Merge pull request #507 from r-world-devs/474-optimize-get_r_package_…

…usage 474 optimize get r package usage
r-world-devs · Oct 17, 2024 · e19f470 · e19f470
2 parents ecb8a9f + 498eb4e
commit e19f470
Show file tree

Hide file tree

Showing 17 changed files with 749 additions and 291 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: GitStats
 Title: Get Statistics from GitHub and GitLab
-Version: 2.1.0.9004
+Version: 2.1.0.9005
 Authors@R: c(
     person(given = "Maciej", family = "Banas", email = "[email protected]", role = c("aut", "cre")),
     person(given = "Kamil", family = "Koziej", email = "[email protected]", role = "aut"),

diff --git a/NEWS.md b/NEWS.md
@@ -1,6 +1,15 @@
 # GitStats (development version)
 
+## Features:
+
+- Optimized `get_R_package_usage()` function: 
+  - you can now pass a vector of packages names (new `packages` parameter replacing old `package_name`) ([#494](https://github.com/r-world-devs/GitStats/issues/494)),
+  - on the other hand, output of the function has been limited to contain only most necessary data (removing all repository stats), making thus process of obtaining package usage faster ([#474](https://github.com/r-world-devs/GitStats/issues/474)).
+  - new `split_output` parameter has been added - when set to `TRUE` a `list` with `tibbles` (every element of the `list` for every package) instead of one `tibble` is returned.
 - Added possibility to get repositories for individual users with `get_repos()` ([#492](https://github.com/r-world-devs/GitStats/issues/492)). Earlier this was only possible for GitHub organizations and GitLab groups.
+
+## Fixes:
+
 - Fixed getting large search responses for GitHub ([#491](https://github.com/r-world-devs/GitStats/issues/491)).
 - Fixed checking token scopes ([#501](https://github.com/r-world-devs/GitStats/issues/501)). If token scopes are insufficient error is returned and `GitHost` is not passed to `GitStats`. This also applies to situation when `GitStats` looks for default tokens (not defined by user). Earlier, if tests for token failed, an empty token was passed and `GitStats` was created, which was misleading for the user.
 - User can now optionally pass public GitHub host name (`github.com` or `https://github.com`) to `set_github_host()` ([#475](https://github.com/r-world-devs/GitStats/issues/475)).

diff --git a/R/EngineRest.R b/R/EngineRest.R
@@ -39,7 +39,7 @@ EngineRest <- R6::R6Class("EngineRest",
     },
 
     # Prepare table for repositories content
-    prepare_repos_table = function(repos_list, verbose = TRUE) {
+    prepare_repos_table = function(repos_list, output = "table_full", verbose = TRUE) {
       repos_dt <- purrr::map(repos_list, function(repo) {
         repo <- purrr::map(repo, function(attr) {
           attr <- attr %||% ""
@@ -51,15 +51,24 @@ EngineRest <- R6::R6Class("EngineRest",
         cli::cli_alert_info("Preparing repositories table...")
       }
       if (length(repos_dt) > 0) {
-        repos_dt <- dplyr::mutate(
-          repos_dt,
-          repo_id = as.character(repo_id),
-          created_at = as.POSIXct(created_at),
-          last_activity_at = as.POSIXct(last_activity_at),
-          forks = as.integer(forks),
-          issues_open = as.integer(issues_open),
-          issues_closed = as.integer(issues_closed)
-        )
+        if (output == "table_full") {
+          repos_dt <- dplyr::mutate(
+            repos_dt,
+            repo_id = as.character(repo_id),
+            created_at = as.POSIXct(created_at),
+            last_activity_at = as.POSIXct(last_activity_at),
+            forks = as.integer(forks),
+            issues_open = as.integer(issues_open),
+            issues_closed = as.integer(issues_closed)
+          )
+        }
+        if (output == "table_min") {
+          repos_dt <- dplyr::mutate(
+            repos_dt,
+            repo_id = as.character(repo_id),
+            created_at = as.POSIXct(created_at)
+          )
+        }
       }
       return(repos_dt)
     }

diff --git a/R/EngineRestGitHub.R b/R/EngineRestGitHub.R
@@ -29,7 +29,7 @@ EngineRestGitHub <- R6::R6Class(
                                  org        = NULL,
                                  filename   = NULL,
                                  in_path    = FALSE,
-                                 raw_output = FALSE,
+                                 output     = "table_full",
                                  verbose    = TRUE,
                                  progress   = TRUE) {
       user_query <- if (!is.null(org)) {
@@ -53,12 +53,12 @@ EngineRestGitHub <- R6::R6Class(
           search_endpoint = search_endpoint,
           total_n = total_n
         )
-        if (!raw_output) {
+        if (output == "table_full" || output == "table_min") {
           search_output <- private$map_search_into_repos(
             search_response = search_result,
             progress        = progress
           )
-        } else {
+        } else if (output == "raw") {
           search_output <- search_result
         }
       } else {
@@ -68,26 +68,39 @@ EngineRestGitHub <- R6::R6Class(
     },
 
