8

What is the difference between absolute / canonical / relative paths?

I guess that absolute starts with a /. Can it contain /../?

Is a canonical path different from an absolute path in that it can't contain /../?

Is foo a relative path?

Googling posix canonical path didn't turn up much... What are the POSIX references for these terms?

Tom Hale
  • 30,455
  • Seems that the stackoverflow question/answer could help you https://stackoverflow.com/questions/1099300/whats-the-difference-between-getpath-getabsolutepath-and-getcanonicalpath/1099370 – igiannak Oct 16 '17 at 13:23
  • See: https://stackoverflow.com/questions/12100299/whats-a-canonical-path – JRFerguson Oct 16 '17 at 13:26

1 Answers1

10

POSIX defines “absolute pathname” as follows:

A pathname beginning with a single or more than two <slash> characters

and “relative pathname” as follows:

A pathname not beginning with a <slash> character.

That’s all there is to it for relative and absolute paths.

Canonical paths aren’t defined in POSIX, but the term usually refers to comparable paths, i.e. if you take two paths to a file system object, and convert them to canonical form, the result should be identical if and only if the two file system objects are identical. This involves removing “..” as you mention, but it also means resolving symbolic links; so a canonical path could be defined as

A pathname whose components are all real directories or files, excluding “.” and “..”, and whose slashes are not repeated

In POSIX terms, a canonical pathname is effectively a resolved pathname (as long as you accept that canonical pathnames can only be determined for file system objects which exist).

Note that this only works because hard-linked directories aren’t allowed...

So to answer your questions:

  • an absolute path can contain /../;
  • a canonical path can not contain /../, nor can it contain /./, // (except arguably in first position), or symbolic links;
  • foo is a relative path.

(Pedantically, they are all pathnames, not just paths.)

Stephen Kitt
  • 434,908
  • Any // or /// or any count //…// should be converted to (interpreted as) one /. In that sense, any // at the middle of (any) path is valid and permissible. –  Oct 16 '17 at 14:19
  • @Arrow The answer doesn't say that // in a path is not valid or permissible. It only states that they are not OK in a canonical path, and I would agree :) – marcelm Oct 16 '17 at 15:40
  • @marcelm After any repeated / (in the middle of the path) is converted to one / (if any exist), the path strings could be compared, yes. As such, I agree now. –  Oct 16 '17 at 16:15
  • 1
    It may be worth mentioning paths starting with ~. Shells will often expand such paths to absolute paths pointing at users' home directories. – John Kugelman Oct 16 '17 at 19:17
  • You can hard-link files, though, so I don't think you can state "the result should be identical if and only if the two file system objects are identical" as a single file can have more than one canonical/absolute path. – Andrew Henle Oct 17 '17 at 12:53
  • 1
    @Andrew hence the “if and only if”, and not mentioning inodes: you can delete one directory entry without affecting any of the others pointing to the same inode, so arguably they’re not the same file system objects. But I agree hard links do make the definition somewhat unsatisfactory. – Stephen Kitt Oct 17 '17 at 12:59