# ShowDict.py
""" Compiles a dictionary of word frequencies
for the Shakespeare sonnet collection.
"""



def GetSonnets():
    """ Returns a list of strings L with the property that
    L[k] is the kth line in the file Sonnets.txt.
    """
    L = []
    with open('TheSonnets.txt',"r") as F:
        for s in F:
            # Add the next line in the file to L without the trailing "\n" (the
            # new line character), and without the trailing "\r" (the carriage
            # return character)
            L.append(s.rstrip("\n").rstrip("\r"))
    return L

def dePunc(s):
    """Returns a string that is the same as s except that all occurrences of the
    period, comma, semicolon, colon, question mark, and explanation point are
    replaced by the null string. Leading and trailing apostrophes are also removed.
     
    Example: dePunc('a.b,c;d:e?f!') returns 'abcdef'
    
    PreC: s is a string.
    """
    if s == '':
        return s
    # s is not the empty string, remove punctuation
    s = s.replace('.','')
    s = s.replace(',','')
    s = s.replace(';','')
    s = s.replace(':','')
    s = s.replace('?','')
    s = s.replace('!','')
    if s[0]=='\'':
        s = s[1:]
    if s[len(s)-1]=='\'':
        s = s[:len(s)-1]
    return s

def WordsInLine(s):
    """ Returns a list of strings, each of which
    is a word in t, where t is s with all punctuation
    removed and all its letters lower-cased.
    
    A word is delimited by blanks.
    
    PreC: s is a string
    """
    s = s.lower()
    s = dePunc(s)
    W = s.split()
    return W

def UpdateFreqD(D,s):
    """ If s is a key in D, then its value is incremented.
    Otherwise, s is added into D as a key and its value is
    set to 1.
    
    PreC: D is a dictionary with keys that are strings and values
    that are ints. s is a string.
    """

    if s in D:
        D[s] +=1
    else:
        D[s] = 1
        
def MakeFreqD(L):
    """ Returns a dictionary whose keys are words that occur
    in L and whose values are their frequency,
    PreC: L is a list of strings
    """
    D = dict()
    for s in L:
        W = WordsInLine(s)
        for w in W:
            UpdateFreqD(D,w)
    return D
          
def printFreq(D,ListOfWords):
    """ Prints each string in ListOfWords and its frquency in D.
      PreC: D is a dictionary with keys that are strings and values
    that are ints.
    """
    for w in ListOfWords:
        if w in D:
            f = D[w]
        else:
            f = 0
        print '%10s  %3d' %(w,f)

if __name__ == '__main__':
    # Read in the sonnets, line-by-line..
    L = GetSonnets()
    # Make the frequency dictionary
    D = MakeFreqD(L)
    # Lets display the frequencies of these words...
    MyWordList = (['love', 'sun','moon','sad','happy','thou','me','flowers',
                   'water','dude','rain','clouds','wonder','forever'] )
    printFreq(D,MyWordList)