Nested Regex matching in Java

Question

I have a scenario where I want to do a nested expression matching in Java.

Consider the following expression:

SUM_ALL(2:3,4:5)>20

where SUM_ALL has a reserved operator meaning in the application. Now, I want to extract the operator name and its arguments from a given expression. For doing the same, I have defined my pattern expression as follows:

Pattern testPattern = Pattern.compile("[^a-zA-Z]*([a-zA-Z_]+)\\s*\\(\\s*([0-9:,]+)\\s*\\).*");

This works fine if the expression is limited simply to the above. Here is the output for the same:

Group 1: SUM_ALL
Group 2: 2:3,4:5

Now, in a given expression, I may not be aware of the number of such operators present. For example consider the following case:

SUM_ALL(4:5,6:7)>MAX(2:3,4:4)+MIN(3:4,5:7)

Now, I want to extract each of the above operators and their respective arguments to perform the calculation according to their reserved meaning and then evaluate a simple math expression.

If there were a nesting capability in the Java pattern matcher it would have helped to extract the operators one by one by considering the rest of the expression once an operator is resolved. I know it is possible to do it by capturing the rest of the expression in a separate group and then running the matcher on that group value and keep doing it until we reach the end of the expression, but I would be more interested to know if the pattern-matcher has an inherent functionality for the same.

When managing such formulae which can have free forms, you should start looking into lexers / parsers, like Parboiled or Javacc... — PhiLho
– PhiLho, Commented Oct 25, 2012 at 11:00
The absolute best resource to learn regex for Java (and Perl, .NET, PHP/PCRE, etc.) is of course the book: Mastering Regular Expressions (3rd Edition) By Jeffrey Friedl. Hands down the most useful book I've ever read. — ridgerunner
– ridgerunner, Commented Oct 25, 2012 at 12:57

anubhava · Accepted Answer · 2012-10-25 11:12:52Z

1

You can use code like this:

String str = "SUM_ALL(4:5,6:7)>MAX(2:3,4:4)+MIN(3:4,5:7)";
Matcher m = 
    Pattern.compile("(?i).*?([a-z_]+)\\s*\\(\\s*([\\d:,]+)\\s*\\)").matcher(str);
while (m.find())
   System.out.printf("%s :: %s%n", m.group(1), m.group(2));

OUTPUT:

SUM_ALL :: 4:5,6:7
MAX :: 2:3,4:4
MIN :: 3:4,5:7

answered Oct 25, 2012 at 11:12

anubhava

790k67 gold badges603 silver badges671 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Abhishek Jain Over a year ago

Thanks for a wonderful answer. Can you please suggest some good sources to read on regex and parsing in java?

anubhava Over a year ago

Java official doc: docs.oracle.com/javase/tutorial/essential/regex and this tutorial here: vogella.com/articles/JavaRegularExpressions/article.html are good starting points.

Javier Diaz · Accepted Answer · 2012-10-25 11:09:29Z

1

Well, have this:

(?:(SUM_ALL|MAX|MIN|addmorehere)\\(((?:\d+:\d+,?){2})\\)[+-><*/addmorehere]?)+

It's not really scaped for java or any language but you get the idea

answered Oct 25, 2012 at 11:09

Javier Diaz

1,8301 gold badge18 silver badges25 bronze badges

Collectives™ on Stack Overflow

Nested Regex matching in Java

2 Answers 2

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related