Extract sentence with a keyword

Question

So I've got this question.

Write a program that extracts from a text all sentences that contain a particular word. We accept that the sentences are separated from each other by the character "." and the words are separated from one another by a character which is not a letter.

Sample text:

We are living in a yellow submarine. We don't have anything else. Inside the submarine is very tight. So we are drinking all the day. We will move out of it in 5 days.

Sample result:

We are living in a yellow submarine.

We will move out of it in 5 days. 

This my code so far.

public static string Extract(string str, string keyword)
    {

        string[] arr = str.Split('.');
        string answer = string.Empty;

        foreach(string sentence in arr)
        {
            var iter = sentence.GetEnumerator();
            while(iter.MoveNext())
            {
                if(iter.Current.ToString() == keyword)
                    answer += sentence;
            }
        }

        return answer;
    }

Well it does not work. I call it with this code:

string example = "We are living in a yellow submarine. We don't have anything else. Inside the submarine is very tight. So we are drinking all the day. We will move out of it in 5 days.";

string keyword = "in";
string answer = Extract(example, keyword);
Console.WriteLine(answer);

which does not output anything. It's probably the iterator part since I'm not familiar with iterators.

Anyhow, the hint for the question says we should use split and IndexOf methods.

Answer 1

sentence.GetEnumerator() is returning a CharEnumerator, so you're examining each character in each sentence. A single character will never be equal to the string "in", which is why it isn't working. You'll need to look at each word in each sentence and compare with the term you're looking for.

Answer 2

Try:

public static string Extract(string str, string keyword)
{
    string[] arr = str.Split('.');
    string answer = string.Empty;

    foreach(string sentence in arr)
    {
        //Add any other required punctuation characters for splitting words in the sentence
        string[] words = sentence.Split(new char[] { ' ', ',' });
        if(words.Contains(keyword)
        {
            answer += sentence;
        }
    }

    return answer;
}

Answer 3

Your code goes through each sentence character by character using the iterator. Unless the keyword is a single-character word (e.g. "I" or "a") there will be no match.

One way of solving this is to use LINQ to check if a sentence has the keyword, like this:

foreach(string sentence in arr)
{
    if(sentence.Split(' ').Any(w => w == keyword))
            answer += sentence+". ";
}

Demo on ideone.

Another approach would be using regular expressions to check for matches only on word boundaries. Note that you cannot use a plain Contains method, because doing so results in "false positives" (i.e. finding sentences where the keyword is embedded inside a longer word).

Another thing to note is the use of += for concatenation. This approach is inefficient, because many temporary throw-away objects get created. A better way of achieving the same result is using StringBuilder.

Answer 4

string input = "We are living in a yellow submarine. We don't have anything else. Inside the submarine is very tight. So we are drinking all the day. We will move out of it in 5 days.";
var lookup = input.Split('.')
                .Select(s => s.Split().Select(w => new { w, s }))
                .SelectMany(x => x)
                .ToLookup(x => x.w, x => x.s);

foreach(var sentence  in lookup["in"])
{
    Console.WriteLine(sentence);
}

Answer 5

I would split the input at the periods and followed by searching each sentence for the given word.

string metin = "We are living in a yellow submarine. We don't have anything else. Inside the submarine is very tight. So we are drinking all the day. We will move out of it in 5 days.";
string[] metinDizisi = metin.Split('.');
string answer = string.Empty;

for (int i = 0; i < metinDizisi.Length; i++)
{
    if (metinDizisi[i].Contains(" in "))
    {
        answer += metinDizisi[i];
    }
}

Console.WriteLine(answer);

Answer 6

You can use sentence.Contains(keyword) to check if the string has the word you are looking for.

public static string Extract(string str, string keyword)
    {
        string[] arr = str.Split('.');
        string answer = string.Empty;

        foreach(string sentence in arr)
            if(sentence.Contains(keyword))
                answer+=sentence;

        return answer;
    }

Answer 7

You could split on the period to get a collection of sentences, then filter those with a regex containing the keyword.

var results = example.Split('.')
    .Where(s => Regex.IsMatch(s, String.Format(@"\b{0}\b", keyword)));