"""
John Rachlin
DS 2000: Intro to Programming with Data


Date: Thu Oct 27 19:50:33 2022

File: climb.py
    
Description: Analyze "The Hill We Climb"
    
    
"""

import dataproc as dp

def count_words(words):
    """ Convert a list of words to a word count dictionary """
    wordcount = {}
    for word in words:
        if word in wordcount:
            wordcount[word] += 1
        else:
            wordcount[word] = 1
    return wordcount


def main():
    text = dp.read_text('the_hill_we_climb.txt')
    #print(text, len(text))
    
    words = dp.text_to_words(text)
    #print(words)
    
    # Find 10 most common words
    wc = count_words(words)
    kv = list(wc.items())
    kv.sort(key=lambda t: t[1], reverse=True)
    print(kv[:10])


    # Does it surprise you that the most common word is 'we'?
    # Why might that be for a poem read at a presidential inaugeral?    

if __name__ == '__main__':
    main()