Replacing letters with numbers with its position in alphabet

Question

If anything in the text isn't a letter, ignore it and don't return it.

a being 1, b being 2, etc.

As an example:

alphabet_position("The sunset sets at twelve o' clock.") Should return "20 8 5 19 21 14 19 5 20 19 5 20 19 1 20 20 23 5 12 22 5 15 3 12 15 3 11" as a string.

It is my naive solution and my code is below in python 2.7

def alphabet_position(text):
    dictt = {'a':'1','b':'2','c':'3','d':'4','e':'5','f':'6','g':'7','h':'8',
    'i':'9','j':'10','k':'11','l':'12','m':'13','n':'14','o':'15','p':'16','q':'17',
    'r':'18','s':'19','t':'20','u':'21','v':'22','w':'23','x':'24','y':'25','z':'26'
    }
    arr = []
    new_text = text.lower()
    for i in list(new_text):
        for k, j in dictt.iteritems():
            if k == i:
                arr.append(j)
    return ' '.join(arr)

Answer 1

First of all, you don't need to hardcode the letters and their positions in the alphabet - you can use the string.ascii_lowercase.

Also, you don't have to call list() on a new_text - you can just iterate over it character by character.

Then, what if we would construct a mapping between letters and letter indexes in the alphabet (with the help of enumerate()). Then, use a list comprehension to create an array of numbers which we then join to produce a result:

from string import ascii_lowercase


LETTERS = {letter: str(index) for index, letter in enumerate(ascii_lowercase, start=1)} 

def alphabet_position(text):
    text = text.lower()

    numbers = [LETTERS[character] for character in text if character in LETTERS]

    return ' '.join(numbers)

Answer 2

Another slightly different approach than what @alecxe proposed (well, not so different ^_^), would be to use Python's builtins count and zip to generate the mapping between letters and their position in the alphabet.

from itertools import count
from string import ascii_lowercase


def letter_indexes(text):
    letter_mapping = dict(zip(ascii_lowercase, count(1)))
    indexes = [
      letter_mapping[letter] for letter in text.lower() 
      if letter in letter_mapping
    ]

    return ' '.join(str(index) for index in indexes)

Answer 3

If you only care about ASCII characters, you can also exploit the fact that their character codes go from 97 for 'a' to 122 for 'z', contiguously, and do something like

def alphabet_position(text):
    nums = [str(ord(x) - 96) for x in text.lower() if x >= 'a' and x <= 'z']
    return " ".join(nums)

Note, however, that it may give an impression of being faster than the @alecxe's solution, but is, in fact, quite a lot slower for long input strings, because calling str() and ord() on every input character is slower than dictionary lookup. Gives about the same or even slightly better performance for repeated calls on short input strings, but only because letters dictionary is constructed anew on every call of @alecxe's function, which is easy to change. (UPD.: no, not anymore). If that matters.

Answer 4

Finally, after a lot of head-banging, I found a way to avoid calling ord(), which, apparently, is very expensive. Below is the test code and results:

from timeit import timeit
from itertools import count
from string import ascii_lowercase


def alphabet_position_Headcrab(text):
    nums = [str(ord(x) - 96) for x in text.lower() if x >= 'a' and x <= 'z']
    return " ".join(nums)


def alphabet_position_wvxvw(text):
    result, i = [32, 32, 32] * len(text), 0
    for c in bytes(text.lower(), 'ascii'):
        if 97 <= c < 106:
            result[i] = c - 48
            i += 2
        elif 106 <= c < 116:
            result[i] = 49
            result[i + 1] = c - 58
            i += 3
        elif 116 <= c <= 122:
            result[i] = 50
            result[i + 1] = c - 68
            i += 3
    return bytes(result[:i-1])


def letter_indexes(text):
    text = text.lower()

    letter_mapping = dict(zip(ascii_lowercase, count(1)))
    indexes = [
      letter_mapping[letter] for letter in text
      if letter in letter_mapping
    ]

    return ' '.join(str(index) for index in indexes)


def test(f):
    data = "The sunset sets at twelve o' clock."
    for _ in range(5):
        f(data)
        data = data + data


def speed_compare():
    results = {
        'wvxvw': timeit(
            'test(alphabet_position_wvxvw)',
            setup='from __main__ import (test, alphabet_position_wvxvw)',
            number=10000,
        ),
        'Headcrab': timeit(
            'test(alphabet_position_Headcrab)',
            setup='from __main__ import (test, alphabet_position_Headcrab)',
            number=10000,
        ),
        'MrGrj': timeit(
            'test(letter_indexes)',
            setup=(
                'from __main__ import (test, letter_indexes)\n'
                'from itertools import count\n'
                'from string import ascii_lowercase\n'
            ),
            number=10000,
        )
    }
    for k, v in results.items():
        print(k, 'scored', v)

Running speed_compare() gives this output:

wvxvw scored 1.7537127458490431
Headcrab scored 2.346936965826899
MrGrj scored 2.2078608609735966

Answer 5

Seems complicated with all the libraries requirements for this simple problem. I'm not proficient in Python syntax but I know this can be done. Let me explain in pseudocode (syntax can be looked up by you):

1. Declare a list (or array as I call it in other langs) of alphabet[]="'a', 'b',....'z'" Their index position is ALREADY their positions...so, the position of 'a' is 0 (because Python index is 0-based; if you don't like it, you can just add 1 to the result to show results counting from 1) (lowercase or uppercase, doesn't matter since you're looking at the position of them in either case not their ord values)
2. Get the string to compare and its length: len(astring) For that length, iterate and find the character (achar for example) in the string (astring for example). Either for loop or some magical python 1-line statements like: for achar in astring If you use a 1-line statement instead of a loop, you'd probably have to load them into another list and loop in that. Not sure, I like for loops, so... Inside the for loop, check each achar's index location if chr(x) == astring[i] for example where i is loop counter, then load that location into a new array (define an empty array outside of the loop first) say, "positions[]"
3. After the loop ends, positions[] should have the positions (0-based) of each matching char. You could add 1 to each value if you want the positions to be 1-based.

Sorry, I couldn't write the whole code mainly because my syntax knowledge isn't very good in Python yet, but the logic should make sense, right?