"""

John Rachlin
DS 2000: Intro to Programming with Data

Filename: climb.py
    
Description: Count the frequency of each word in the poem: 
    "The Hill We Climb" by Amanda Gorman

    
"""

def clean_word(word):
    """ Clean a word, converting to lowercase
    and removing punctuation
    
    Parameters: word - A word to be cleaned
    Return:     cleaned up word
    """
    
    punctuation = "',?;:!."
    for c in punctuation:
        word = word.replace(c,'')
    word = word.lower()
    return word


def extract_words(filename):
    """ Extract a list of words. 
    Parameters: filename - the name of a file to be read
    Return:     a list of words """
    
    wordlist = []
    with open(filename, 'r') as file:
        for line in file:
            words = line.strip().split()
            for word in words:
                wordlist.append(clean_word(word))
                
    return wordlist
            
            
def count_words(words):
    """ Convert a list of words to a dictionary wordcount """
    D = {}
    for w in words:
        if w in D:
            D[w] += 1
        else:
            D[w] = 1
    return D


def main():
    
    # Read file and extract a list of words
    filename = "the_hill_we_climb.txt"
    words = extract_words(filename)
    
    # Count occurences of each word
    wc = count_words(words)
    
    # Output results of words occuring at least 5 times
    for word, count in wc.items():
        if count>=5: 
            print(word,":",count)

    # Sorting by count - this is advanced
    kv = list(wc.items())
    kv.sort(key=lambda t: t[1], reverse=True)
    print(kv[:10])
    

if __name__ == '__main__':
    main()