如何计算每个单词在字符串中出现的次数？

Question

I'm learning JavaScript now and I'm working on a function that will count how many times words show up in a string, then spits out the answer as an object. Per some suggestions on similar threads on this site, I've decided to use .split and a counter function to make a two-dimensional array, then populate an object with the results. I was running tests with sample texts, when I ran into some trouble with certain strings. I can't figure out why some of the counters are showing up as undefined.

function countWords(str) {
  var answer = {};
  if (str === '') {
    return answer;
  }
  var strArray = [];
  strArray = str.split(' ');
  //strArray is an array that holds the words as separate strings
  console.log('strArray: ' + strArray);
  var resultWords = [];
  var resultCount = [];
  var counter = 0;

  // only if the word doesnt show up in resultWords, push it in, and increase the counter. if it has shown up, disregard
  for (var i = 0; i<strArray.length; i++) {
    counter = 0;
    if (resultWords.indexOf( strArray[i] ) === -1)  {
      resultWords.push(strArray[i]);
      counter += 1;
      // if the word shows up again, increase the counter
      for (var j = i + 1; j < strArray.length; j++) {
        if (resultWords[i] === strArray[j]) {
          counter += 1;
        }
        // push to resultCount the counter for each word
        resultCount[i] = counter;
      }
    }
    // create an object where the key is the word from resultWords and the value is the number from wordCount
    for (var k = 0; k < resultWords.length; k++) {
      answer[ resultWords[k] ] = resultCount[k];            
    }
  }
  console.log('resultWords: ' + resultWords);
  console.log('resultCount: ' + resultCount);
  return answer;
}

var sample = 'how now brown cow how now';
console.log(sample);
var output = countWords( sample ); 

I find that when I use the words "this" and "is" in my sample, I'm often returning "undefined" as the word count for those words. For example, "to be or not to be" returns "undefined" for "that" and "is." Can someone help elucidate what's going on here? Thank you.

Answer 1

Your code is unreadable:

• too long blocks of code
• non-informative variable names (eg answer )

Next, your algorithm is very slow: you can count words by parsing the whole array only once.

Last but not least, you should create the answer array outside the loop.

Here's a shorter implementation that use modern javascript features (for the sake of learning):

function countWords(str) {
  const wordCounts = new Map()
  str.split(' ').forEach(word => {
    const currentWordCount = wordCounts.get(word) || 0
    wordCounts.set(word, currentWordCount+1)
  })

  /* Reproduce your output */
  const resultWords = [...wordCounts.keys()]
  const resultCount = [...wordCounts.values()]
  console.log('resultWords: ' + resultWords);
  console.log('resultCount: ' + resultCount);

  return wordCounts
}

On older js environment, you cannot use Map and arrow functions:

function countWords(str) {
  const wordCounts = {}
  str.split(' ').forEach(function(word) {
    const currentWordCount = wordCounts[word] || 0
    wordCounts[word] = currentWordCount+1
  })

  /* Reproduce your output */
  const resultWords = Object.keys(wordCounts)
  const resultCount = resultWords.map(function(word) { return wordCounts[word] })
  console.log('resultWords: ' + resultWords);
  console.log('resultCount: ' + resultCount);

  return wordCounts
}

Go back to your code, you get some undefined because of this line:

// push to resultCount the counter for each word
resultCount[i] = counter;

The index i is the index of the current word in strArray . You can fix it by removing this line, and do

resultCount.push(counter)

After the end of the loop that starts with for (var j = i + 1; j < strArray.length; j++) .