Place random character in string if not a certain word

Question

For instance, I have the following string:

Hello how are you today, [name]?

How would I go about randomly placing characters between a random choice of words but not [name]? I already I have this following piece of code but I was hoping there is a better way of going about it.

string = 'Hello how are you today, [name]?'
characters = 'qwertyuioplkjhgfdsazxcvbnm,. '
arr = string.rsplit(" ")

for i in range(0, len(arr)):
    x = arr[i]
    if x == '[name]':
        continue
    if (random.randint(0,2)==1) :
        rnd=random.randint(1,len(x)-2)
        tmp1 = random.randint(0,len(characters))
        rndCharacter = characters[tmp1:tmp1+1]
        x = x[0:rnd] + rndCharacter + x[rnd+1:]
        arr[i] = x

" ".join(arr)

> Hellio how are yoy todsy, [name]?"

Though this replaces the character with a another random character. What way would I go about having it randomly replace or place a random character after or before a character as well?

Basically I'm just trying to simulate a sort of typo generator.

Thanks

Update on my code so far:

string = 'Hey how are you doing, [name]?'
characters = 'aeiou'
arr = string.rsplit(" ")
for i in range(0, len(arr)):
    x = arr[i]
    if x == '[name]': continue
    if len(x) > 3:
        if random.random() > 0.7:
            rnd = random.randint(0,len(x)-1)
            rndCharacter = random.choice(characters)
            if random.random() > 0.7:
                x = x[0:rnd] + rndCharacter + x[rnd+1:]
            else:
                x = x[:rnd] + rndCharacter + x[rnd:]
            arr[i] = x
    else:
        if random.random() > 0.7:
            rnd = random.randint(0,len(x)-1)
            rndCharacter = random.choice(characters)
            x = x[:rnd] + rndCharacter + x[rnd:]
            arr[i] = x
print " ".join(arr)

> Hey houw are you doiang, [name]?

UPDATE:

Maybe my final update for the code, hopefully this will help someone out some point in the future

def misspeller(word):
    typos = { 'a': 'aqwedcxzs',
              'b': 'bgfv nh',
              'c': 'cdx vf',
              'd': 'desxcfr',
              'e': 'e3wsdfr4',
              'f': 'fredcvgt',
              'g': 'gtrfvbhyt',
              'h': 'hytgbnju',
              'i': 'i8ujko9',
              'j': 'juyhnmki',
              'k': 'kiujm,lo',
              'l': 'loik,.;p',
              'm': 'mkjn ,',
              'n': 'nhb mjh',
              'o': 'o9ikl;p0',
              'p': 'p0ol;[-',
              'q': 'q1asw2',
              'r': 'r4edft5',
              's': 'swazxde',
              't': 't5rfgy6',
              'u': 'u7yhji8',
              'v': 'vfc bg',
              'w': 'w2qasde3',
              'x': 'xszcd',
              'y': 'y6tghu7',
              'z': 'zaZxs',
              ' ': ' bvcnm',
              '"': '"{:?}',
              '\'': '[;/\']',
              ':': ':PL>?"{',
              '<': '<LKM >',
              '>': '>:L<?:',
              ';': ';pl,.;[',
              '[': '[-p;\']=',
              ']': '=[\'',
              '{': '{[_P:"}+',
              '}': '}=[\']=',
              '|': '|\]\'',
              '.': '.l,/;',
              ',': ',lkm.'
            }

    index = random.randint(1,len(word)-1)
    letter = list(word[:index])[-1].lower()
    try:
        if random.random() <= 0.5:
            return word[:index] + random.choice(list(typos[letter])) + word[index:]
        else:
            return word[:index-1] + random.choice(list(typos[letter])) + word[index:]
    except KeyError:
        return word

def generate(self, s, n, safe_name):
    misspelled_s = ''
    misspelled_list = []
    for item in s.split(' '):
        if n:
            if safe_name in item:
                misspelled_list.append(item)
            else:
                r = random.randint(0,1)
                if r == 1 and len(re.sub('[^A-Za-z0-9]+', '', item)) > 3:
                    misspelled_list.append(misspeller(item))
                    n -= 1
                else:
                    misspelled_list.append(item)
        else:
            misspelled_list.append(item)
    return ' '.join(misspelled_list)

Answer 1

You are "hoping there is a better way of going about it". Well, here's some suggestions, and some code demonstrating those suggestions. Some of the suggestions are about making the code more pythonic or easy to read, not just the mechanics of changing strings.

1. Use module re for regular expressions to detect "[name]". This will pay dividends as soon as you have more keywords than just this one.
2. for x in string.rsplit(" ") is a more pythonic way to loop through the words.
3. Get real-numbered randoms, and compare to a probability setting in range 0.0-1.0. More flexible than getting integers 0,1.
4. Use x[:rnd] + ... + x[rnd:x] as recommended by others, for easier string manipulation.
5. Use x if condition else y for a concise choice between alternatives, in this case between an index which causes overwriting and an index which causes insertion.
6. Your example output shows a typo inserted in "you", but your example code only inserts typos if len(x) > 3. I follow your code, but this is easy to change.

