Question 205687 in Javascript Question posted in
A very good W3school tutorial can be found here.

Javascript – Join texts based on condition in an array of dictionaries in JS/React

I have data like this:

Record= [
  { "tid": 1, "token_text": "Canis", "spanid": 1, "label": "Name" },
  { "tid": 2, "token_text": "Familiaris", "spanid": 1, "label": "Name" },
  { "tid": 3, "token_text": "is" },
  { "tid": 4, "token_text": "the" },
  { "tid": 5, "token_text": "scientific" },
  { "tid": 6, "token_text": "name" },
  { "tid": 7, "token_text": "of" },
  { "tid": 8, "token_text": "dog", "spanid": 2, "label": "species" },
  { "tid": 9, "token_text": "." }
]

I want to create another array of dictionaries like the data where the token_text strings that have the same spanid are joined and the tid and label_name can be the same as that of the first occurrence. I want to do this while keeping the order of the elements in the array the same as that of the Record array. I want to do this in JS/ React. I will appreciate any leads in this matter. The output will look like the following:

Record = [
  { "tid": 1, "token_text": "Canis Familiaris", "spanid": 1, "label": "Name" },
  { "tid": 3, "token_text": "is" },
  { "tid": 4, "token_text": "the" },
  { "tid": 5, "token_text": "scientific" },
  { "tid": 6, "token_text": "name" },
  { "tid": 7, "token_text": "of" },
  { "tid": 8, "token_text": "dog", "spanid": 2, "label": "species" },
  { "tid": 9, "token_text": "." }
]

I tried looping but couldn’t figure out exactly.

for (let j = 0; j < spanid.length; j++) {
  var arr = [];
  Records.forEach(i => {
    if (i.spanid === spanid[j]) {
      arr.push(i.tid);
    }
  })
}

Is there a way to do it like this to join the token_texts with same spanids and delete the instances where that have duplicate spanids?

Tags:

3

Answers


  1. 0

    One solution is to use a for loop and an if statement to iterate over the array and check the spanid value of each object.

    For example:

    for (let i = 0; i < Record.length - 1; i++) {
  if (Record[i].hasOwnProperty("spanid")) {
    if (Record[i].spanid === Record[i + 1].spanid) {
      Record[i].token_text += " " + Record[i + 1].token_text;
      Record.splice(i + 1, 1);
      i--;
    }
  }
}

    You can use a nested loop to handle the cases where there are more than two token_text strings with the same spanid. This way, you can combine all those strings into one with a space between them.

    For example:

    for (let i = 0; i < Record.length - 1; i++) {
  if (Record[i].hasOwnProperty("spanid")) {
    let token_text = Record[i].token_text;
    let spanid = Record[i].spanid;
    for (let j = i + 1; j < Record.length && Record[j].spanid === spanid; j++) {
      token_text += " " + Record[j].token_text;
      Record.splice(j, 1);
      j--;
    }
    Record[i].token_text = token_text;
  }
}
    Login or Signup to reply.
  2. 0 
    const record = [
  { "tid": 1, "token_text": "Canis", "spanid": 1, "label": "Name" },
  { "tid": 2, "token_text": "Familiaris", "spanid": 1, "label": "Name" },
  { "tid": 3, "token_text": "is" },
  { "tid": 4, "token_text": "the" },
  { "tid": 5, "token_text": "scientific" },
  { "tid": 6, "token_text": "name" },
  { "tid": 7, "token_text": "of" },
  { "tid": 8, "token_text": "dog", "spanid": 2, "label": "species" },
  { "tid": 9, "token_text": "." }
];

const transformedRecord = [];
const spanidSet = new Set();

record.forEach(item => {
  if (item.spanid && !spanidSet.has(item.spanid)) {
    const joinedText = record
      .filter(i => i.spanid === item.spanid)
      .map(i => i.token_text)
      .join(' ');
      
    transformedRecord.push({
      tid: item.tid,
      token_text: joinedText,
      spanid: item.spanid,
      label: item.label
    });
    
    spanidSet.add(item.spanid);
  } else if (!item.spanid) {
    transformedRecord.push(item);
  }
});

console.log(transformedRecord);
    Login or Signup to reply.
  3. 0

    You can reduce the array of tokens, and check the previous token’s label against the current.

    In order to append to the previous token_text you need to verify three things:

    1. There is a previous (only false for the first iteration)
    2. The current token has a label
    3. The current and previous token labels match

    Working example

    Note: You could shorten the if-condition (below) to simply:

    if (token.label && token.label === prev?.label) {

    const record = [
  { "tid": 1, "token_text": "Canis", "spanid": 1, "label": "Name" },
  { "tid": 2, "token_text": "Familiaris", "spanid": 1, "label": "Name" },
  { "tid": 3, "token_text": "is" },
  { "tid": 4, "token_text": "the" },
  { "tid": 5, "token_text": "scientific" },
  { "tid": 6, "token_text": "name" },
  { "tid": 7, "token_text": "of" },
  { "tid": 8, "token_text": "dog", "spanid": 2, "label": "species" },
  { "tid": 9, "token_text": "." }
];

const joinTokensByLabel = (tokens) => {
  return tokens.reduce((result, token) => {
    const prev = result[result.length - 1];
    if (prev && token.label != null && token.label === prev.label) {
      prev.token_text += ` ${token.token_text}`;
    } else {
      result.push(structuredClone(token));
    }
    return result;
  }, []);
};

const fixed = joinTokensByLabel(record);

console.log(fixed);
    .as-console-wrapper { top: 0; max-height: 100% !important; }
    Login or Signup to reply.
Please signup or login to give your own answer.

Back To Top
Search