Suppose i have a string " kk a.b.cjkmkc jjkocc a.b.c. jjj 'a.b.ckkkkkkkkkkkkkkkk ' "
I want to replace the substring a.b.c in the string which are only outside the single quote , but it is not working.
Here is my code `
String str = " kk a.b.cjkmkc jjkocc a.b.c. jjj 'a.b.ckkkkkkkkkkkkkkkk ' ";
Pattern p = Pattern.compile("a\\.b\\.c");
Matcher m = p.matcher(str);
int x = m.find()
`
use this pattern : a\.b\.c(?=(([^']*'){2})*[^']*$) Demo
a.b.cxxx'xxx. It is not clear whether the input is always expected to have an even number of quote marks.
' a.b.cxxx xxx "outside single quotes" by default means quotes should come in pairs, right?a.b.c in that case. And whether quotes "should" come in pairs doesn't really matter; programs have to be able to handle incorrect inputs.'a'b' which letter is between quotes and which one is not?To search for a substring outside quotes, you can do something like this:
Pattern pat = Pattern.compile("^(?:[^']|'[^']*')*?a\\.b\\.c");
The first part will skip over:
every character that isn't a quote mark ([^']), or
every sequence of non-quote-mark characters enclosed in quotes ('[^']*').
Once those are skipped, then if it sees the pattern you want, it will know that it isn't inside quote marks.
This will handle a simple case. If things start getting more complicated, e.g. you want to allow \' to quote a quote mark in your input string the way C or Java does in a string literal, the regex starts getting more complicated, and you can quickly reach a point whether either your regex is unreadable or regexes aren't suitable solutions.
EDIT: fixed to put "reluctant" qualifier after second *, so that the first a.b.c will be found.
EDIT 2: If you want to replace the substring you find, it gets trickier. The above pattern matches the entire beginning of the string up through a.b.c, and I couldn't get a look-behind to work so that the match would be only the a.b.c part. I think you'll need to put the beginning of the string in a group, and then use $1 in the replacement string to copy the beginning:
Pattern pat = Pattern.compile("^((?:[^']|'[^']*')*?)a\\.b\\.c");
Matcher m = pat.matcher(source);
if (m.find()) {
result = m.replaceFirst("$1replacement");
}
I'm not sure replaceAll works with this, so if you want to replace all of them, you may need to loop.
I wouldn't mess with REGEX.
public static void main(String[] args) {
String str = " kk a.b.cjkmkc jjkocc a.b.c. jjj 'a.b.ckkkkkkkkkkkkkkkk ' ";
String[] s = str.split("'");
str = s[0].replaceAll("[abc]", "") + "'"+ s[1]+"'"
+ s[2].replaceAll("[abc]", "");
System.out.println(str);
}
OP:
kk ..jkmk jjko ... jjj 'a.b.ckkkkkkkkkkkkkkkk '
Inefficient.. but works
