", rda_file)
+                rda_file_renamed <- rda_file
+              }
+              write.csv(data, file = file.path(folder_path, csv_file), row.names = FALSE)
+
+              # create associated documentation inside R folder
+              dataset_name <- gsub("\\.rda$", "", rda_file_renamed)
+              doc_string <- paste(
+                '#', sprintf('Dataset %s', dataset_name),
+                '#',
+                sprintf('%s dataset', dataset_name),
+                '#',
+                sprintf('@name %s', dataset_name),
+                '# @docType data',
+                '# @format A data frame with', ncol(data), 'columns:',
+                '   \\describe{',
+                sapply(names(data), function(col_name) {
+                  paste('#     \\item{', col_name, '}{Description goes here}')
+                }, USE.NAMES = FALSE),
+                '#   }',
+                sprintf('# @source Generated from ${{ github.repository_name }}.'),
+                '# @references None',
+                sprintf('# @examples\n# load(%s)', dataset_name),
+                '',
+                collapse = '\n'
+              )
+
+              doc_dir <- "datasets_doc"
+              if (!file.exists(doc_dir)) {
+                dir.create(doc_dir, recursive = TRUE)
+              }
+              writeLines(doc_string, con = file.path(doc_dir, paste0(dataset_name, ".R")))
             }
-            write.csv(data, file = file.path(folder_path, csv_file), row.names = FALSE)
           }
         shell: Rscript {0}
 
       # zip templates data
-      - name: zip templates data
+      - name: zip artifacts data
         run: |
           find "$GITHUB_WORKSPACE/tmp" -type f \( -name "*.rda" -o -name "*.csv" \) -exec zip -j "$GITHUB_WORKSPACE/data.zip" {} \;
+          find "$GITHUB_WORKSPACE/datasets_doc" -type f \( -name "*.R" -o \) -exec zip -j "$GITHUB_WORKSPACE/doc.zip" {} \;
 
       # store templates data as artifacts
-      - name: Archive production artifacts
+      - name: Archive templates data
         uses: actions/upload-artifact@v3
         with:
           name: data_templates
           path: ${{ github.workspace }}/data.zip
 
+      # store templates data as artifacts
+      - name: Archive doc
+        uses: actions/upload-artifact@v3
+        with:
+          name: doc_templates
+          path: ${{ github.workspace }}/doc.zip
+
       - name: Checkout repo (PR) 🛎
         uses: actions/checkout@v3
         with:
@@ -255,6 +306,7 @@ jobs:
         run: |
           head_data_folder=$(git ls-tree -r HEAD data) # check if data folder exists on remote main branch
           unzip "${GITHUB_WORKSPACE}/data.zip" -d "${GITHUB_WORKSPACE}/inst/extdata/"
+          unzip "${GITHUB_WORKSPACE}/doc.zip" -d "${GITHUB_WORKSPACE}/R"
           # move every rda files inside data folder
           mkdir -p data
           for rda_file in inst/extdata/*.rda; do
@@ -296,7 +348,7 @@ jobs:
         uses: stefanzweifel/git-auto-commit-action@v4
         with:
           commit_message: templates data updates from ${{ github.repository }}
-          file_pattern: 'inst/extdata/* data/*'
+          file_pattern: 'inst/extdata/* data/* R/*'
           commit_user_name: pharmaverse-bot
           commit_user_email: 113703390+pharmaverse-bot@users.noreply.github.com
           branch: "${{ steps.branch.outputs.source-branch }}"