Getting phrases which have each word starting with an uppercase character

Question

Currently i have tried getting words that start with an upper case in a sentence using Character.isUpperCase. However now i would like to only retrieve phrases in a sentence where all the 1st letter in every word of the phrase is upper case. How should i go about doing it.

E.g "This is a sample sentence so that Ang Mo Kio Avenue 1 is part of Ang Mo Kio."

I would retrieve "Ang Mo Kio Avenue 1" and "Ang Mo Kio".

String s = "This is a sample sentence so that Ang Mo Kio Avenue 1 is part of Ang Mo Kio.";
String[] words = s.split("[^a-zA-Z']+");
for (int i = 0; i < words.length; i++) {
  if (Character.isUpperCase(words[i].charAt(0))) {
  System.out.println(words[i]);
}}

The real intention is to extract 3 or more uppercase words, optionally followed by a number

I don't see how *1* is uppercase, and why *This* is not included in expected output even if it's uppercase. Also by splitting by `\\s` you include punctuation marks into words. — m0skit0, Nov 18 '13 at 11:28

Masudul · Accepted Answer · 2013-11-18T12:40:56.697

2

i would like to only retrieve phrases in a sentence where all the 1st letter in every word of the phrase is upper case

For that you need to capture consecutive upper case words, and append them in StringBuilder. If lowercase letter come then initialize StringBuilder.

Try,

StringBuilder answer = new StringBuilder();
String s
 = "This is a sample sentence so that Ang Mo Kio Avenue 1 is part of Ang Mo Kio.";
    String[] words = s.split("\\s+");
    int count=0;
    for (int i = 0; i < words.length; i++) {
        char firstChar=words[i].charAt(0);
        if (Character.isUpperCase(firstChar) 
                 ||(count>0  && Character.isDigit(firstChar))) {
            answer.append(" "+words[i]);
            count++;
        } else {
            //To avoid less than 3 word apply this logic.
            if(count>2){
            System.out.println(answer);
            }
            count=0;
            answer = new StringBuilder();
        }
    }
    System.out.println(answer);// Also need to print answer here.

Output:

 Ang Mo Kio Avenue 1
 Ang Mo Kio.

edited Nov 18 '13 at 12:40

answered Nov 18 '13 at 11:30

Masudul

21,823
5
43
58

Just for the sake of my OCD, replace `isPrase` with `isPhrase` and this seems like a decent answer. Also Kudos for using `StringBuilder`, although implementing it so that it `stores` the phrases is probably more attractive. – christopher Nov 18 '13 at 11:31
You also need to print the answer at the end of program so that if the last word starts with a capital letter, it gets printed. – Chandranshu Nov 18 '13 at 11:31
I think the real intention is to extract 2 or more uppercase words, optionally followed by a street number, but the OP might need to clarify on this. – Matt Nov 18 '13 at 11:35
ya as Matt says sorry for the confusion. – user2541163 Nov 18 '13 at 11:36
hi what expression should i use if i would only like to accept letters and numbers. The current \\s includes punctuations – user2541163 Nov 18 '13 at 11:47
\\s is regex, which accept whitespaces, to capture only letters and numbers use \\w. – Masudul Nov 18 '13 at 11:49
Hi i tried your code but the output still includes a single word – user2541163 Nov 18 '13 at 11:52
@user2541163, what type of single word you want to avoid? – Masudul Nov 18 '13 at 12:20
Any word. I would like to capture phrases that have 3 or more words – user2541163 Nov 18 '13 at 12:27
@user2541163, Use count instead of isPhrase, I have updated my answer. – Masudul Nov 18 '13 at 12:34

score 1 · Answer 2 · answered Nov 18 '13 at 11:54

As basic starting code, you might try the following function:

private static void printStreetNames(String text) {
    List<String> words = new ArrayList<String>();

    for (String w : text.split("\\s+")) {
        if (Character.isUpperCase(w.charAt(0))) {
            words.add(w);
            continue;
        }

        if (w.matches("\\d+") && words.size() > 1) {
            words.add(w);
            continue;
        }

        if (words.size() >= 2) {
            System.out.println(words);
        }
        words = new ArrayList<String>();
    }

    if (words.size() >= 2) {
        System.out.println(words);
    }
}

Output:

[Ang, Mo, Kio, Avenue, 1]
[Ang, Mo, Kio.]

There are some caveats though. For example the following would not parse correctly: Ang Mo Kio 1 1 (because we do not check if we already added a street number). Also it does not remove trailing . from the parsed street names (e.g. Kio.) But I'll leave it up to you as an excercise.

Getting phrases which have each word starting with an uppercase character

2 Answers2

Output: