Bought to you by vincevincevince over at webmasterworld.
The engine itself:
print preg_replace(“/\{([^\{]{1,100}?)\}/e”,”$$1″,file_get_contents(“template.tpl”));
Format of template.tpl file:
<html>
<head>
<title>{title}</title>
</head>
<body>
<h1>{header}</h1>
{text}
</body>
</html>
Setting variables:
$title=”Example page”;
$header=”My Examples”;
$text=”See the placeholders replaced?”;
“/\{([^\{]{1,100}?)\}/e”,”$$1″
I’m delimiting the regular expression with / / and using the modifier ‘e’ which causes the second argument to be evaluated by php.
Explanation
The pattern looks for an opening curly-brace ( \{ ) – the end of the pattern is a closing curly-brace ( \} ).
In between the two braces I look for any character which isn’t an opening curly-brace [^\{], avoiding mistaken nesting of tags.
I match between 1 and 100 of these non-{ characters by writing {1,100} and then I make the match non-greedy (try to find the shortest strings between { and }, not the longest) by adding a?. (? after *, + or {a,b} expressions changes them to non-greedy – in other situations? means 0 or 1 of the preceding).
The full string of non-{ characters is matched and stored as string $1 by surrounding that part of the pattern with brackets ().
Finally, the second argument of the preg_replace is “$$1”, using variable variables. If the pattern encounters “{title}” then the matched string $1 is “title” and so $$1 is $title.