Files stored on OSF can be downloaded locally by passing an osf_tbl_file
that contains the files and folders of interest. Use path to specify
where the files should be downloaded, otherwise they are downloaded to
your working directory by default.
osf_download(
x,
path = NULL,
recurse = FALSE,
conflicts = "error",
verbose = FALSE,
progress = FALSE
)The same osf_tbl_file passed to x with a new column,
"local_path", containing paths to the local files.
An osf_tbl_file containing a single file or directory.
Path pointing to a local directory where the downloaded files will be saved. Default is to use the current working directory.
Applies only to OSF directories. If TRUE, a directory is
fully recursed and all nested files and subdirectories are downloaded.
Alternatively, a positive number will determine the number of levels
to recurse.
This determines what happens when a file with the same name exists at the specified destination. Can be one of the following:
"error" (the default): throw an error and abort the file transfer operation.
"skip": skip the conflicting file(s) and continue transferring the
remaining files.
"overwrite": replace the existing file with the transferred copy.
Logical, indicating whether to print informative messages
about interactions with the OSF API (default FALSE).
Logical, if TRUE progress bars are displayed for each file
transfer. Mainly useful for transferring large files. For tracking lots of
small files, setting verbose = TRUE is more informative.
Directories are always downloaded from OSF as zip files that contain its
entire contents. The logic for handling conflicts and recursion is
implemented locally, acting on these files in a temporary location and
copying them to path as needed. This creates a gotcha if you're
downloading directories with large files and assuming that setting conflicts = "skip" and/or limiting recursion will reduce the number of files you're
downloading. In such a case, a better strategy would be to use
osf_ls_files() to list the contents of the directory and pass that output
to osf_download().
While osf_download() and osf_upload() can be used to conveniently shuttle
files back and forth between OSF and your local machine, it's important to
note that they are not file synchronization functions. In contrast to
something like rsync,
osf_download()/osf_upload() do not take into account a file's contents or
modification time. Whether you're uploading or downloading, if overwrite = TRUE, osfr will overwrite an existing file regardless of whether the
existing file is the more recent copy. You have been warned.
osf_upload() for uploading files to OSF.
osf_ls_files() for listing files and directories on OSF.
if (FALSE) {
# download a single file
analysis_plan <- osf_retrieve_file("2ryha") %>%
osf_download()
# verify the file was downloaded locally
file.exists(analysis_plan$local_path)
}
Run the code above in your browser using DataCamp Workspace