Java ::使用正则表达式解析多行文本

Question

I want to parse a multiline text, so I wrote something like this:

String text = "[timestamp1] INFO - Message1 \r\n"
            + "[timestamp2] ERROR - Message2 \r\n"
            + "[timestamp3] INFO - Message3 \r\n"
            + "Message3_details1......... \r\n"
            + "Message3_details2 ......... \r\n";
String regex = "\\[(.*)\\] (.*) - (.*)";
Pattern p = Pattern.compile(regex, Pattern.DOTALL);
Matcher m = p.matcher(text);
while (m.find()) {
    System.out.println("G1: " + m.group(1));
    System.out.println("G2: " + m.group(2));
    System.out.println("G3: " + m.group(3));
    System.out.println();
}

What I want to get is this:

G1: timestamp1
G2: INFO
G3: message1

G1: timestamp2
G2: ERROR
G3: message2

G1: timestamp3
G2: INFO
G3: message3
    message_details1....
    message_details2...

But what I get is like this:

G1: timestamp1] INFO - Message1
    [timestamp2] ERROR - Message2
    [timestamp3
G2: INFO
G3: Message3
    Message3_details1........
    Message3_details2........

I'm not able to solve that even with Google's help.

Answer 1

You have used greedy quantifier in your regex. So, .* in [(.*)] will consume everything till the last found ] . You need to use reluctant quantifier. Add a ? after .* .

Also, for the last .* , you need to use a look-ahead, to make it stop before the next [ .

The following code would work:

String text = "[timestamp1] INFO - Message1 \r\n"
            + "[timestamp2] ERROR - Message2 \r\n"
            + "[timestamp3] INFO - Message3 \r\n"
            + "Message3_details1......... \r\n"
            + "Message3_details2 ......... \r\n";

String regex = "\\[(.*?)\\] (.*?) - (.*?)(?=\\[|$)";

Pattern p = Pattern.compile(regex, Pattern.DOTALL);
Matcher m = p.matcher(text);
while (m.find()) {
    System.out.println("G1: " + m.group(1));
    System.out.println("G2: " + m.group(2));
    System.out.println("G3: " + m.group(3));
    System.out.println();
}

The last part of the regex - (.*?)(?=\\\\[|$) matches everything till the [ in the next line, or till the end ( $ ). $ is required for the last two lines to be captured in group 3 of the last match.

Output:

G1: timestamp1
G2: INFO
G3: Message1 


G1: timestamp2
G2: ERROR
G3: Message2 


G1: timestamp3
G2: INFO
G3: Message3 
Message3_details1......... 
Message3_details2 ......... 

Answer 2

尝试"\\\\[(.*?)\\\\] (.*?) - (.*?) \\\\r\\\\n"