From 55a5b339cc7374d1e2deb649ecee1b5746323089 Mon Sep 17 00:00:00 2001
From: Tan Ho <tan@tanho.ca>
Date: Sun, 29 May 2022 12:57:43 -0400
Subject: [PATCH] 0.0.5 - granular file type control in nflverse_save

---
 DESCRIPTION             |  4 ++--
 NAMESPACE               |  1 +
 NEWS.md                 |  4 ++++
 R/upload.R              | 36 ++++++++++++++++++++++--------------
 man/nflverse_archive.Rd | 14 ++++++++++++++
 man/nflverse_save.Rd    |  4 +++-
 6 files changed, 46 insertions(+), 17 deletions(-)
 create mode 100644 man/nflverse_archive.Rd

diff --git a/DESCRIPTION b/DESCRIPTION
index 5482ae0..00cfb83 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: nflversedata
 Title: nflverse data storage functions
-Version: 0.0.4
+Version: 0.0.5
 Authors@R: 
     person("Tan", "Ho", , "tan@tanho.ca", role = c("aut", "cre"),
            comment = c(ORCID = "0000-0001-8388-5155"))
@@ -28,4 +28,4 @@ Remotes:
 Encoding: UTF-8
 LazyData: true
 Roxygen: list(markdown = TRUE)
-RoxygenNote: 7.1.2
+RoxygenNote: 7.2.0
diff --git a/NAMESPACE b/NAMESPACE
index e44a47f..439be1e 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -1,4 +1,5 @@
 # Generated by roxygen2: do not edit by hand
 
+export(nflverse_archive)
 export(nflverse_save)
 export(nflverse_upload)
diff --git a/NEWS.md b/NEWS.md
index 0aed5db..0357f6a 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,3 +1,7 @@
+# nflversedata 0.0.5 
+
+* Added file_types argument which takes a vector of file types and defaults to `c("csv","rds","qs","parquet")`
+
 # nflversedata 0.0.4
 
 * Added `nflverse_archive()` which automates archiving the rds version of every nflverse-data release asset and pushes it to https://github.com/nflverse/nflverse-data-archives
diff --git a/R/upload.R b/R/upload.R
index 8956cab..94404a7 100644
--- a/R/upload.R
+++ b/R/upload.R
@@ -40,6 +40,7 @@ update_release_timestamp <- function(tag){
 #' @param nflverse_type metadata: name/information to add to data
 #' @param release_tag name of release to upload to
 #' @param .token a GitHub token, defaults to gh::gh_token()
+#' @param file_types one or more of c("rds","csv","parquet","qs","csv.gz")
 #'
 #' @export
 nflverse_save <- function(data_frame,
@@ -47,8 +48,8 @@ nflverse_save <- function(data_frame,
                           nflverse_type,
                           release_tag,
                           .token = gh::gh_token(),
-                          include_gz = FALSE
-                          ){
+                          file_types = c("rds","csv","parquet","qs")
+){
 
   stopifnot(
     is.data.frame(data_frame),
@@ -56,29 +57,36 @@ nflverse_save <- function(data_frame,
     is.character(nflverse_type),
     is.character(release_tag),
     is.character(.token),
+    is.character(file_types),
     length(file_name) == 1,
     length(nflverse_type) == 1,
     length(release_tag) == 1,
-    length(.token) == 1
+    length(.token) == 1,
+    length(file_types) >= 1
   )
 
   attr(data_frame,"nflverse_type") <- nflverse_type
   attr(data_frame,"nflverse_timestamp") <- Sys.time()
 
   temp_dir <- tempdir(check = TRUE)
+  ft <- rlang::arg_match(file_types,
+                         values = c("rds","csv","csv.gz","parquet","qs"),
+                         multiple = TRUE)
 
-  saveRDS(data_frame,file.path(temp_dir,paste0(file_name,".rds")))
-  data.table::fwrite(data_frame, file.path(temp_dir,paste0(file_name,".csv")))
-  if(include_gz) data.table::fwrite(data_frame, file.path(temp_dir,paste0(file_name,".csv.gz")))
-  arrow::write_parquet(data_frame, file.path(temp_dir, paste0(file_name,".parquet")))
-  qs::qsave(data_frame,
-            file.path(temp_dir,paste0(file_name,".qs")),
-            preset = "custom",
-            algorithm = "zstd_stream",
-            compress_level = 22,
-            shuffle_control = 15)
+  if("rds" %in% ft) saveRDS(data_frame,file.path(temp_dir,paste0(file_name,".rds")))
+  if("csv" %in% ft) data.table::fwrite(data_frame, file.path(temp_dir,paste0(file_name,".csv")))
+  if("csv.gz" %in% ft) data.table::fwrite(data_frame, file.path(temp_dir,paste0(file_name,".csv.gz")))
+  if("parquet" %in% ft) arrow::write_parquet(data_frame, file.path(temp_dir, paste0(file_name,".parquet")))
+  if("qs" %in% ft){
+    qs::qsave(data_frame,
+              file.path(temp_dir,paste0(file_name,".qs")),
+              preset = "custom",
+              algorithm = "zstd_stream",
+              compress_level = 22,
+              shuffle_control = 15)
+  }
 
-  .filetypes <- if(include_gz) c(".rds",".csv",".csv.gz",".parquet",".qs") else c(".rds",".csv",".parquet",".qs")
+  .filetypes <- paste0(".",ft)
 
   .file_names <- file.path(temp_dir, paste0(file_name,.filetypes))
 
diff --git a/man/nflverse_archive.Rd b/man/nflverse_archive.Rd
new file mode 100644
index 0000000..1c31399
--- /dev/null
+++ b/man/nflverse_archive.Rd
@@ -0,0 +1,14 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/archive.R
+\name{nflverse_archive}
+\alias{nflverse_archive}
+\title{Archive nflverse}
+\usage{
+nflverse_archive(release_name)
+}
+\arguments{
+\item{release_name}{}
+}
+\description{
+Archive nflverse
+}
diff --git a/man/nflverse_save.Rd b/man/nflverse_save.Rd
index 9d14117..2b82bfc 100644
--- a/man/nflverse_save.Rd
+++ b/man/nflverse_save.Rd
@@ -10,7 +10,7 @@ nflverse_save(
   nflverse_type,
   release_tag,
   .token = gh::gh_token(),
-  include_gz = FALSE
+  file_types = c("rds", "csv", "parquet", "qs")
 )
 }
 \arguments{
@@ -23,6 +23,8 @@ nflverse_save(
 \item{release_tag}{name of release to upload to}
 
 \item{.token}{a GitHub token, defaults to gh::gh_token()}
+
+\item{file_types}{one or more of c("rds","csv","parquet","qs","csv.gz")}
 }
 \description{
 This functions attaches nflverse attributes like type and timestamp, saves