Merge pull request #153 from jhudsl/ki/refactor

start functionalizing make screenshots

.github/workflows/R-CMD-check.yaml

@@ -36,7 +36,7 @@ jobs:
           rcmdcheck::rcmdcheck(args = c("--no-manual"), check_dir = "check")
         shell: Rscript {0}
         env:
-          GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
+          GH_PAT: ${{ secrets.GH_PAT }}
 
       - name: Check testthat
         id: check_check

DESCRIPTION

@@ -35,13 +35,15 @@ Imports:
     rprojroot,
     rvest,
     stringr,
+    webshot2,
     xml2,
     yaml
 Suggests:
-    remotes,
     testthat (>= 3.0.0),
     tibble,
     utils
+Remotes: 
+    jhudsl/cow
 VignetteBuilder: 
     knitr
 ByteCompile: true

NAMESPACE

@@ -23,6 +23,7 @@ export(get_gs_pptx)
 export(get_image_from_slide)
 export(get_image_link_from_slide)
 export(get_object_id_notes)
+export(get_pages_url)
 export(get_slide_id)
 export(get_slide_page)
 export(get_yaml_spec)
@@ -32,6 +33,7 @@ export(gs_png_download)
 export(gs_png_url)
 export(include_slide)
 export(make_embed_markdown)
+export(make_screenshots)
 export(output_destination)
 export(parse_q_tag)
 export(parse_quiz)
@@ -46,6 +48,7 @@ export(set_up_leanpub)
 export(unzip_pptx)
 export(website_to_embed_leanpub)
 export(xml_notes)
+import(dplyr)
 importFrom(httr,GET)
 importFrom(httr,accept_json)
 importFrom(httr,config)
@@ -57,10 +60,13 @@ importFrom(jsonlite,fromJSON)
 importFrom(magrittr,"%>%")
 importFrom(openssl,aes_cbc_decrypt)
 importFrom(readr,write_tsv)
+importFrom(rprojroot,find_root)
+importFrom(rprojroot,has_dir)
 importFrom(utils,download.file)
 importFrom(utils,installed.packages)
 importFrom(utils,menu)
 importFrom(utils,unzip)
+importFrom(webshot2,webshot)
 importFrom(xml2,read_xml)
 importFrom(xml2,xml_find_all)
 importFrom(xml2,xml_text)

R/get_pages_url.R

@@ -0,0 +1,81 @@
+#' Retrieve pages url for a repo
+#'
+#' Given an repository on GitHub, retrieve the pages URL for it.
+#'
+#' @param repo_name The full name of the repo to get bookdown chapters from.
+#' e.g. "jhudsl/OTTR_Template"
+#' @param git_pat If private repositories are to be retrieved, a github personal
+#' access token needs to be supplied. If none is supplied, then this will attempt to
+#' grab from a git pat set in the environment with usethis::create_github_token().
+#' Authorization handled by \link[cow]{get_git_auth}
+#' @param verbose TRUE/FALSE do you want more progress messages?
+#' @param keep_json verbose TRUE/FALSE keep the json file locally?
+#'
+#' @return a data frame with the repository with the following columns:
+#' data_level, data_path, chapt_name, url, repository name
+#'
+#' @importFrom magrittr %>%
+#' @import dplyr
+#'
+#' @export
+#'
+#' @examples \dontrun{
+#'
+#' usethis::create_github_token()
+#'
+#' get_pages_url("jhudsl/Documentation_and_Usability")
+#' }
+get_pages_url <- function(repo_name,
+                          git_pat = NULL,
+                          verbose = FALSE,
+                          keep_json = FALSE) {
+  page_url <- NA
+
+  # Try to get credentials other way
+  auth_arg <- get_git_auth(git_pat = git_pat, quiet = !verbose)
+
+  git_pat <- try(auth_arg$password, silent = TRUE)
+
+  if (grepl("Error", git_pat[1])) {
+    warning("Cannot retrieve page info without GitHub credentials. Passing an NA.")
+  }
+
+  # We can only retrieve pages if we have the credentials
+  if (!grepl("Error", git_pat[1])) {
+    exists <- check_git_repo(
+      repo_name = repo_name,
+      git_pat = git_pat,
+      verbose = FALSE
+    )
+
+    if (exists) {
+      # Get repo info
+      repo_info <- get_repo_info(
+        repo_name = repo_name,
+        git_pat = git_pat
+      )
+
+      # Declare URL
+      url <- paste0("https://api.github.com/repos/", repo_name, "/pages")
+
+      # Github api get
+      response <- httr::GET(
+        url,
+        httr::add_headers(Authorization = paste0("token ", auth_arg$password)),
+        httr::accept_json()
+      )
+
+      if (httr::http_error(response)) {
+        if (verbose) {
+          warning(paste0("url: ", url, " failed"))
+        }
+      } else {
+        # Get content as JSON
+        page_info <- httr::content(response, as = "parsed")
+
+        page_url <- page_info$html_url
+      }
+    }
+  }
+  return(page_url)
+}

