Parsing in Python

// under python

In Python, everything is an object. Even the built-in types themselves (i.e. int, str, dict, list, etc.) are object's of type type. You can use the built-in type function to get an object's type.

type(int) # Returns type. So int is an object of type type.
isinstance(int, type) # Returns True
isinstance(int, object) # Returns True
type(type) # Returns type. So type is an object of type type. Truth.
type(object) # Returns type. So object is an object of type type. Mind blown.

Since everything in Python is an object, you can treat them as such. This means you can store the types themselves in a list.

my_types = [int, float, str, dict]

It turns out this is really useful when writing scripts to parse a text file. Let's say I'm writing a script to parse the CSV file below. (In reality, I should just use the pandas library.)

Bob, Jones, 26, 184.2
Thomas, Apple, 19, 155.9
Jake, Wilson, 31, 172.2

The parser script can store the types of each field in a list, and utilize it to convert each field value to its appropriate type with a simple generator expression:

field_names = ['first', 'last', 'age', 'weight']
field_types = [str, str, int, float]

def parse_line(l):
    """
    Parse a single line and return a dictionary.
    """
    field_values = l.split(',')
    return dict((fn, ft(fv)) for fn,ft,fv in zip(field_names, field_types, field_values))

def parse(filename):
    """
    Parse all of the lines in the file.
    """
    return [parse_line(l) for l in open(filename)]

To summarize, storing types in a list is really useful for writing concise scripts to parse text.

Go Top