Hope this helps.

import random
import re

string = 'Hello how are you today, [name]?'
characters = 'qwertyuioplkjhgfdsazxcvbnm,. '
words = []

for x in string.rsplit(" "):
    if    None == re.search('[^\]]*\[[a-z]+\].*', x) \
      and len(x) > 3 and random.random()<=0.5:
        # rnd: index of char to overwrite or insert before
        rnd = random.randint(2,len(x)-2)
        # rnd1: index of 1st char after modification
        # random.random <= 0.x is probability of overwriting instead of inserting
        rnd1 = rnd + 1 if random.random() <= 0.5 else 0
        x = x[:rnd] + random.choice(characters) + x[rnd1:]

    words.append(x)

typos = " ".join(words)
print typos

Update: fixed indentation mistake in the code.

Update 2: made the code for choosing overwrite versus insert to be more concise.

Answer 2

import random

def misspeller(word):
    characters = 'qwertyuioplkjhgfdsazxcvbnm,. '
    rand_word_position = random.randint(-1,len(word))
    rand_characters_position = random.randint(0,len(characters)-1)

    if rand_word_position == -1:
        misspelled_word = characters[rand_characters_position] + word 
    elif rand_word_position == len(word):
        misspelled_word = word + characters[rand_characters_position] 
    else:
        misspelled_word = list(word)
        misspelled_word[rand_word_position] = characters[rand_characters_position]
        misspelled_word = ''.join(misspelled_word)        
    return misspelled_word

s = 'Hello how are you today, [name]?'
misspelled_s = ''
misspelled_list = []
for item in s.split(' '):
    if '[name]' in item:
        misspelled_list.append(item)
    else:
        misspelled_list.append(misspeller(item))
misspelled_s = ' '.join(misspelled_list)
print misspelled_s

Examples of what I'm getting from misspelled_s are:

'Hellk howg ars youf poday, [name]?'
'Heylo how arer y,u todab, [name]?'
'Hrllo hfw  are zyou totay, [name]?'

Edited to clean up a couple of mistakes and omissions on first copy.

Edit 2 If you don't want every word to be affected you can modify the for loop in the following way:

for item in s.split(' '):
    n = random.randint(0,1)
    if '[name]' in item:
        misspelled_list.append(item)
    elif n == 1:
        misspelled_list.append(misspeller(item))
    else:
        misspelled_list.append(item)

You can modify the probability that a word is modified by changing how n is generated e.g. n = random.randint(0,10)

Answer 3

You can also use split('[name]'), and work on the sub-string, this way you'll be sure (see note below) of don't change '[name]'.

You may have problem splitting on every [name] occurance catching some substring of some longer name, but if you:

• Use real and common names in capital letter (like Jonh).
• Avoid the use of similar names.

Then the following code should work fine:

def typo(string):
    index = random.randint(1,len(string)-1)   # don't change first or last
    return string[:index] + random.choice(characters) + string[index:]

def generate(string, n, safe_name):
    sub_strings = string.split(safe_name)
    while n:
        sub_index = random.randint(0,len(sub_strings) - 1)
        sub = sub_strings[sub_index]
        if len(sub) <= 2:   # if too short don't change
            continue

        sub_strings[sub_index] = typo(sub)
        n -= 1
    return safe_name.join(sub_strings)

Example adding 3 new random charachter:

>>> string = 'Hello how are you today, Alice?'
>>> generate(string, 3, 'Alice')
'Hellov howj are yoiu today, Alice?'

With the name occurring more then one time:

>>> string = 'Hello Alice, how are you today, Alice?'
>>> generate(string, 3, 'Alice')
'Hello Alice, hoiw arfe you todayq, Alice?'

Answer 4

I think @sgallen's answer will work, but I have a few tips (for your previous code, and going forward).

for i in range(0, len(arr)):
    x = arr[i]

# is the same as

for i,x in enumerate(arr):

---

else:
    if random...:

# to

elif random...:

---

Using string as the name of a variable, isn't a good practice. The reason being, there is a string module. It might even come in handy for this because of the string constants. Alternatives could be inp, or data, or sentence.

# For example

>>> import string
>>> string.lowercase
'abcdefghijklmnopqrstuvwxyz'

---

By the way, if anyone notices errors in the above, leave a comment. Thanks.

Answer 5

For the example you have given, it looks like we can split it on the comma and put the typo(s) in the first part of the string.

If this is correct, you need to do three things randomly before generating the typo:

• Choose which character to do the typo on or near
• Choose the typo character
• Choose one of three actions - Replace, Prefix, Append

Would this fit the bill?

(BTW, since you are familiar with random, I didn't give any code.)

Answer 6

If you want to place a letter before or after instead of replacing, just fix the indices in your splicing so that they don't jump over a letter - i.e. use

x = x[:rnd] + rndCharacter + x[rnd:]

That way the new character will be inserted in the middle, instead of replacing an existing one.

Also, you can use rndCharacter = random.choice(characters) instead of using tmp1 like that.