RegEx to split a String either at "," or L," or ",N

Question

I'm an absolute beginner on learning RegEx. I need to split a string into it's different Substrings, based on multiple conditions.

String:

"abc","def",NULL,"ghi",NULL,"jkl"

should be splitted to

[abc, def, NULL, ghi, NULL, jkl]

Currently I'm using String[] split = line.split("\",\""); to generate all substrings which are enclosed within "..." and seperated by ,. This works fine, but if there is a NULL value (which is not enclosed by "..."), the substrings are splittet incorrect.

Is it possible to split the String into it's Substrings by using a RegEx which splits the String if one of the following conditions is given?

","
L,"
",N

Thanks in advance!

Can you show the actual string value....with characters properly escaped? — RobOhRob, Mar 12 '20 at 18:48
Do your string values contain commas or why don't you use `split(",")`? — f1sh, Mar 12 '20 at 18:49
Note: Rather than split, **capture** all values *surrounded* by quotes, or all `NULL` that follow or preceed a comma — OneCricketeer, Mar 12 '20 at 18:50

Ryan · Answer 1 · 2020-03-12T19:38:35.460

0

    String input = "\"abc\",\"def\",NULL,NULL,\"ghi\",NULL,\"jkl\"";
    String [] split = input.split("(\",\")|(\",)|(,\")|(L,)");
    for (int i = 0; i < split.length; i++) {
        if (split[i].startsWith("\"")) {
            split[i] = split[i].substring(1);
        }
        if (split[i].endsWith("\"")) {
            split[i] = split[i].substring(0, split[i].length() -1);
        }
        if (split[i].equals("NUL")) {
            split[i] = "NULL";
        }
    }

OUTPUT

[abc, def, NULL, NULL, ghi, NULL, jkl]

edited Mar 12 '20 at 19:38

answered Mar 12 '20 at 18:43

Ryan

590
2
7

1

this messes up string values that contain a comma. Which is probably the reason why OP doesn't just use `split(",")`. – f1sh Mar 12 '20 at 18:45
Updated to handle commas within valid strings. – Ryan Mar 12 '20 at 19:09
now it doesnt handle something like `NULL,NULL` – f1sh Mar 12 '20 at 19:27
I think that's the limit of regex. The only other thing I can think to do is adding a case for (L,) and then replacing all NUL with NULL. – Ryan Mar 12 '20 at 19:35

score 0 · Answer 2 · answered Mar 12 '20 at 19:09

We can first replace all double quotes with an empty string then we can split the resulting string with a delimiter comma. As shown below

String[] strArray = str.replaceAll("\"" , "").split(",");

 String str = "\"abc\",\"def\",NULL,\"ghi\",NULL,\"jkl\"";
 String[] strArray = str.replaceAll("\"" , "").split(",");
 Arrays.asList(strArray).stream().forEach(System.out::println);

score -1 · Answer 3 · answered Mar 12 '20 at 18:44

-1

I think you don't need regexp. Your problem may be solved with split() and replace() methods.

public static void main(String[] args) {
    String str = "\"abc\",\"def\",NULL,\"ghi\",NULL,\"jkl\"";
    List<String> list = Arrays.stream(str.split(","))
            .map(part -> part.replace("\"", ""))
            .collect(Collectors.toList());
    System.out.println(list);
}

Output:

[abc, def, NULL, ghi, NULL, jkl]

answered Mar 12 '20 at 18:44

ekiryuhin

803
4
19

1

this messes up string values that contain a comma. Which is probably the reason why OP doesn't just use `split(",")`. – f1sh Mar 12 '20 at 18:45
The argument to split is always a regex, btw – OneCricketeer Mar 12 '20 at 18:51

RegEx to split a String either at "," or L," or ",N

3 Answers3