正则表达式用于句子的开头和结尾

Question

Is there a way to match start and end of sentence in Java? The easiest case is ending with simple (.) dot. In some other cases it could end with colum (:) or a shortcut ended with colum (.:).

For example some random news text:

Cliffs have collapsed in New Zealand during an earthquake in the city of Christchurch on the South Island. No serious damage or fatalities were reported in the Valentine's Day quake that struck at 13:13 local time. Based on the med. report everybody were ok.

My goal is to get the shortcut of a word + the context of it, but if possible only the sentence in which the shortcut belonds.

So the successfull output for me will be if I would be able to get something like this:

selected word -> collapsed

context -> Cliffs have collapsed in New Zealand during an earthquake in the city of Christchurch on the South Island.

selected word -> med.

context -> Based on the med. report everybody were ok.

Thanks

Answer 1

what you are looking for is a natural language processing toolkit. for java you can use: CoreNLP and they already have some example cases on their tutorials page. you can certainly make a regex expression that looks for all chars inbetween the set of chars (.:? etc...), and it would look something like this:

\.*?(?=[\.\:])\

then you would have to loop through the matched results and find the relevant sentences which have your words in them. but i recommend you use a NLP to achieve this.

Answer 2

The code:

import java.util.HashMap;
import java.util.Map;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

public class Main {

   public static void main( String[] args ) {
      final Map<String, String> dict = new HashMap<>();
      dict.put( "med", "medical" );
      final String text =
         "Cliffs have collapsed in New Zealand during an earthquake in the "
         + "city of Christchurch on the South Island. No serious damage or "
         + "fatalities were reported in the Valentine's Day quake that struck "
         + "at 13:13 local time. Based on the med. report everybody were ok.";
      final Pattern p = Pattern.compile( "[^\\.]+\\W+(\\w+)\\." );
      final Matcher m = p.matcher( text );
      int pos = 0;
      while(( pos < text.length()) && m.find( pos )) {
         pos = m.end() + 1;
         final String word = m.group( 1 );
         if( dict.containsKey( word )) {
            final String repl            = dict.get( word );
            final String beginOfSentence = text.substring( m.start(), m.end());
            final String endOfSentence;
            if( m.find( pos )) {
               endOfSentence = text.substring( m.start() - 1, m.end());
            }
            else {
               endOfSentence = text.substring( m.start() - 1);
            }
            System.err.printf( "Replace '%s.' in '%s%s' with '%s'\n",
               word, beginOfSentence, endOfSentence, repl );
            final String sentence =
               ( beginOfSentence + endOfSentence ).replaceAll( word+'.', repl );
            System.err.println( sentence );
         }
      }
   }
}

The execution:

Replace 'med.' in 'Based on the med. report everybody were ok.' with 'medical'
Based on the medical report everybody were ok.

Answer 3

You spot the sentence easily. It starts with a capital letter and ends with one of .:!? chars followed by space and another capital letter or reached the end of the whole string.

Compare the difference time. Based time. Based and med. report med. report .

So the regex capturing the whole sentence should look like this:

([A-Z][a-z].*?[.:!?](?=$| [A-Z]))

Take a look! Regex101