Re: RegEx - Chk for special chars

M Kinnear wrote:

I want to check a string only contains a-z 0-9 ( ) . and #

I've used

ereg("^[a-zA-z0-9().#]*)$"),$instr)

Firstly, it looks to me like a couple of extra brackets slipped into the
code you posted, as I don't think the above would parse at all. I'm
assuming what you're actually using is:

ereg("^[a-zA-z0-9().#]*$", $instr)

Firstly, use preg_match. It runs faster that ereg; and ereg is being
phased out in newer versions of PHP. This gives you:

preg_match("/^[a-zA-z0-9().#]*$/", $instr)

Now, the dot (.) in regular expressions has a special meaning. It means
"absolutely any character you want". This means that you are allowing
absolutely any character at all -- that is, your regular expression is
doing nothing. The dot needs to be escaped with a backslash:

preg_match("/^[a-zA-z0-9()\.#]*$/", $instr)

Also, parentheses (these) have a special meaning within regular
expressions, marking out subexpressions. IIRC, this shouldn't be a problem
in your case because that special meaning disappears within square
brackets. However, it doesn't hurt to escape them, and it looks a bit
clearer to people just glancing at the expression:

preg_match("/^[a-zA-z0-9\(\)\.#]*$/", $instr)

Now try it.

Second q: is there clarification on whether characters need to be
escaped when part of a regex statement? i.e., should it be [$] or [\$] -
some websites/posts state they should be escaped, others say they
shouldn't

Generally it never hurts to escape, so when something could go either way,
err on the side of escaping things. Note that there are two levels of
escapes coming into play here -- firstly as you're using "double quotes"
you need to escape against PHP's built in string-parsing. Secondly, you
need to escape against special meanings of characters in regular
expressions.

To make things simpler, it is usually preferable to use 'single quotes',
so that you only have to worry about one set of escaping.

preg_match('/^[a-zA-z0-9\(\)\.#]*$/', $instr)

--
Toby A Inkster BSc (Hons) ARCS
http://tobyinkster.co.uk/
Geek of ~ HTML/SQL/Perl/PHP/Python*/Apache/Linux

* = I'm getting there!
.

Relevant Pages

  • Re: RegEx - Chk for special chars
    ... the dot in regular expressions has a special meaning. ... Generally it never hurts to escape, so when something could go either way, ...
    (comp.lang.php)
  • Re: A string-replace?
    ... is there a good way to escape ... regular expressions so they would be treated as strings by ppcre (the ... You can write a wrapper around REGEX-REPLACE, ...
    (comp.lang.lisp)
  • String.Format and Curley Braces
    ... I am trying my hand at RegEx and came across a tangent; curley braces upset ... string.format expressions. ... Is there a way to escape them with out making ...
    (microsoft.public.dotnet.general)
  • Re: Regular expression to allow only selected string values
    ... I get confused with when to escape and when not to ... it's easier to use single quotes if ... sed uses basic regular expressions. ... You also get some extras like "+" with extended regular ...
    (comp.unix.shell)
  • Re: Regular Expression Matches
    ... The escaping stuff gets a little confusing because the regular expressions ... You need to escape the '' characters in your Regurlar Expression. ... Regex indexRegex = new Regex(categoryListRegex, ... MatchCollection indexMatches = indexRegex.Matches; ...
    (microsoft.public.dotnet.languages.csharp)