Note: this vignette is pre-computed to utilize the Microsoft Graph login that is only available locally to me as a logged in SharePoint user.
Downloading and reading files from SharePoint
You can use download_sp_item() to download files or
folders from SharePoint:
docx_url <- "https://bmore.sharepoint.com/:w:/r/sites/MayorsOffice-DataGovernance/Policy%20Documents/Data%20Classification%20Standard.docx?d=w54a9ae7eaa894e94b6d6d14516f3aaa4&csf=1&web=1&e=ee7ZSX"
download_sp_item(
path = docx_url,
new_path = tempdir()
)
#> Loading Microsoft Graph login for default tenant
#>
ℹ Downloading SharePoint item to
#> '/var/folders/3f/50m42dx1…
#> Error: Path exists and overwrite is FALSE
#>
✖ Downloading SharePoint item to '/var/folders/3f/50m42dx1…For files on SharePoint, read_sharepoint() extends
download_sp_item() by downloading the selected item to a
temporary folder by default and, depending on the file extension, tries
to read the file using readr, readxl,
officer, or sf.
docx <- read_sharepoint(docx_url)
#> Loading Microsoft Graph login for default tenant
#>
ℹ Downloading SharePoint item to '/var/folders/3f/50m42dx1…
✔ Downloading SharePoint item to '/var/folders/3f/50m42dx1…
#>
ℹ Reading item with `officer::read_docx()`
✔ Reading item with `officer::read_docx()` [44ms]
docx
#> rdocx document with 62 element(s)
#>
#> * styles:
#> Normal heading 1
#> "paragraph" "paragraph"
#> heading 2 Default Paragraph Font
#> "paragraph" "character"
#> Normal Table No List
#> "table" "numbering"
#> Heading 1 Char Table Grid
#> "character" "table"
#> List Paragraph Heading 2 Char
#> "paragraph" "character"
#> Title Title Char
#> "paragraph" "character"
#> Normal (Web) header
#> "paragraph" "paragraph"
#> Header Char footer
#> "character" "paragraph"
#> Footer Char markedcontent
#> "character" "character"
#> TOC Heading toc 1
#> "paragraph" "paragraph"
#> toc 2 Hyperlink
#> "paragraph" "character"
#> toc 3 toc 4
#> "paragraph" "paragraph"
#> toc 5 toc 6
#> "paragraph" "paragraph"
#> toc 7 toc 8
#> "paragraph" "paragraph"
#> toc 9 Unresolved Mention
#> "paragraph" "character"
#> FollowedHyperlink Revision
#> "character" "paragraph"
#> Unresolved Mention1 annotation reference
#> "character" "character"
#> annotation text Comment Text Char
#> "paragraph" "character"
#> annotation subject Comment Subject Char
#> "paragraph" "character"
#> Balloon Text Balloon Text Char
#> "paragraph" "character"
#> footnote text Footnote Text Char
#> "paragraph" "character"
#> footnote reference
#> "character"
#>
#> * Content at cursor location:
#> level num_id text style_name content_type
#> 1 NA NA heading 1 paragraphWriting and uploading files to SharePoint
You can use upload_sp_item() to upload a local file to a
SharePoint folder or document library.
folder_url <- "https://bmore.sharepoint.com/:f:/r/sites/MayorsOffice-DataGovernance/Shared%20Documents/RStats?csf=1&web=1&e=S1XxVU"
upload_sp_item(
file = system.file("gpkg/nc.gpkg", package = "sf"),
dest = folder_url
)
#> Loading Microsoft Graph login for default tenant
#>
ℹ Uploading file 'nc.gpkg' to SharePoint drive
✔ File upload complete [1.9s] Using read_sharepoint(), we can confirm that the file
has been uploaded:
sp_drive <- get_sp_drive(folder_url)
#> Loading Microsoft Graph login for default tenant
nc <- read_sharepoint(
"RStats/nc.gpkg",
drive = sp_drive
)
#>
ℹ Downloading SharePoint item to
#> '/var/folders/3f/50m42dx1…
#>
✔ Downloading SharePoint item to '/var/folders/3f/50m42dx1…
#>
ℹ Reading item with `sf::read_sf()`
✔ Reading item with `sf::read_sf()` [27ms]
plot(nc["AREA"])
write_sharepoint() extends upload_sp_item()
by allowing you to pass an R object instead of a file path. Like
read_sharepoint() tries to guess the appropriate input
function, write_sharepoint() tries to guess the appropriate
output function based on the object class.
write_sharepoint(
mtcars,
file = "mtcars.csv",
dest = folder_url
)
#> Loading Microsoft Graph login for default tenant
#>
ℹ Uploading file 'mtcars.csv' to SharePoint drive
✔ File upload complete [970ms] To wrap up this example, we need to remove the uploaded files from SharePoint to keep a tidy shared file system.
delete_sp_item() supports the option to use a shared
item URL to select which file to remove but, in this case, it is easier
to set the drive argument along with a relative
filepath:
# Remove the file
delete_sp_item(
file.path("RStats", "nc.gpkg"),
drive = sp_drive,
confirm = FALSE
)
delete_sp_item(
file.path("RStats", "mtcars.csv"),
drive = sp_drive,
confirm = FALSE
)Listing files
If you do not know the URL or file path for an item on SharePoint you
can also use the directory info functions to list items typically using
the sp_dir_info() function. This function supports
recursive listings but this can be slow depending on the number of items
in the SharePoint library.
This last example is not computed but it shows how to list and remove empty nested directories left over from a failed manual import;
# List directories
dir_info <- sp_dir_info("<SharePoint Folder URL>", type = "directory", recurse = TRUE)
# Filter to empty directories and sort by depth
empty_dirs <- dir_info |>
filter(size == 0) |>
mutate(
path_depth = str_count(name, "/")
) |>
arrange(desc(path_depth))
# Get drive
drive <- get_sp_drive(
drive_name = "<SharePoint Document Library Name>",
site_url = "<SharePoint Site URL>"
)
# Delete empty directories and skip confirmation
walk(
empty_dirs[["id"]],
\(x) {
delete_sp_item(
item_id = x,
drive = drive,
confirm = FALSE
)
}
)Overall, the intent of this package is to maximize flexibility in how and where you read and write files from SharePoint. Suggestions are welcome so please share your own tips on working with the Microsoft SharePoint API and the Microsoft365R package.