using regex to replace words not enclosed in <> signs

Question

I really need some help forming a regex pattern in java.

I am trying to replace all occurrences of a given word provided that it is not enclosed within a less-than or greater-than sign. It should not replace it even if the less/greater-than signs are not directly next to the word. Here is my most recent attempt:

    stringIn = stringIn.replaceAll("((?:<[^>]>[^<]*)+" + word + "(?:<[^>]>[^<]*)+)", "$1<" + newWord + ">$2");

But this does not seem to replace the word not in brackets. Just to clarify I made up an example replacement:

    "foo word <word> <foo word foo>" should go to
    "foo <newWord> <word> <foo word foo>

PS Just saw this SO question and it looks similar but I don't really understand it and cannot guaranty it as it refers to PHP not Java.

@SaadAttieh For that, use the syntax `(?:abc)`. `(?abc)` won't work. — The Guy with The Hat, Jan 20 '14 at 14:13
@Sujith PS not sure exactly what you mean but basically it might have an input of word and only the second "word" should be replaced as the first two are within less/greater than signs. — Saad Attieh, Jan 20 '14 at 14:18

The Guy with The Hat · Accepted Answer · 2014-01-20T15:32:35.907

1

I would use the regex

word(?=[^>]*(?:$|<))

The items contained in {} in the following explanation show what part of the regex is explained by the previous phrase: The regex looks for a word {word} that is followed by {(?=...)} a sequence of zero or more characters {[...]*} that does not include the character > {^>} and is immediately followed by the end of the string {$} or {(?:...|...)} the character < {<}.

You can find a demonstration and another explanation here: http://regex101.com/r/tY5cT7

Implemented in Java, it would look like this:

stringIn = stringIn.replaceAll(word + "(?=[^>]*(?:$|<))", "<" + newWord + ">");

^{Note: this assumes that all <...> are balanced, i.e. for every < there is one and only one corresponding >, and for every > there is one and only one corresponding <.}

edited Jan 20 '14 at 15:32

answered Jan 20 '14 at 14:20

The Guy with The Hat

10,836
8
57
75

Great Answer! Thanks - could you possibly explain why this works? just not sure as you are able to put the word to replace at the beginning of the regex. Also what does the = sign do? – Saad Attieh Jan 20 '14 at 14:48
@SaadAttieh I edited my answer to include a better explanation. – The Guy with The Hat Jan 20 '14 at 15:32

score 0 · Answer 2 · answered Jan 20 '14 at 14:40

Here is code that should work for you:

String newWord = "-REPLACED-";
Pattern p = Pattern.compile( "(<[^>]*>)|(\\bword\\b)" );
Matcher m = p.matcher( "foo word <word> <foo word foo> word" );
StringBuffer sb = new StringBuffer();
while (m.find()) {
    if (m.groupCount()>1 && m.group(2) != null)
        m.appendReplacement(sb, "<" + newWord + ">");
}
m.appendTail(sb);
System.out.println( sb );

OUTPUT

foo <-REPLACED-> <word> <foo word foo> <-REPLACED->

score -1 · Answer 3 · answered Jan 20 '14 at 14:26

-1

You can use something like this :

String val="<blablabla word blablabla> word123 <word>";
String REGEX="(?<=>)[^<]+";
String replaceString="replace";
System.out.println(val.replaceAll(REGEX, replaceString));

Output

<blablabla word blablabla>replace<word>

Refer

answered Jan 20 '14 at 14:26

Sujith PS

4,776
3
34
61

Try this on the input `word`. – The Guy with The Hat Jan 20 '14 at 14:29

using regex to replace words not enclosed in <> signs

3 Answers3

OUTPUT

Linked