How to replace an unknown string? (only pattern is known) - php

Question

I want to replace a string which might appear within a URL.
The string has the following pattern:
%26TID%3D123456
I want to replace the 123456 part, to a specific value such as: 777777.
To be on the safe side though, I don't want to assume that the relevant part of the original string has necessarily 6 digits after the %3D part; I want to assume that the original string might contain a few more or few less characters (and I also can't tell the real value of each digit).
In addition, when I replace the string, since that string will usually appear in the middle of the URL, I need to replace it without modifying the rest of the URL. After that string, there would usually be another %26 string which I want to keep including whatever that is after it, but to be on the safe side, I don't want to assume that the original string is necessarily followed by %26.
What is the best practice to make such a replacement, that would stand up to all my above conditions?

Answer 1

The general rule is to specify the boundary (or an "anchor") (here, a starting one) and then match whatever you want with the more generic pattern.
Here, the "anchor" is the literal text TID%3D. The more generic pattern is one or more digits: \d+.
Since you need to replace the first occurrence, you need to pass 1 as the limit argument value in preg_replace.
So, combining all that:
$re = '~TID%3D\d+~';
$str = "%26TID%3D123456 %26PID%3D123456 %26TID%3D123456";
$subst = 'TID%3D7652';
echo $result = preg_replace($re, $subst, $str, 1);
// = > %26TID%3D7652 %26PID%3D123456 %26TID%3D123456
See IDEONE demo
If you do not want (or do not know) the "anchor" text, use a capturing mechanism (demo):
$re = '~(TID%\w{2})\d+~'; // (...) specify a capturing group referenced with ${1} later
$str = "%26TID%3D123456 %26PID%3D123456 %26TID%3D123456";
$subst = '${1}7652';
echo $result = preg_replace($re, $subst, $str, 1);
// = > %26TID%3D7652 %26PID%3D123456 %26TID%3D123456
You can also use a lookbehind approach, but it is less efficient:
$re = '~(?<=TID%3D)\d+~'; // (?<=TID%3D) makes sure digits are preceded with TID%3D substring
$str = "%26TID%3D123456 %26PID%3D123456 %26TID%3D123456";
$subst = '${1}7652';
echo $result = preg_replace($re, $subst, $str, 1);