How can I print the steps of a regular expression?

Only the first 2 options are for Java, the rest I include because they are relevant (clarifying why they are included).

jRegexAnalyser

It generates a debug under the java engine and is free and Open Source. It's not very visually pleasing and it's even hard to understand what each of the steps represents, but it does what you're looking for. It is mainly used to see the number of total steps ("Matcher Operations", "Op" column).

Link: http://www.schwebke.com/index.php/10/18/

Execution time ( benchmark )

While measuring the number of steps is a good start, the only way to really see how efficient a regular expression is is to measure the time it takes to execute, comparing 2 or more regular expressions with different text. In this way, you can also consider the time it takes to compile.

Roughly, the code would be:

int ciclos = 100000;  //Un número grande de repeticiones

long antes = System.nanoTime();  //grabar la hora en que comienza

for (int n=0; n<ciclos; n++)
{
    
    //Poner acá el código con la expresión regular a evaluar
    
}

long despues = System.nanoTime();  //La hora en que termina

//Imprimir cuánto tardó
System.out.println("Tiempo de ejecución: " + (despues-antes)/1e6 + " ms");

Ejemplo en ideone

regex101.com

Developed under pcre , it does not use java . It is one of the best visually accomplished tools and it helps enormously to understand what is happening at the internal engine level of a regular expression, so it is worth mentioning it. Especially since I am interested in clarifying the following:

• All Perl-like regular expression engines (" Perl-like ": PCRE, .Net, Python, Java, Perl, etc.) follow the same number of theoretical internal steps.
• Java and PCRE share most of the structures. A Java regular expression can most likely be used in PCRE ( they only differ in very few constructs ).
• The biggest difference between the engines lies in internal optimizations that help reduce the number of steps to evaluate. However, by generating a regular expression that evaluates fewer steps in one engine, it will almost certainly evaluate fewer steps in another engine.
• Tools used for debugging may count each step differently (1 step in one tool can be represented by multiple steps in another tool), but that doesn't mean more processing.

Link: https://regex101.com

_{* The request for regex101 to support the Java engine has been open for some time , but there is no sign of it moving forward.}

RegexBuddy

Windows 2000/XP/Vista/7/8/8.1/10 (or on Linux using Wine )

It doesn't use java , but it has its own engine that accepts structs and has options for Java regular expressions (it emulates it). By far the best desktop application for evaluating regular expressions. Unfortunately, it is not free.

Link: https://www.regexbuddy.com

With Debugexx link you can parse the steps of your regular expression. I hope it helps you, greetings!

Try this expression on the site

\b\d{1,3}.\d{1,3}.\d{1,3}.\d{1,3}\b