🦸 Querying with Regexes

In the last step, you learned how to query for documents that match a specific value. In this step, you'll learn how to query for documents that match a pattern.

This is similar to doing a LIKE query in SQL.

Edit the regex endpoint

In the server/controllers/movies.mjs file, you will notice a method called findWithRegex. This is where we'll add the code to query the collection with a regex.

Using the almost same code you used in the previous step, add the code needed to query the data using a regex. To do so, you need to change the value passed in the query to a regular expression rather than the parameter directly.

To do so, you will need to use the $regex operator. You can find more about this operator in the documentation.

Answer

  async findWithRegex(searchTerm) {
    const movies = await collection
      .find({ title: {$regex: new RegExp(searchTerm, "i")} }, {
        title: 1,
        year: 1,
        "imdb.rating": 1,
        fullplot: 1,
        poster: 1,
        released: 1,
        genres: 1
      })
      .limit(20)
      .toArray();
    return movies;
  }

Test it out

Try it out. In the client, change the dropdown to Regex Query, and type in a partial name. You should see the results filtered by the regex. Something like Terminator should now return multiple results.

Problems with this approach

This approach works, but it's not ideal. The reason it's not very efficient is that it has to scan the entire collection to find the matches. This is because the regex is not indexed.

That doesn't mean you should never use regexes to query. If you have a small collection, it's not a big deal. But if you have a large collection, it can be a problem.

A good use case for a regex is when you want to match documents starting with a certain string. For example, if you wanted to find all movies whose title starts with "A", you could use the regex /^A/. This would be a good use case for a regex as it can still leverage the index on the title field.