'''
    DS2000
    Spring 2022
    Sentiment Analysis algorithm

    In this version we make a list of pos words, and a list of
    neg words. We read all the comments in from a reddit file
    (different than the reddit flie on HW6, we have just one 
     comment per line). 

   Then, we pick one comment at random and clean it up (remove
    punctuation and numbers, and turn everything lowercase) and 
    apply our sentiment analysis algorithm ----- +1 for every
    positive word, -1 for every negative word, divided by
    number of words in the comment.
'''

import random

REDDIT = "reddit.txt"

POSITIVE = ["good", "happy", "relieved", "relief",
            "glad", "finally", "normal", "excited", 
            "proud", "well", "healthy", "lol", "great"]

NEGATIVE = ["bad", "angry", "frustrated", "stressed",
            "stress", "stupid", "scared", "scary", "scaring",
            "hate", "hated", "annoying", "annoyed",
            "tired", "disappointed", "lol", "afraid", "lmao"]

def sentiment_score(comment, pos, neg):
    ''' Function: sentiment_score
        Parameters: comment (a string), list of pos words,
                    list of neg words
        Returns: sentiment score (float between -1 and +1)
    '''
    score = 0
    words = comment.split()
    for word in words:
        if word in pos:
            score += 1
        if word in neg:
            score -= 1
    return score / len(words)
    
def read_comments(filename):
    ''' Function: read_comments
        Parameter: filename, string
        Returns: list of strings, one per line in the file
    '''
    comments = []
    with open(filename, "r") as infile:
        while True:
            comment = infile.readline()
            if not comment:
                break
            comments.append(comment)
    return comments
    

def clean_string(input_st):
    ''' Function: clean_string
        Parameter: one string
        Returns: new version of the string, cleaned up
    '''
    output_st = ""
    for letter in input_st:
        if letter.isalpha() or letter == " ":
            output_st += letter.lower()
    return output_st

def main():
    # string = input("Enter a string...\n")
    # cleaned = clean_string(string)
    # print("Cleaned version:", cleaned)
    
    # Step one: read data from the file, one comment per
    # element in a list
    comments = read_comments(REDDIT)
    
    # Step two: computation --- clean up one random comment
    # to remove puncutation/numbers and make it all lowercase
    comment = random.choice(comments)
    clean = clean_string(comment)
    score = sentiment_score(clean, POSITIVE, NEGATIVE)
    
    # Step three: communication --- print the cleaned
    # up comment and its score
    print(clean)
    print("Sentiment score:", score)
    
    
main()