Matching Regular Expressions java when word contains embedded &

Question

I use a regular expression to match words (and not match embedded strings) that looks looks like this:

word.trim().matches("(\\w+)(\\.|\\?|\\!|\\<)?");

This works great if the word is a string like this word = " blue ";

But it fails if the word has an embedded &, or any special character, like this
word = " A&P "; or this word = " A-P ";

How can I alter my regular expression so that it picks up certain special characters like "&" and "-"

possible duplicate of [Reference - What does this regex mean?](http://stackoverflow.com/questions/22937618/reference-what-does-this-regex-mean) — Biffen, Mar 13 '15 at 10:58
@TheLostMind It's a more compact version of `(\\.|\\?|\\!|\\ — Biffen, Mar 13 '15 at 11:08
@Biffen - Nope. I meant. *why not include it as part of the first charset*? — TheLostMind, Mar 13 '15 at 11:12
@Elliott There seems to be some confusion about what you want the regex to match. Could you supply some samples, preferably both positive and negative ones? — Biffen, Mar 13 '15 at 11:25
@Bohemian - yes.. But not captured :P.. I am really confused with what the OP wants.. So, I can't say much :P — TheLostMind, Mar 13 '15 at 11:31
The Op does say "embedded" so that suggests, between, but not starting with and not ending with special characters & and -. (\\w+([-&]?\\w+)*)[.?! — Les, Mar 13 '15 at 11:36

Bohemian · Accepted Answer · 2015-03-13T11:21:11.720

1

This seems about the simplest:

str.matches("(?i)[a-z]+([-&][a-z]+)?[.!?<]?")

Also note the re-phrasing of your alternation to a character class, since each term was a single character and being a character class you don't even need to escape dot or "?".

The (?i) means "ignore case".

I'm not confident about the "<" being a legitimate "word ending"; you might consider removing it from the regex.

edited Mar 13 '15 at 11:21

answered Mar 13 '15 at 11:15

Bohemian

365,064
84
522
658

Matching Regular Expressions java when word contains embedded &

1 Answers1