Unit files

This unit contains file- and pathname-oriented procedures. It uses the regex unit.

Pathname operations

absolute-pathname?

(absolute-pathname? PATHNAME) procedure

Returns #t if the string PATHNAME names an absolute pathname, and returns #f otherwise.

decompose-pathname

(decompose-pathname PATHNAME) procedure

Returns three values: the directory-, filename- and extension-components of the file named by the string PATHNAME. For any component that is not contained in PATHNAME, #f is returned.

make-pathname

make-absolute-pathname

(make-pathname DIRECTORY FILENAME [EXTENSION]) procedure
(make-absolute-pathname DIRECTORY FILENAME [EXTENSION]) procedure

Returns a string that names the file with the components DIRECTORY, FILENAME and (optionally) EXTENSION with SEPARATOR being the directory separation indicator (usually / on UNIX systems and \ on Windows, defaulting to whatever platform this is running on). DIRECTORY can be #f (meaning no directory component), a string or a list of strings. FILENAME and EXTENSION should be strings or #f. make-absolute-pathname returns always an absolute pathname.

pathname-directory

pathname-file

pathname-extension

(pathname-directory PATHNAME) procedure
(pathname-file PATHNAME) procedure
(pathname-extension PATHNAME) procedure

Accessors for the components of PATHNAME. If the pathname does not contain the accessed component, then #f is returned.

pathname-replace-directory

pathname-replace-file

pathname-replace-extension

(pathname-replace-directory PATHNAME DIRECTORY) procedure
(pathname-replace-file PATHNAME FILENAME) procedure
(pathname-replace-extension PATHNAME EXTENSION) procedure

Return a new pathname with the specified component of PATHNAME replaced by a new value.

pathname-strip-directory

pathname-strip-extension

(pathname-strip-directory PATHNAME) procedure
(pathname-strip-extension PATHNAME) procedure

Return a new pathname with the specified component of PATHNAME stripped.

normalize-pathname

(normalize-pathname PATHNAME [PLATFORM]) procedure

Performs a simple "normalization" on the PATHNAME, suitably for PLATFORM, which should be one of the symbols windows or unix and defaults to on whatever platform is currently in use. All relative path elements and duplicate separators are processed and removed. If NAME ends with a / or is empty, the appropriate slash is appended to the tail. Tilde ~ and variable $<name>/... expansion is also done.

No directories or files are actually tested for existence; this procedure only canonicalises path syntax.

directory-null?

(directory-null? DIRECTORY) procedure

Does the DIRECTORY consist only of path separators and the period?

DIRECTORY may be a string or a list of strings.

decompose-directory

(decompose-directory DIRECTORY) procedure

Returns 3 values: the base-origin, base-directory, and the directory-elements for the DIRECTORY.

base-origin
a string or #f. The drive, if any.
base-directory
a string or #f. A directory-separator when DIRECTORY is an absolute-pathname.
directory-elements
a list-of string or #f. The non-directory-separator bits.

DIRECTORY is a string.

  • On Windows (decompose-directory "c:foo/bar") => "c:" #f ("foo" "bar")

Temporary files and directories

create-temporary-file

(create-temporary-file [EXTENSION]) procedure

Creates an empty temporary file and returns its pathname. If EXTENSION is not given, then .tmp is used. If the environment variable TMPDIR, TEMP or TMP is set, then the pathname names a file in that directory. If none of the environment variables is given the location of the temporary file defaults to /tmp if it exists or the current-directory

create-temporary-directory

(create-temporary-directory) procedure

Creates an empty temporary directory and returns its pathname. If the environment variable TMPDIR, TEMP or TMP is set, then the temporary directory is created at that location.

Deleting a file without signalling an error

delete-file*

(delete-file* FILENAME) procedure

If the file FILENAME exists, it is deleted and #t is returned. If the file does not exist, nothing happens and #f is returned.

File move/copy

file-copy

(file-copy ORIGFILE NEWFILE #!optional CLOBBER BLOCKSIZE) procedure

Copies ORIGFILE (a string denoting some filename) to NEWFILE, BLOCKSIZE bytes at a time. BLOCKSIZE defaults to 1024, and must be a positive integer. Returns the number of bytes copied on success, or errors on failure. CLOBBER determines the behaviour of file-copy when NEWFILE is already extant. When set to #f (default), an error is signalled. When set to any other value, NEWFILE is overwritten. file-copy will work across filesystems and devices and is not platform-dependent.

file-move

(file-move ORIGFILE NEWFILE #!optional CLOBBER BLOCKSIZE) procedure

Moves ORIGFILE (a string denoting some filename) to NEWFILE, with the same semantics as file-copy, above. file-move is safe across filesystems and devices (unlike rename-file). It is possible for an error to be signalled despite partial success if NEWFILE could be created and fully written but removing ORIGFILE fails.


Previous: Unit ports

Next: Unit extras