regular expression help

oochie

Hi, im trying to create a regular expression to validate user input on a web site. i want to check for the '$' character in the input string and validate what comes after each occurrence of '$'.

e.g. if my user is input is something like:
"some text here $validstring1 more text $validstring2 some more text $invalidstring"

i want a regular expression to tell me that this is invalid input as the $invalidstring part is invalid. There is a pre-defined set of about 20 strings that can occur after the '$' character. The user's input string could potentially contain none or all of these strings.

conorod

Questions:
1. What programming language are you using?
2. Do you know how many strings the user will enter?

oochie

1. its a Java web app but the regular expression is in a commons validator config file. But i guess its processed by some Java class or other. I want to try validate the input automatically using a regular expression & without having to examine the string in my Java code.
My config file has entries like this to validate input for an email address:

<var-value>
	<![CDATA[^[a-zA-Z0-9.]+@mydomain.com$ ]]>
</var-value>

2. No i dont know how many strings the user will enter. It will be a paragraph of free text.

conorod

I'll think about it, I'm not the best with Java and regex, sorry. Hopefully someone else can figure it out if I can't.

Inspector Gadget

I'm going to go ahead and assume that whatever's applying the regex to the string will recursively match a pattern against every hit it finds in the string, but assuming that's so, something simple like the following should do:

\$(\w+)

That will match a dollar sign followed by any combination of A-Z (upper or lower case), 0-9, and underscores (_); for example $test, or $test_1, but will strip out the dollar sign when returning it. If you want to do this with a search and replace in mind, you'll have to match the dollar sign too, so you could try this:

(\$(?:\w+))

I tried these in jEdit's search and replace thingy, which as far as I remember uses java.util.regex, and they worked fine - sorry, not a Java head. It'll work fine in Perl, PHP or Python, or anything else using PCRE (i.e. Perl-compatible regex) though. Here's what Python says, for example:

Python 2.5 (r25:51918, Sep 19 2006, 08:49:13) 
[GCC 4.0.1 (Apple Computer, Inc. build 5341)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import re
>>> teststring = "some text here $validstring1 more text $validstring2 some more text $invalidstring"
>>> print re.findall("\$(\w+)", teststring)
['validstring1', 'validstring2', 'invalidstring']
>>> print re.findall("(\$(?:\w+))", teststring)
['$validstring1', '$validstring2', '$invalidstring']
>>> 

That's what you want, right?
Gadget

Hobbes

Using Inspector Gadgets regex expression and RegExBuddy I asked it for the code to to move through the matches.

try {
	Pattern regex = Pattern.compile("\\$(\\w+)", Pattern.CASE_INSENSITIVE | Pattern.UNICODE_CASE | Pattern.MULTILINE);
	Matcher regexMatcher = regex.matcher(subjectString);
	while (regexMatcher.find()) {
		// matched text: regexMatcher.group()
		// match start: regexMatcher.start()
		// match end: regexMatcher.end()
	} 
} catch (PatternSyntaxException ex) {
	// Syntax error in the regular expression
}

This would probably be the best way. Then have your string of valid entries in one string and try to match your result. For example.

String t = "$one$two$three";
String r = "$two";

if ( t.contains(r)) System.out.println("Found it");