Regex to match word surrounded by non-alphanumeric characters

Question

I want to match and find index of word either surrounded by space or special characters. For example:

To find: test
this is input test : True
this is#input_ : True
this isinput : False
thisisinputtest: False
this @test is right: True.

How do I match this and find index. My current regex fails: (?i)[^a-zA-Z0-9]test[^a-zA-Z0-9]

This is a different case because in my case, #test_ should return index of test not #test. — Maxsteel, Nov 13 '16 at 09:46
`_` is a word, not a special character. Besides, `this is#input_` cannot be `True`, where is the word `test` there? The `#test_ ` is missing in your test cases. — Wiktor Stribiżew, Nov 13 '16 at 09:46
*Non alphanumeric. Besides, on a match I want index of "test". All other examples gives a full match! — Maxsteel, Nov 13 '16 at 09:47
@WiktorStribiżew I understand that, but the answer in the linked question will match #test and give start index of `#` not test. Which is what I want. — Maxsteel, Nov 13 '16 at 09:48
Then what is your code? Please share. Still, please check the `this is#input_ : True` test case. — Wiktor Stribiżew, Nov 13 '16 at 09:49
Let us [continue this discussion in chat](http://chat.stackoverflow.com/rooms/127981/discussion-between-maxsteel-and-wiktor-stribizew). — Maxsteel, Nov 13 '16 at 09:50
Besides, since you need the index of the search word, it is easier to use lookarounds - `String pat = "(?<!\\p{Alnum})test(?!\\p{Alnum})";` — Wiktor Stribiżew, Nov 13 '16 at 09:52

score 5 · Accepted Answer · answered Nov 13 '16 at 09:56

5

I think what you need to use lookarounds in your case:

(?<!\p{Alnum})test(?!\p{Alnum})

The negative lookbehind (?<!\p{Alnum}) will fail the match if there is an alphanumeric char present to the left of the test, and the negative lookahead (?!\p{Alnum}) will fail the match if there is an alphanumeric char right after test.

See the testing screenshot:

Java demo:

String str = "this is#test_ :";
Pattern ptrn = Pattern.compile("(?<!\\p{Alnum})test(?!\\p{Alnum})");
Matcher matcher = ptrn.matcher(str);
while (matcher.find()) {
    System.out.println(matcher.start());
}

Alternative way: match and capture the search word, and print the start position of the 1st capturing group:

Pattern ptrn = Pattern.compile("\\P{Alnum}(test)\\P{Alnum}");
...
System.out.println(matcher.start(1));

See this Java demo

NOTE that in this scenario, the \P{Alnum} is a consuming pattern, and in some edge cases, test might not get matched.

answered Nov 13 '16 at 09:56

Wiktor Stribiżew

607,720
39
448
563

Although I wasn't clear in the question (sorry about that!), this is exactly what I wanted! Thanks a lot! – Maxsteel Nov 13 '16 at 10:06
For those who only need alphabetic characters and not alphanumeric, you can use `"(?<!\\p{Alpha})test(?!\\p{Alpha})"` – Dat Nguyen Jul 13 '17 at 03:10
1

@DatNguyen: Note that `\p{Alpha}` works with ASCII only letters by default (if you do not specify the `Pattern.UNICODE_CHARACTER_CLASS` flag). To match any Unicode letters without depending on flags, use `"(?<!\\p{L})test(?!\\p{L})"`. – Wiktor Stribiżew Jul 13 '17 at 07:18
Good to know. Thanks @WiktorStribiżew! – Dat Nguyen Jul 13 '17 at 07:19

score 0 · Answer 2 · answered Nov 13 '16 at 09:58

I am just trying to understand your question. You are looking for test that is surrounded by a special character (_ included) or a space? But yet you say this is#input_ : True. I am not sure if I am picking this up wrong but how is that true in that case?

Anyway I have got the regex [\W\s_](input|test)[\W\s_] that matches all your cases defined as true.

\W matches any non words
\s matches any whitespace
_ matches any underscore - Has to be defined on its own as this is a word
From my confusion of the test mentioned I have the regex search for both input and test

Also I user this site anytime I work with Regex as I find it so useful.

Not sure if this is the answer your looking for but let me know if I am wrong and I will try again

Regex to match word surrounded by non-alphanumeric characters

2 Answers2