     # Retrieve only important info from repositories response
-    tailor_repos_response = function(repos_response) {
+    tailor_repos_response = function(repos_response, output = "table_full") {
       repos_list <- purrr::map(repos_response, function(repo) {
-        list(
-          "repo_id" = repo$id,
-          "repo_name" = repo$name,
-          "default_branch" = repo$default_branch,
-          "stars" = repo$stargazers_count,
-          "forks" = repo$forks_count,
-          "created_at" = gts_to_posixt(repo$created_at),
-          "last_activity_at" = if (!is.null(repo$pushed_at)) {
-            gts_to_posixt(repo$pushed_at)
-          } else {
-            gts_to_posixt(repo$created_at)
-          },
-          "languages" = repo$language,
-          "issues_open" = repo$issues_open,
-          "issues_closed" = repo$issues_closed,
-          "organization" = repo$owner$login,
-          "repo_url" = repo$html_url
-        )
+        if (output == "table_full") {
+          repo_data <- list(
+            "repo_id" = repo$id,
+            "repo_name" = repo$name,
+            "default_branch" = repo$default_branch,
+            "stars" = repo$stargazers_count,
+            "forks" = repo$forks_count,
+            "created_at" = gts_to_posixt(repo$created_at),
+            "last_activity_at" = if (!is.null(repo$pushed_at)) {
+              gts_to_posixt(repo$pushed_at)
+            } else {
+              gts_to_posixt(repo$created_at)
+            },
+            "languages" = repo$language,
+            "issues_open" = repo$issues_open,
+            "issues_closed" = repo$issues_closed,
+            "organization" = repo$owner$login,
+            "repo_url" = repo$html_url
+          )
+        }
+        if (output == "table_min") {
+          repo_data <- list(
+            "repo_id" = repo$id,
+            "repo_name" = repo$name,
+            "default_branch" = repo$default_branch,
+            "created_at" = gts_to_posixt(repo$created_at),
+            "organization" = repo$owner$login,
+            "repo_url" = repo$html_url
+          )
+        }
+        return(repo_data)
       })
       return(repos_list)
     },

diff --git a/R/EngineRestGitLab.R b/R/EngineRestGitLab.R
@@ -62,11 +62,11 @@ EngineRestGitLab <- R6::R6Class(
     #   filtering by language. For more information look here:
     #   https://gitlab.com/gitlab-org/gitlab/-/issues/340333
     get_repos_by_code = function(code,
-                                 org = NULL,
+                                 org      = NULL,
                                  filename = NULL,
-                                 in_path = FALSE,
-                                 raw_output = FALSE,
-                                 verbose = TRUE,
+                                 in_path  = FALSE,
+                                 output   = "table_full",
+                                 verbose  = TRUE,
                                  progress = TRUE) {
       search_response <- private$search_for_code(
         code = code,
@@ -75,37 +75,53 @@ EngineRestGitLab <- R6::R6Class(
         org = org,
         verbose = verbose
       )
-      if (raw_output) {
+      if (output == "raw") {
         search_output <- search_response
-      } else {
+      } else if (output == "table_full" || output == "table_min") {
         search_output <- search_response %>%
           private$map_search_into_repos(
             progress = progress
-          ) %>%
-          private$get_repos_languages(
-            progress = progress
           )
+        if (output == "table_full") {
+          search_output <- search_output %>%
+            private$get_repos_languages(
+              progress = progress
+            )
+        }
       }
       return(search_output)
     },
 
     # Retrieve only important info from repositories response
-    tailor_repos_response = function(repos_response) {
+    tailor_repos_response = function(repos_response, output = "table_full") {
       repos_list <- purrr::map(repos_response, function(project) {
-        list(
-          "repo_id" = project$id,
-          "repo_name" = project$name,
-          "default_branch" = project$default_branch,
-          "stars" = project$star_count,
-          "forks" = project$fork_count,
-          "created_at" = project$created_at,
-          "last_activity_at" = project$last_activity_at,
-          "languages" = paste0(project$languages, collapse = ", "),
-          "issues_open" = project$issues_open,
-          "issues_closed" = project$issues_closed,
-          "organization" = project$namespace$path,
-          "repo_url" = project$web_url
-        )
+        if (output == "table_full") {
+          repo_data <- list(
+            "repo_id" = project$id,
+            "repo_name" = project$name,
+            "default_branch" = project$default_branch,
+            "stars" = project$star_count,
+            "forks" = project$fork_count,
+            "created_at" = project$created_at,
+            "last_activity_at" = project$last_activity_at,
+            "languages" = paste0(project$languages, collapse = ", "),
+            "issues_open" = project$issues_open,
+            "issues_closed" = project$issues_closed,
+            "organization" = project$namespace$path,
+            "repo_url" = project$web_url
+          )
+        }
+        if (output == "table_min") {
+          repo_data <- list(
+            "repo_id" = project$id,
+            "repo_name" = project$name,
+            "default_branch" = project$default_branch,
+            "created_at" = project$created_at,
+            "organization" = project$namespace$path,
+            "repo_url" = project$web_url
+          )
+        }
+        return(repo_data)
       })
       return(repos_list)
     },