DEV Community

Cover image for How to create a MongoDB full text index
Adrian Matei for Codever

Posted on • Edited on • Originally published at codever.dev

How to create a MongoDB full text index

When you are searching your code snippets on Codever, depending on the search terms you use, relevant results are displayed. Behind this there is a MongoDb full text search index. Let's see how it is created and used.

Full text search is supported in Mongo by using a text index. Text indexes can include any field whose value is a string or an array of string elements, to which you can give weights. For a text index, the weight of an indexed field denotes the significance of the field relative to the other indexed fields in terms of the text search score.

db.snippets.createIndex(
  {
    title: "text",
    tags: "text",
    "codeSnippets.comment": "text",
    "codeSnippets.code": "text",
    sourceUrl: "text"
  },
  {
    weights: {
      title: 8,
      tags: 13,
      "codeSnippets.comment": 3,
      "codeSnippets.code": 1,
      sourceUrl: 1
    },
    name: "full_text_search",
    default_language: "none",
    language_override: "none"
  }
);
Enter fullscreen mode Exit fullscreen mode

For each indexed field in the document, MongoDB multiplies the number of matches by the weight and sums the results. Using this sum, MongoDB then calculates the score for the document. You then can use the $meta operator for details on returning and sorting by text scores, as in the snippet below:

let getPublicBookmarksForSearchedTerms = async function (nonSpecialSearchTerms, page, limit, sort, specialSearchFilters, searchInclude) {

  let filter = {
    public: true
  }

  if ( nonSpecialSearchTerms.length > 0 ) {
    if(searchInclude === 'any') {
      filter.$text = {$search: nonSpecialSearchTerms.join(' ')}
    } else {
      filter.$text = {$search: bookmarksSearchHelper.generateFullSearchText(nonSpecialSearchTerms)};
    }
  }

  addSpecialSearchFiltersToMongoFilter(specialSearchFilters, filter);

  let sortBy = {};
  if ( sort === 'newest' ) {
    sortBy.createdAt = -1;
  } else {
    sortBy.score = {$meta: "textScore"}
  }

  let bookmarks = await Bookmark.find(
    filter,
    {
      score: {$meta: "textScore"}
    }
  )
    .sort(sortBy)
    .skip((page - 1) * limit)
    .limit(limit)
    .lean()
    .exec();

  return bookmarks;
}
Enter fullscreen mode Exit fullscreen mode

Shared with ❤️ from Codever.   👉 use the copy to mine functionality to add it to your personal snippets collection.

Top comments (0)