正则表达式提取方程中的数学变量

Question

I have a mathematical expression given as a String and I have to extract all the variables which are identified as a letter, possibly followed by a number (eg x or x0 ). It works for simple expressions but if I try it with a more complicated equation I pick also numbers which I don't want since my goal is to determinate if the two equations use the same variables.

     String expression = "((x0+(2.0^x))/(21.1-x0))";
     for (String variable : expression.split("[^a-z0-9?]")) {
       if(!variable.isEmpty()){
         System.out.print(variable + " ");
     };

and the output is:

x0 2 0 x 21 1 x0

where I wanted

x0 x x0

why it takes also digits without a letter before? I already tried every possible combination of \\b and I didn't find anything online.

Answer 1

Instead of splitting on what you don't want, you can also match what you are looking for.

\b[a-z]\d*\b

Regex demo | Java demo

String regex = "\\b[a-z]\\d*\\b";
String s = "((x0+(2.0^x))/(21.1-x0))";
List<String> matches = new ArrayList<String>();
Matcher m = Pattern.compile(regex, Pattern.CASE_INSENSITIVE).matcher(s);

while (m.find()) {
    matches.add(m.group());
}

for (String match : matches) {
    System.out.println(match);
}

Output

x0
x
x0

Answer 2

The expression keeps the digits because they are not included in the regex search for the split method when creating the String variable .

Try splitting at one or many non-alphanumeric characters ( \W+ ), which may be followed by zero or many digits ( \d* ).

"\W+\d*"

Adding \d* to the end of your existing regex should also work.

"[^a-z0-9?]\d*"

Tested on regex101 with Java 8.

Please let me know whether this resolved your question.