What is the difference (if any) between path.normalize(your_path)
and path.resolve(your_path)
?
I know path.resolve(...)
can accept
path.normalize
gets rid of the extra .
, ..
, etc. in the path. path.resolve
resolves a path into an absolute path. Example (my current working directory was /Users/mtilley/src/testing
):
> path.normalize('../../src/../src/node')
'../../src/node'
> path.resolve('../../src/../src/node')
'/Users/mtilley/src/node'
In other words, path.normalize
is "What is the shortest path I can take that will take me to the same place as the input", while path.resolve
is "What is my destination if I take this path."
Note however that path.normalize()
is much more context-independent than path.resolve()
. Had path.normalize()
been context-dependent (i.e. if it had taken into consideration the current working directory), the result in the example above would've been ../node
, because that's the shortest path one could take from /Users/mtilley/src/testing
to /Users/mtilley/src/node
.
Ironically, this means that path.resolve()
produces a relative path in absolute terms (you could execute it anywhere, and it would produce the same result), whereas path.normalize()
produces an absolute path in relative terms (you must execute it in the path relative to which you want to calculate the absolute result).
From the docs:
Another way to think of resolve is as a sequence of cd commands in a shell.
Links to path.resolve and path.normalize in the documentation. I mostly don't want to just provide links in an answer but the Node.js docs are very decent.