Course of Raku / Objects, I/O, and exceptions / Input and output

Deleting files and directories

Just as spurt and mkdir create things on disk, two routines remove them.

The unlink routine deletes a file:

spurt 'temp.txt', 'scratch data';
say 'temp.txt'.IO.e; # True

unlink 'temp.txt';
say 'temp.txt'.IO.e; # False

unlink deletes the named file. Like the other file-system routines, it also has a method form on a path object — 'temp.txt'.IO.unlink.

Deletion is immediate and permanent: there is no wastebasket to recover from. Note that after a file is gone, you must test for it with .e (existence) rather than .f, because the type tests .f and .d throw when the path is not there.

The rmdir routine removes a directory, but only an empty one:

mkdir 'reports';
rmdir 'reports';
say 'reports'.IO.e; # False

If the directory still contains something, rmdir removes nothing and leaves it in place. So to delete a directory that holds files, you empty it first and then remove it:

mkdir 'reports';
spurt 'reports/jan.txt', 'January';
spurt 'reports/feb.txt', 'February';

.unlink for 'reports'.IO.dir;   # delete every entry
rmdir 'reports';                # now the directory is empty
say 'reports'.IO.e;             # False

'reports'.IO.dir lists the directory’s entries (the previous topic covered it), and calling .unlink on each removes them. Once the directory is empty, rmdir can take it away. (This flat loop assumes the directory holds only files; a tree with nested sub-directories would need to recurse into them first.)

Course navigation

← Solution: File or directory?   |   Delete a file →