Working with IrodsPath

The Python API uses IrodsPath objects to represent locations inside an iRODS server.
They behave similarly to pathlib paths, but operate on the iRODS namespace.
All interactions with collections and data objects begin with an IrodsPath.

An IrodsPath always belongs to a Session and resolves either absolute or relative iRODS paths.

Creating an IrodsPath

Absolute paths must start with a slash:

from ibridges.path import IrodsPath

p = IrodsPath(session, "/tempZone/home/christine/demo")
print(p)

The operator / extends paths:

sub = p / "experiment1"
print(sub)

But let us first explore the home and the current working directory a bit more.

Home and current working directory

The Session defines two important locations:

• session.home
  
• session.cwd

The home is your default working directory in iRODS.
It can be set in three ways:

1. In irods_environment.json
   
2. When creating the Session
   
3. By assigning session.home manually

If not set, it defaults to /zone/home/username.

The cwd is optional and defaults to the home.
The dot symbol refers to the cwd:

IrodsPath(session, ".")

The tilde symbol refers to the home:

IrodsPath(session, "~")

You can set your demo collection as current working directory:

session.cwd = IrodsPath(session, "~", "demo")
print("cwd:", session.cwd)
print("home:", session.home)

Relative paths are resolved against the session’s cwd:

p = IrodsPath(session, "subcoll")
print(p)

You can create instances of IrodsPath, however, that does not mean that they are present on the iRODS server.

Let’s have a look at how to check whether a path exists and how to create a new collection.

Checking existence

print(p)
print("Exists:", p.exists())
print("Is a collection:", p.collection_exists())
print("Is a data object:", p.dataobject_exists())

Creating collections

Creating an IrodsPath does not create anything on the server.
To create a collection:

coll = p.create_collection()

This creates the entire path if needed.

print("Exists:", p.exists())
print("Is a collection:", p.collection_exists())
print("Is a data object:", p.dataobject_exists())

And it returns the collection (not the path). We will investigate that later when we work with data.

Renaming or moving

print(p)
q = p.rename("new_name")
print(q)

The old variable p still contains our old path. But it does not exist any longer:

p.exists()

The rename function returns the new IrodsPath which does exist:

q.exists()

You can also use the rename function to move a collection or data object to a different location:

target = IrodsPath(session, "/tempZone/home/christine/archive/my_experiment")
new_p = q.rename(target)

print(new_p)
print("Is a collection:", new_p.collection_exists())

We have moved our collection new_name to a completely new branch in the tree of iRODS.

Removing collections or data objects

Let us remove the whole new archive collection. We will use the parent of the path to navigate one level up and then apply the remove function to this collection:

new_p.parent.remove()
new_p.collection_exists()
new_p.parent.collection_exists()

The IrodsPath object remains, but the server location is deleted.

Inspecting properties

p.absolute()
p.parts
p.parent
p.size

These help you understand the structure and contents of the path.

Exercises (15 minutes)

Exercise 1: Basic path handling

• Create a session using interactive_auth
  
• Construct an IrodsPath pointing to your home using “~”
  
• Verify that the path exists
  
• Print its parts and parent

Exercise 2: Creating and navigating collections

• Verify the existence of session.home / "demo" and create it if necessary
  
• Extend the path with “experiment1” using /
  
• Create the subcollection

Exercise 3: Renaming and moving

• Rename experiment1 to experimentA
  
• Move experimentA into a new collection session.home / “archive”
  
• Verify the new location exists

Exercise 4: Removing collections

• Remove the archive/experimentA collection