A python decorator that adds a column to your pandas dataframe -- the MD5 hash of the specified column

md5_decorator.py

import pandas as pd
from hashlib import md5

def text_to_hash(text):
    return md5(text.encode("utf8")).hexdigest()

def add_hash(column_name="document"):
    """ 
    Decorator. Wraps a function that returns a dataframe, must have column_name in columns.
    
    Adds an additional column to the returned df: "hashid", the MD5 hash of column_name
    """
    def apply_hash(df_fn):
        def new_df_fn(*args, **kwargs):
            df = df_fn(*args, **kwargs)
            df.loc[:, "hashid"] = df.loc[:, column_name].apply(text_to_hash)
            return df
        return new_df_fn
    return apply_hash


# Test text_to_hash #
assert text_to_hash("hello") == text_to_hash("hello")
assert text_to_hash("hey hows it going") != text_to_hash("hey hey hows it going")  # extra hey
assert text_to_hash("hey hows it going") != text_to_hash("hey how's it going")  # hows != how's


# Test add_hash decorator #
@add_hash(column_name="document")
def test_df_fn(text_list):
    return pd.DataFrame({"document": text_list})
    
test_df = test_df_fn(["hi", "hello"])
assert "hashid" in test_df.columns
test_df