code for counting number of sentences, words and characters in an input file

问题

I have written the following code to count the number of sentences, words and characters in the input file sample.txt, which contains a paragraph of text. It works fine in giving the number of sentences and words, but does not give the precise and correct number of characters ( without whitespaces and punctuation marks)

lines,blanklines,sentences,words=0,0,0,0
num_chars=0

print '-'*50

try:
    filename = 'sample.txt'
    textf = open(filename,'r')c
except IOError:
    print 'cannot open file %s for reading' % filename
    import sys
    sys.exit(0)

for line in textf:
    print line
    lines += 1
    if line.startswith('\n'):
        blanklines += 1
    else:

    sentences += line.count('.')+ line.count ('!')+ line.count('?')

    tempwords = line.split(None)
    print tempwords 
    words += len(tempwords)


textf.close()

print '-'*50
print "Lines:", lines
print "blank lines:",blanklines
print "sentences:",sentences
print "words:",words

import nltk
import nltk.data
import nltk.tokenize

with open('sample.txt' , 'r') as f:
    for line in f:
        num_chars += len(line)

num_chars = num_chars - (words +1 )

pcount = 0
from nltk.tokenize import TreebankWordTokenizer
with open('sample.txt','r') as f1:
    for line in f1:
        #tokenised_words = nltk.tokenize.word_tokenize(line)
        tokenizer = TreebankWordTokenizer()
        tokenised_words = tokenizer.tokenize(line)
    for w in tokenised_words:
        if ((w=='.')|(w==';')|(w=='!')|(w=='?')):
            pcount = pcount + 1
print "pcount:",pcount
num_chars = num_chars - pcount
print "chars:",num_chars

pcount is the number of punctuation marks. Can some suggest the changes I need to make in order to find out the exact number of characters without spaces and punctuation marks?

回答1:

import string

#
# Per-line counting functions
#
def countLines(ln):      return 1
def countBlankLines(ln): return 0 if ln.strip() else 1
def countWords(ln):      return len(ln.split())

def charCounter(validChars):
    vc = set(validChars)
    def counter(ln):
        return sum(1 for ch in ln if ch in vc)
    return counter
countSentences = charCounter('.!?')
countLetters   = charCounter(string.letters)
countPunct     = charCounter(string.punctuation)

#
# do counting
#
class FileStats(object):
    def __init__(self, countFns, labels=None):
        super(FileStats,self).__init__()
        self.fns    = countFns
        self.labels = labels if labels else [fn.__name__ for fn in countFns]
        self.reset()

    def reset(self):
        self.counts = [0]*len(self.fns)

    def doFile(self, fname):
        try:
            with open(fname) as inf:
                for line in inf:
                    for i,fn in enumerate(self.fns):
                        self.counts[i] += fn(line)
        except IOError:
            print('Could not open file {0} for reading'.format(fname))

    def __str__(self):
        return '\n'.join('{0:20} {1:>6}'.format(label, count) for label,count in zip(self.labels, self.counts))

fs = FileStats(
    (countLines, countBlankLines, countSentences, countWords, countLetters, countPunct),
    ("Lines",    "Blank Lines",   "Sentences",    "Words",    "Letters",    "Punctuation")
)
fs.doFile('sample.txt')
print(fs)

results in

Lines                   101
Blank Lines              12
Sentences                48
Words                   339
Letters                1604
Punctuation             455

回答2:

You can also use a regex to replace all non-alphanumeric characters and then count the number of characters in each line.

回答3:

Once thing you could do is when you read the line iterate through it and increment number of characters:

for character in line:
    if character.isalnum():
        num_chars += 1

P.S. you might want to change if statement condition to satisfy your particular needs, i.e. if you want to count $ for example.

回答4:

Try this for count number of words and number of sentences and get probability for similar words,

from nltk.tokenize import word_tokenize
from nltk.tokenize import sent_tokenize


text_file = open("..//..//static//output.txt", "r")
lines = text_file.readlines()
x=0
tokenized_words = [word_tokenize(i) for i in lines]
for i in tokenized_words:

    print(i) #array contain with tokens
    print(str(len(i))) #word count

    for j in i:
        if j== 'words': #simple algo for count number of 'words' to be count
            x = x+1

tokenized_sents = [sent_tokenize(k) for k in lines]

for  k in tokenized_sents:
    print("Sentences"+str(k)) #array contain with sentences
    print("number of sentences "+str(len(k))) #number of sentences

print("number of word"+str(x))
print("Probability of 'word' in text file "+str(x/len(i)))

来源：https://stackoverflow.com/questions/5094351/code-for-counting-number-of-sentences-words-and-characters-in-an-input-file