Java: RegEx Problem (using the '.' all character symbol)

Question

I have some document stored as a large String. In the String I have some inline XML tags and I want to get out the words inbetween the tags. The documents may also contain HTML tags, as the documents are often web sites.

Example Document:

"< tr > My name is < b >< PERSON >Bobby< /PERSON >< /b >, I live in the USA."

Current RegEx:

Pattern p = Pattern.compile("<(LOCATION|PERSON|ORGANIZATION)>[\\w[ '\"/\\!%$\\(\\)\\-\\+]]*</(LOCATION|PERSON|ORGANIZATION)>");

Matcher m = p.matcher("I'm <PERSON>Graham Brown</PERSON> I went to the <LOCATION>USA'S</LOCATION>");

while(m.find()){
    System.out.println(m.group());
}

Result = < PERSON >Bobby< /PERSON > < LOCATION >USA< /LOCATION >

This works fine with pretty much most puntuation and grammer, but the Regex should allow any character pattern to be found between the tags. When I try using '.' (any character), as below it returns the whole String.

"< tr > My name is < b >< PERSON >Bobby< /PERSON >< /b >, I live in the USA."

 Pattern p = Pattern.compile("<(LOCATION|PERSON|ORGANIZATION)>.</(LOCATION|PERSON|ORGANIZATION)>");

How do I return any characters between the angular openinng and closing tags?

EDIT: Thanks for your responses. Just and for helping get the correct answer. For clarification I have marked Named Entites using NER. If you are unware of what this is please see some of the papers I have referenced at the bottom.

All I am interested in is getting the text between the three opening and closing tags. There are no other tags and the documents are not XML files and I am not parsing all the HTML tags nor I am I interested in them. All I am interested in is parsing the XML tags that I have created hence I though RegEx would be the simplest way to do so.

Papers to be added later...

Answer 1

Put a question mark after .*? for non-greedy processes.

Pattern p = Pattern.compile("<(LOCATION|PERSON|ORGANIZATION)>.*?</(LOCATION|PERSON|ORGANIZATION)>");

PS: I am just correcting your regex, but it does not mean it's the solution. Using parsers are always better idea.

Answer 2

Please use an XML parser for XML snippets. It's the right tool for your problem.

Edit: And use a HTML sanitizer to preprocess the HTML file. Additionally define a strict XML schema for the XML to ensure the XML structure.

Answer 3

Parsing Html The Cthulhu Way, by Jeff Atwood

Answer 4

There's only one answer for this problem: You can't parse HTML with regex