3

I have an array and I want to search it for strings which start with "test" (for example); what is the most efficient way to search for these set prefixes? Regular expressions or if statements?

Regex:

    boolean found = false;
    for (String line: ArrayList){
                Pattern pattern = 
                Pattern.compile("^test"); //regex

                Matcher matcher = 
                pattern.matcher(line);


                while (matcher.find()) {
                    found = true;
                }
                if(found){
                    doSomething();
                         } 
                    }
                }

if Statement:

for (String line : ArrayList) {
       if (line.startsWith("test"){
            doSomething();
            }

Which is most efficient?

Which method is most effective for longer strings?

If I want to find Strings that start with "test" but then only ones which have "foo" after "test", which method is better?

If Regex is the answer, what is the correct syntax for saying starts with "test" followed by "foo" or "bar" but not both?

Improve this question
3
  • 5
    A regex for this IMO is too much... Commented Jul 12, 2012 at 15:02
  • 1
    Just a side note: even if it would make sense to use regular expressions here, you should compile the pattern outside the loop. Commented Jul 12, 2012 at 15:05
  • 1
    if you have cases in which one block should be responsible for more prefixes take a look into StringUtils.startsWithAny(String string, String[] searchStrings) Commented Jul 12, 2012 at 15:11

6 Answers 6

Reset to default
6

4 lines of code compared with 16 lines? Never mind which runs faster, the 4 line version is more efficient to write, more efficient to maintain.

If you have some code and have benchmarked it and identified a specific bottleneck then maybe consider making it more complicated, otherwise go with the simpler version every time.

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

2 Comments

What happens with each if I want to find more than just "foo" (~20 different suffixes) or more than just "test"? I asked this question because I received OutOfMemory errors for if statements.
If you have something more complex to match then something like a regex may be appropriate, but it's going to depend on what you're really doing. For the question you asked the answer is just to use line.startsWith(...)
5

Just use startsWith. Regex is a bit overkill, unless you want to accept String with leading spaces.

startsWith can work with "test" or even "testfoo". If you mean that "foo" can appear anywhere in the input after "test" (i.e. "testokokokfoonothing"), then regex should be used here.

Your code for regex version can be shortened to:

for (String line: ArrayList){
    if (line.matches("^test.*") {
        doSomething();
    }
}

matches() check if the whole input matches the regex, so a bit of modification to the regex is necessary. The code above is slightly slower, since the Pattern is recompiled.

Improve this answer

Comments

2

Regular expressions incur pretty significant overhead; if you're doing something for which there's a "direct" implementation -- like startsWith -- that will almost always be faster.

Improve this answer

Comments

1

I like the IF statements. Personally, I avoid Regex if at all possible. They get complicated and hard to code. When it comes to coding, one of the most important things is maintainability and Regex don't provide that.

Stay with the if statement.

Improve this answer

2 Comments

Well, that depends. There are cases that are easier to code and read using regex - if you understand them.
Maybe, but like you said IF, you understand them. When writing code, you aren't always the one who maintains it; therefore, you want to use what is going to be easiest for everyone. If you are the only one who knows how to use Regex, you will be the only one who will find it easy to work with, and that is not a good practice
1

You should prefer startsWith(). It is more idiomatic, concise, and it is unlikely that you will implement something faster, especially with java.util.regex, as it is relatively easy to write an inefficient regular expression.

In general, you should only be using java.util.regex when the standard methods on the String object will not suffice.

A couple notes on your Regex code, if you are going to use a regex for something else, you shouldn't run Pattern.compile inside your loop. you can simplify your code further by saying:

if (matcher.find()) {
    doSoemthing();
}
Improve this answer

Comments

0

The first thing yopu have to do : put Pattern pattern = Pattern.compile("^test"); //regex outside the for because compile eats time, and have to do once and only once.

   Pattern pattern = Pattern.compile("^test.*"); //regex
   for (String line: ArrayList){
       if(patter.matches(line)){
            oSomething();
       }
   }
Improve this answer

Comments

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.