import sys
import string
import re
file = open ( "corpora.txt", "r" )
text = file.read ( )
file.close ( )

word_freq ={ }

word_list = string.split ( text )

for word in word_list:
    count = word_freq.get ( string.lower ( word ), 0 )
    word_freq[string. lower ( word )] = count + 1

keys = word_freq.keys ( )
keys.sort ( )
i=0
while i<10:
 for word in keys:
     
     print word, word_freq[word]
     i=1+1

# count words in a text and show the first ten items
# by decreasing frequency
 
# sample text for testing
text = """\
My name is Fred Flintstone and I am a famous TV
star.  I have as much authority as the Pope, I
just don't have as many people who believe it.
"""
 
word_freq = {}
 
word_list = text.split()
 
for word in word_list:
    # word all lower case
    word = word.lower()
    # strip any trailing period or comma
    word = word.rstrip('.,')
    # build the dictionary
    count = word_freq.get(word, 0)
    word_freq[word] = count + 1
 
# create a list of (freq, word) tuples
freq_list = [(freq, word) for word, freq in word_freq.items()]
 
# sort the list by the first element in each tuple (default)
freq_list.sort(reverse=True)
 
for n, tup in enumerate(freq_list):
    # print the first ten items
    if n < 10:
        freq, word = tup
        print freq, word
        # or
        #print word, freq
 
"""
my output -->
3 i
3 as
2 have
1 who
1 tv
1 the
1 star
1 pope
1 people
1 name
"""

# a look at the Tkinter Text widget
# use ctrl+c to copy, ctrl+x to cut selected text, 
# ctrl+v to paste, and ctrl+/ to select all
 
import Tkinter as tk
 
def get_text():
    # get text widget contents between start_index and end_index
    # start_index = "%d.%d" % (line, column)  here "1.0"
    # line starts with 1 and column with 0
    # here end_index = tk.END    
    # set the label text to the typed-in text
    v1.set(text1.get(1.0, tk.END))
    # clear the text
    text1.delete(1.0, tk.END)      
    text1.insert(tk.INSERT, ' new text')
    text1.insert(tk.INSERT, '\n and more text')
 
# this sets the window title caption too
# without the leading space Text will be text!?
root = tk.Tk(className = " Text, Button, Label ...")
 
# text entry field, width=width chars, height=lines text
text1 = tk.Text(root, width=50, height=2, bg='yellow')
text1.pack()
 
# function listed in command will be executed on button click
button1 = tk.Button(root, text='get the text', command=get_text)
button1.pack(pady=5)
 
# define a variable to hold the label text
v1 = tk.StringVar()
 
# label text will always be the textvariable's value
# width/height in char size
label1 = tk.Label(root, textvariable=v1, width=50, height=2) 
label1.pack(pady=5)
 
# do some caculation and format result
pi_approx = 355/113.0
str1 = "%.4f" % (pi_approx)    # 3.1416
# show result in text widget
text1.insert(tk.INSERT, str1)
 
# start cursor in text1
text1.focus()
 
root.mainloop()

# a look at the Tkinter Text widget

# use ctrl+c to copy, ctrl+x to cut selected text,

# ctrl+v to paste, and ctrl+/ to select all
import Tkinter as tk


def most_frequant_word():

      # count words in a text and show the first ten items
    # by decreasing frequency
     
    # sample text for testing

    import sys
    import string
    import re
    v1.set(text1.get(1.0, tk.END))
    text1.delete(1.0, tk.END)
    file = open ("arb.txt", "r")
    text = file.read ( )
    file.close ( )
     
    word_freq = {}
     
    word_list = text.split()
     
    for word in word_list:
        # word all lower case
        word = word.lower()
        # strip any trailing period or comma
        word = word.rstrip('.,/"-_;\[]()')
        # build the dictionary
        count = word_freq.get(word, 0)
        word_freq[word] = count + 1
     
    # create a list of (freq, word) tuples
    freq_list = [(freq, word) for word, freq in word_freq.items()]
     
    # sort the list by the first element in each tuple (default)
    freq_list.sort(reverse=True)
     
    for n, tup in enumerate(freq_list):
        # print the first ten items
        if n < 10:
            text1.insert(tk.INSERT, freq)
            text1.insert(tk.INSERT, word)
            text1.insert(tk.INSERT, "\n")
            freq, word = tup
            print freq, word
root = tk.Tk(className = " most_frequant_word")


# text entry field, width=width chars, height=lines text


text1 = tk.Text(root, width=50, height=20, bg='green')
text1.pack()
# function listed in command will be executed on button click
button1 = tk.Button(root, text='result', command=most_frequant_word)
button1.pack(pady=5)

# define a variable to hold the label text
v1 = tk.StringVar()
# label text will always be the textvariable's value
# width/height in char size
label1 = tk.Label(root, textvariable=v1, width=50, height=20)
label1.pack(pady=5)

# start cursor in text1.
text1.focus()
root.mainloop()

 text1.insert(tk.INSERT, freq)
            text1.insert(tk.INSERT, word)
            text1.insert(tk.INSERT, "\n")