base (version 3.2.1)

files: File Manipulation

Description

These functions provide a low-level interface to the computer's file system.

Usage

file.create(..., showWarnings = TRUE) file.exists(...) file.remove(...) file.rename(from, to) file.append(file1, file2) file.copy(from, to, overwrite = recursive, recursive = FALSE, copy.mode = TRUE, copy.date = FALSE) file.symlink(from, to) file.link(from, to) windows Sys.junction(from, to)

Arguments

..., file1, file2
character vectors, containing file names or paths.
from, to
character vectors, containing file names or paths. For file.copy and file.symlink windows and Sys.junction to can alternatively be the path to a single existing directory.
overwrite
logical; should existing destination files be overwritten?
showWarnings
logical; should the warnings on failure be shown?
recursive
logical. If to is a directory, should directories in from be copied (and their contents)? (Like cp -R on POSIX OSes.)
copy.mode
logical: should file permission bits be copied where possible?
copy.date
logical: should file dates be preserved where possible? See Sys.setFileTime.

Value

These functions return a logical vector indicating which operation succeeded for each of the files attempted. Using a missing value for a file or path name will always be regarded as a failure.If showWarnings = TRUE, file.create will give a warning for an unexpected failure.

Case-insensitive file systems

Case-insensitive file systems are the norm on Windows and OS X, but can be found on all OSes (for example a FAT-formatted USB drive is probably case-insensitive). These functions will most likely match existing files regardless of case on such file systems: however this is an OS function and it is possible that file names might be mapped to upper or lower case.

Details

The ... arguments are concatenated to form one character string: you can specify the files separately or as one vector. All of these functions expand path names: see path.expand.

file.create creates files with the given names if they do not already exist and truncates them if they do. They are created with the maximal read/write permissions allowed by the ‘umask’ setting (where relevant). By default a warning is given (with the reason) if the operation fails.

file.exists returns a logical vector indicating whether the files named by its argument exist. (Here ‘exists’ is in the sense of the system's stat call: a file will be reported as existing only if you have the permissions needed by stat. Existence can also be checked by file.access, which might use different permissions and so obtain a different result. Note that the existence of a file does not imply that it is readable: for that use file.access.) What constitutes a ‘file’ is system-dependent, but should include directories. (However, directory names must not include a trailing backslash or slash on Windows.) Note that if the file is a symbolic link on a Unix-alike, the result indicates if the link points to an actual file, not just if the link exists. Lastly, note the different function exists which checks for existence of R objects.

file.remove attempts to remove the files named in its argument. On most Unix platforms ‘file’ includes empty directories, symbolic links, fifos and sockets. On Windows, ‘file’ means a regular file and not, say, an empty directory.

file.rename attempts to rename files (and from and to must be of the same length). Where file permissions allow this will overwrite an existing element of to. This is subject to the limitations of the OS's corresponding system call (see something like man 2 rename on a Unix-alike): in particular in the interpretation of ‘file’: most platforms will not rename files across file systems. (On Windows, file.rename can move files but not directories between volumes.) On platforms which allow directories to be renamed, typically neither or both of from and to must a directory, and if to exists it must be an empty directory.

file.append attempts to append the files named by its second argument to those named by its first. The R subscript recycling rule is used to align names given in vectors of different lengths.

file.copy works in a similar way to file.append but with the arguments in the natural order for copying. Copying to existing destination files is skipped unless overwrite = TRUE. The to argument can specify a single existing directory. If copy.mode = TRUE file read/write/execute permissions are copied where possible, restricted by ‘umask’. (On Windows this applies only to files.) Other security attributes such as ACLs are not copied. On a POSIX filesystem the targets of symbolic links will be copied rather than the links themselves, and hard links are copied separately.

file.symlink and file.link make symbolic and hard links on those file systems which support them. For file.symlink the to argument can specify a single existing directory. (Unix and OS X native filesystems support both. Windows has hard links to files on NTFS file systems and concepts related to symbolic links on recent versions: see the section below on the Windows version of this help page. What happens on a FAT or SMB-mounted file system is OS-specific.)

See Also

file.info, file.access, file.path, file.show, list.files, unlink, basename, path.expand.

dir.create.

Sys.glob to expand wildcards in file specifications.

file_test, Sys.readlink.

http://en.wikipedia.org/wiki/Hard_link and http://en.wikipedia.org/wiki/Symbolic_link for the concepts of links and their limitations.

Examples

Run this code
cat("file A\n", file = "A")
cat("file B\n", file = "B")
file.append("A", "B")
file.create("A")
file.append("A", rep("B", 10))
if(interactive()) file.show("A")
file.copy("A", "C")
dir.create("tmp")
file.copy(c("A", "B"), "tmp")
list.files("tmp")
 unix
setwd("tmp")
file.remove("B")
file.symlink(file.path("..", c("A", "B")), ".")
setwd("..")
unlink("tmp", recursive = TRUE)
file.remove("A", "B", "C")

Run the code above in your browser using DataCamp Workspace