File I/O and Data Persistence

A Python object normally exists in your program's memory while the program is running. The pickle module provides a way to convert that object into a byte stream for storage and later retrieve it back into memory. This creates a simple persistence lifecycle: an object is created, serialized into bytes, written to a file, read later, and reconstructed as a Python object.

The file is independent of the variable that originally referred to the object. Once pickle.dump() has stored the object in the file, deleting the variable removes that variable from memory but does not remove the stored data from the file.

To save an object, call pickle.dump(object, file). The first argument is the Python object to serialize. The second argument is an opened file. For pickle operations, the file must be opened in wb mode: binary write mode. The dump operation converts the object into a byte stream and writes that stream to the file.

To retrieve the stored object, open the file in rb mode: binary read mode. Then call pickle.load(file). The load operation reads the stored byte stream and returns the object back into memory. The returned object can be assigned to a new variable such as storedlist.

The important distinction is between a variable in memory and the data stored on disk. del shoplist destroys the reference to the list in the program, but it does not erase the pickled file. Loading the file later creates access to the stored data through storedlist.

These are different stages of the same persistence process. The Python object is the in-memory value your program works with. The byte stream is the serialized form produced for storage. The file holds that stored byte stream. pickle.load() reverses the storage direction by retrieving the data and returning it as an object in memory.