preg_replace not working on groups of symbols - php

Question

So I am trying to make a morse code encoder/decoder; I got the encoder done, but the decoder is giving me some problems.
So, if I use the function test and input "ab" it will return "ab". If however, I input "a b" it returns "c d" (as it should, 100% working)
function test($code){
$search = array('/\ba\b/', '/\bb\b/');
$replace = array('c', 'd');
return preg_replace($search, $replace, $code);
}
BUT when I use the function morsedecode and input ".- -..." it doesn't do anything and retuns ".- -...".
function morsedecode($code){
$search = array('/\b.-\b/', '/\b-...\b/');
$replace = array('a', 'b');
return preg_replace($search, $replace, $code);
}
I am stuck because it doesn't seem to be working for symbols, as it does for letters and words. Does anyone know the reason for this and is there anyway to work around this in PHP?

Answer 1

Update
If all your characters are surrounded by spaces (or beginning/end of line), you will probably find it easier to use strtr rather than a regex based approach. Since strtr replaces longest matches first, you don't have to worry about (for example) -.- (k) being partially replaced as -a.
function morsedecode($code){
$search = array('.-', '-...');
$replace = array('a', 'b');
return strtr($code, array_combine($search, $replace));
}
echo morsedecode(".- -...");
Output:
a b
Demo on 3v4l.org
Original Answer
Your problem is that \b matches a word boundary i.e. the place where the character to the left is a word character (a-zA-Z0-9_) and the character to the right a non-word character (or vice versa). Since you have no word characters in your input string, you can never match a word boundary. Instead, you could use lookarounds for a character which is not a dot or a dash:
function morsedecode($code){
$search = array('/(?<![.-])\.-(?![.-])/', '/(?<![.-])-\.\.\.(?![.-])/');
$replace = array('a', 'b');
return preg_replace($search, $replace, $code);
}
echo morsedecode(".- -...");
Output
a b
Demo on 3v4l.org
Note that . is a special character in regex (matching any character) and needs to be escaped, otherwise it will match a - as well as a ..

Answer 2

\b is a word boundary, which is any of the following.
Before the first character in the string, if the first character is a word character.
After the last character in the string, if the last character is a word character.
Between two characters in the string, where one is a word character and the other is not a word character.
'/\b.-\b/'
The first does \b not match in .- -... because of #1. Specifically if the first character is a word character
A word character = ASCII letter, digit or underscore so . is not a word character.
Also, you need to escape . characters like \..
Try looking for \s* (any number of white spaces) instead of a word boundary.
function morsedecode($code){
$search = array('/\s*\.-\s*/', '/\s*-\.\.\.\s*/');
$replace = array('a', 'b');
return preg_replace($search, $replace, $code);
}
Example
https://regex101.com/r/LCZXCn/1

Answer 3

I ended up coming up with my own little fix for the problem:
function morsedecode($code){
$bd_code = str_replace(array('.', '-', '/'), array('dot', 'dash', '~slash~'), $code);
$search = array('/\bdotdash\b/', '/\bdashdotdotdot\b/', '/\bdashdotdashdot\b/', 'etc..');
$replace = array('a', 'b', 'c', 'etc..');
$string = preg_replace($search, $replace, $bd_code);
return str_replace(array(' ', '~slash~'), array('', ' '), $string);
}
Definitely not the most efficient but gets the job done. #Nick answer is definitely an efficient way to go.