using `_.pick` to extract relevant details from objects



~3 min read


437 words

Building on lessons from yesterday’s dive into groupBy, I found an opportunity to explore pick. This time it was because I wanted to send data from the client to my server for processing and committing to the database.

My APIs are written in Typescript and expect a particular shape of data. Since, I wanted to get the benefits of Typescript, my goal was not to rely on typing it as any. Which, of course, means that I needed to shape the data before sending it over.

In my case, I was sending an array of objects. Each object on the client side, however, was carrying around way too much.

Imagine an object like:

const exampleElement = {
  id: "abc123",
  keyProp: "yippee",
  otherProp: 1,
 otherProp2: ...,
  otherProp100: "done?",
  otherProp200: "not quite!",
 otherProp289: "finally!"

But, all we really needed to send across the wire was the id and the keyProp . So, at a high level we need to do the following:

  1. Iterate over the array
  2. Look at the element (an object)
  3. Return a new object (or mutate the original, though that won’t be our approach) with just the id and keyProp

Using vanilla Javascript, we could do this with a .map and custom function to construct the new object (since .map returns a new array, we would use the value to construct a new object).

For example:

const myArrOfObj = [{id:"abc123", keyProp: "yippee", ...}, ... ]

const newObjArr = => {
    let newEl = {} =;
    newEl.val = el.val;
    return newEl;

While this works, it’s a bit cumbersome.

Lodash offers a method in pick to do this with relative ease. We still need the map (since we’re picking from the objects elements of the array, not the array itself.

Pick, however, has a very declarative API indicating exactly what you’re trying to do - whereas with the example above, you need to read the code to see what’s happening.

const myArrOfObj = [{id:"abc123", keyProp: "yippee", ...}, ... ]
const picked = => _.pick(val, ["id", "val"]))

Two comparable ways to do the same thing. The benefit of the Lodash method in this case is that it makes it clear what is being accomplished. Reaching for a library is not always helpful - particularly if there’s a native way to achieve the same objective (as I did here), however, I gladly do it when it means that the program is more readable / declarative.

Note: Thank you Kyle Simpson for giving me the language to speak about this (imperative vs declarative code).

Hi there and thanks for reading! My name's Stephen. I live in Chicago with my wife, Kate, and dog, Finn. Want more? See about and get in touch!