25. Context managers¶
Context managers allow you to allocate and release resources precisely
when you want to. The most widely used example of context managers is
with statement. Suppose you have two related operations which
you’d like to execute as a pair, with a block of code in between.
Context managers allow you to do specifically that. For example:
with open('some_file', 'w') as opened_file: opened_file.write('Hola!')
The above code opens the file, writes some data to it and then closes it. If an error occurs while writing the data to the file, it tries to close it. The above code is equivalent to:
file = open('some_file', 'w') try: file.write('Hola!') finally: file.close()
While comparing it to the first example we can see that a lot of
boilerplate code is eliminated just by using
with. The main
advantage of using a
with statement is that it makes sure our file
is closed without paying attention to how the nested block exits.
A common use case of context managers is locking and unlocking resources and closing opened files (as I have already showed you).
Let’s see how we can implement our own Context Manager. This would allow us to understand exactly what’s going on behind the scenes.
25.1. Implementing Context Manager as a Class:¶
At the very least a context manager has an
__exit__ methods defined. Let’s make our own file opening Context
Manager and learn the basics.
class File(object): def __init__(self, file_name, method): self.file_obj = open(file_name, method) def __enter__(self): return self.file_obj def __exit__(self, type, value, traceback): self.file_obj.close()
Just by defining
__exit__ methods we can use it in
with statement. Let’s try:
with File('demo.txt', 'w') as opened_file: opened_file.write('Hola!')
__exit__ function accepts three arguments. They are required by
__exit__ method which is a part of a Context Manager class.
Let’s talk about what happens under-the-hood.
withstatement stores the
- It calls the
__enter__method opens the file and returns it.
- the opened file handle is passed to
- we write to the file using
withstatement calls the stored
__exit__method closes the file.
25.2. Handling exceptions¶
We did not talk about the
arguments of the
__exit__ method. Between the 4th and 6th step, if
an exception occurs, Python passes the type, value and traceback of the
exception to the
__exit__ method. It allows the
to decide how to close the file and if any further steps are required.
In our case we are not paying any attention to them.
What if our file object raises an exception? We might be trying to access a method on the file object which it does not supports. For instance:
with File('demo.txt', 'w') as opened_file: opened_file.undefined_function('Hola!')
Let’s list down the steps which are taken by the
with statement when
an error is encountered.
- It passes the type, value and traceback of the error to the
- It allows the
__exit__method to handle the exception.
__exit__returns True then the exception was gracefully handled.
- If anything else than True is returned by the
__exit__method then an exception is raised by the
In our case the
__exit__ method returns
None (when no return
statement is encountered then the method returns
None). Therefore, the
with statement raises the exception.
Traceback (most recent call last): File "<stdin>", line 2, in <module> AttributeError: 'file' object has no attribute 'undefined_function'
Let’s try handling the exception in the
class File(object): def __init__(self, file_name, method): self.file_obj = open(file_name, method) def __enter__(self): return self.file_obj def __exit__(self, type, value, traceback): print("Exception has been handled") self.file_obj.close() return True with File('demo.txt', 'w') as opened_file: opened_file.undefined_function() # Output: Exception has been handled
__exit__ method returned True, therefore no exception was raised
This is not the only way to implement context managers. There is another way and we will be looking at it in this next section.
25.3. Implementing a Context Manager as a Generator¶
We can also implement Context Managers using decorators and generators. Python has a contextlib module for this very purpose. Instead of a class, we can implement a Context Manager using a generator function. Let’s see a basic, useless example:
from contextlib import contextmanager @contextmanager def open_file(name): f = open(name, 'w') yield f f.close()
Okay! This way of implementing Context Managers appears to be more intuitive and easy. However, this method requires some knowledge about generators, yield, and decorators. In this example we have not caught any exceptions which might occur. It works in mostly the same way as the previous method.
Let’s dissect this method a little.
- Python encounters the
yieldkeyword. Due to this it creates a generator instead of a normal function.
- Due to the decoration, contextmanager is called with the function name (open_file) as it’s argument.
contextmanagerfunction returns the generator wrapped by the
GeneratorContextManageris assigned to the
open_filefunction. Therefore, when we later call
open_filefunction, we are actually calling the
So now that we know all this, we can use the newly generated Context Manager like this:
with open_file('some_file') as f: f.write('hola!')