R/screenshot.R

@@ -0,0 +1,74 @@
+#' A function to make screenshots from an OTTR bookdown website
+#' @description This function creates screenshots of course chapters that are stored in a created output directory 
+#'
+#' @param git_pat required argument; a Git secret -- see https://docs.github.com/en/authentication/keeping-your-account-and-data-secure/managing-your-personal-access-tokens for more info
+#' @param repo required argument; GitHub repository name, e.g., jhudsl/OTTR_Template
+#' @param output_dir default is "resources/chapt_screen_images"; Output directory where the chapter's screen images should be stored. For OTTR courses, don't change this unless you've changed the downstream functions accordingly. 
+#' @param base_url default is NULL; rendered bookdown URL where screenshots are taken from, if NULL, the function will use the repo_name and and git_pat to find the base_url
+#'
+#' @return the file path for file where chapter urls are saved
+#'
+#' @import dplyr
+#' @importFrom webshot2 webshot
+#' @importFrom magrittr %>%
+#' @importFrom rprojroot find_root has_dir
+#' 
+#' @author Candace Savonen
+#'
+#' @export
+#'
+#' @examples \dontrun{
+#' 
+#'  make_screenshots(Sys.getenv("secrets.GH_PAT"), "jhudsl/OTTR_Template")
+#'  
+#' }
+make_screenshots <- function(git_pat, repo, output_dir = "resources/chapt_screen_images", base_url = NULL){
+
+  # Find .git root directory
+  root_dir <- find_root(has_dir(".git"))
+
+  output_folder <- file.path(output_dir)
+  if (!dir.exists(output_folder)) {
+    dir.create(output_folder, recursive = TRUE)
+  }
+
+  if (is.null(base_url)){
+    base_url <- ottrpal::get_pages_url(repo_name = repo, git_pat = git_pat) #what if these arguments are still NULL/not supplied?
+    base_url <- gsub("/$", "", base_url)
+  }
+
+  # Collect all the chapter pages for the url given
+  chapt_df <- ottrpal::get_chapters(html_page = file.path(root_dir, "docs", "index.html"),
+                                    base_url = base_url)
+
+  # Now take screenshots for each
+  file_names <- lapply(chapt_df$url, function(url){
+    file_name <- gsub(".html", 
+                      ".png", 
+                      file.path(output_folder, basename(url))
+                      )
+
+    # Get rid of special characters because leanpub no like
+    file_name <- gsub(":|?|!|\\'",
+                      "",
+                      file_name
+                      )
+
+    # Take the screenshot
+    webshot(url, file = file_name)
+
+    return(file_name)
+  })
+
+  # Save file of chapter urls and file_names
+  chapt_df <- chapt_df %>%
+    dplyr::mutate(img_path = unlist(file_names))
+
+  chapt_df %>% 
+    readr::write_tsv(file.path(output_folder, "chapter_urls.tsv"))
+
+  message(paste("Image Chapter key written to: ", file.path(output_folder, "chapter_urls.tsv")))
+
+  return(file.path(output_folder, "chapter_urls.tsv"))
+
+}