The Weekly Challenge - 365

TASK #2: Valid Token Counter
You are given a sentence.

Write a script to split the given sentence into space-separated tokens and count how many are valid words.
A token is valid if it contains no digits, has at most one hyphen surrounded by lowercase letters, and
at most one punctuation mark (!, ., ,) appearing only at the end.

#!/usr/bin/perl
use strict;
use warnings;

# The instruction is didactic in nature and outlines the step-by-step construction of a regular expression. 

# 1. A token is valid if it contains no digits (but letters: see examples), 
# /^[a-z]+$/

# - ^ -> start of the word
# - [a-z]+ -> one or more letters (lowercase)
# - $ -> end of the word
# - This regex matches: "hello", "c", but does not match: "hello,", "a-bcdef", "-oops"

# 2. has at most one hyphen surrounded by lowercase letters, 
# /^[a-z]+(-[a-z]+)?$/

# - (-[a-z]+)? -> zero or one occurrences of a hyphen followed by letters
# - This regex matches: "a-bcdef", "cd-xy" but does not match words with punctuation at the end like "hello," or "c!"

# 3. and at most one punctuation mark (!, ., ,) appearing only at the end.
# /^[a-z]+(-[a-z]+)?[!.,]?$/

# - [!.,]? -> optional single punctuation character at the end (? = zero or one)
# - This regex matches: "hello,", "c!", "cd-xy!"

# So the complete regex is: /^[a-z]+(-[a-z]+)?[!.,]?$/ 

# Now the counting part:
# - grep works on an array
# - grep filters the words that match the regex
# - grep returns an array
# - in scalar context, grep returns the number of matches

# So the result is with space-separated tokens as input:
# my $input_str = "hello, c! a-bcdef -oops ab- cd-xy!";
# my $no_valid_words = scalar grep { /^[a-z]+(-[a-z]+)?[!.,]?$/ } split(/\s+/, $input_str);
# return $no_valid_words;
# shorter: without variable $no_valid_words and simply: return scalar grep ...

# study below: die built-in; prints message and exits with error


# Implementation

sub valid_token_counter {
    my ($input_str) = @_;
    die "Invalid input: use lowercase letters, spaces, and . , ! - only (at least one letter required)\n"
        unless $input_str =~ /^(?=.*[a-z])[a-z.,! \-]+$/; # see footnote (1)
    
    return scalar grep { /^[a-z]+(-[a-z]+)?[!.,]?$/ } (split /\s+/, $input_str);
}


# Tests

my $str;

# Example 1
$str = "cat and dog"
printf "%d\n", valid_token_counter($str); # Output: 3

# Example 2
$str = "a-b c! d,e"
printf "%d\n", valid_token_counter($str); # Output: 2

# Example 3
$str = "hello-world! this is fun"
printf "%d\n", valid_token_counter($str); # Output: 4

# Example 4
$str = "ab- cd-ef gh- ij!"
printf "%d\n", valid_token_counter($str); # Output: 2

# Example 5
$str = "wow! a-b-c nice."
printf "%d\n", valid_token_counter($str); # Output: 2

# Footnote
# (1): ^(?=.*[a-z]) means "Starting at the beginning, ensure that somewhere in the string there is at least one lowercase letter”