Capturing multiple text blocks with regex in java

Question

What regex should be used for extracting multiple text blocks delimited by theirs headers that also should be parsed, for example:

some text info before message sequence
============
first message header that should be parsed (may contain = character)
============
first multiline
message body that
should also be parsed
(may contain = character)
============
second message header that should be parsed
============
second multiline
message body that
should also be parsed
... and so on

I was trying to use:

String regex = "^=+$\n"+
        "^(.+)$\n"+
        "^=+$\n"+
        "((?s:(?!(^=.+)).+))";
Pattern p = Pattern.compile(regex, Pattern.MULTILINE);

But ((?s:(?!(^=.+)).+)) eats second message as weel. This is a test showing a problem:

import java.util.regex.Matcher;
import java.util.regex.Pattern;
import org.junit.Assert;
import org.junit.Test;
public class ParsingTest {
@Test
public void test() {
    String fstMsgHeader = "first message header that should be parsed (may contain = character)";
    String fstMsgBody = "first multiline\n"+
                        "message body that\n"+
                        "should also be parsed\n"+
                        "(may contain = character)";
    String sndMsgHeader = "second message header that should be parsed";
    String sndMsgBody = "second multiline\n"+
            "message body that\n"+
            "should also be parsed\n"+
            "... and so on";
    String sample = "some text info before message sequence\n"+
                    "============\n"+
                    fstMsgHeader+"\n"+
                    "============\n"+
                    fstMsgBody+"\n"+
                    "============\n"+
                    sndMsgHeader+"\n"+
                    "============\n"+
                    sndMsgBody +"\n";
    System.out.println(sample);
    String regex =  "^=+$\n"+
                    "^(.+)$\n"+
                    "^=+$\n"+
                    "((?s:(?!(^=.+)).+))";
    Pattern p = Pattern.compile(regex, Pattern.MULTILINE);
    Matcher matcher = p.matcher(sample);
    int blockNumber = 1;
    while (matcher.find()) {
        System.out.println("Block "+blockNumber+": "+matcher.group(0)+"\n_________________");
        if (blockNumber == 1) {
            Assert.assertEquals(fstMsgHeader, matcher.group(1));
            Assert.assertEquals(fstMsgBody, matcher.group(2));
        } else {
            Assert.assertEquals(sndMsgHeader, matcher.group(1));
            Assert.assertEquals(sndMsgBody, matcher.group(2));
        }
    }
}

}

What output do you expect to have, and which one do you actually have? — sp00m
– sp00m, Commented Aug 20, 2013 at 15:12
Reg. split usage: I've finished with split, but it seems that capturing message and its header with one regex makes code more clear (one while loop with group accessors). So I am trying to consider this variant. — Mikhail Tsaplin
– Mikhail Tsaplin, Commented Aug 20, 2013 at 15:25
2 sp00m: for each while iteration i want to extract message header and its body. First iteration successfuly extracts first message header (Assert.assertEquals(fstMsgHeader, matcher.group(1)); - passes, but matcher.group(2) captures first mesage body plus the rest of string). — Mikhail Tsaplin
– Mikhail Tsaplin, Commented Aug 20, 2013 at 15:34

Pshemo · Accepted Answer · 2013-08-20 16:01:13Z

1

I am not sure if that is what you are looking for but maybe this regex will help

String regex = 
        "={12}\n" +   // twelve '=' marks and new line mark
        "(.+?)" +     // minimal match that has
        "\n={12}\n" + // new line mark with twelve '=' marks after it
        "(.+?)(?=\n={12}|$)"; // minimal match that will have new line
                              // character and twelve `=` marks after
                              // it or end of data $

and to make it work you should make dot to also match new line characters with Pattern.DOTALL flag.

Pattern p = Pattern.compile(regex, Pattern.DOTALL);

edited Aug 20, 2013 at 16:01

answered Aug 20, 2013 at 15:54

Pshemo

125k26 gold badges194 silver badges280 bronze badges

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

2 Comments

Mikhail Tsaplin Over a year ago

Pshemo, thank you it is working. Can you describe what (.+?) means?

Pshemo Over a year ago

@MikhailTsaplin normally (.+?) is greedy so it will try to find maximal possible. If you add ? it will make + quantifier reluctant so it will try to find minimal match. More info at docs.oracle.com/javase/tutorial/essential/regex/quant.html.

Collectives™ on Stack Overflow

Capturing multiple text blocks with regex in java

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related