Null-safety with JSON and PostgreSQL

If you’re using the native JSON support in PostgreSQL you may, from time to time, come across the issue of null safety. For example, if you’re trying to query a nested JSON object which contains null values. Attempting to use one of the JSON operators on a null will result in an error like this:

ERROR:  cannot extract element from a scalar

PostgreSQL’s -> operator is generally pretty good at soaking up null references. You can chain it on a nested object and it’ll always return null if it encounters a key which doesn’t exist. However, if it encounters a key whose value is actually null. you’ll get the above error.

The easiest way to avoid this is to make sure you never set a JSON key to null if you’re ever going to treat that key as an object. If you can’t make that guarantee though, you’ve got a few other options to navigate safely around the null.

Test table

We’ll use this little table for testing:

CREATE TABLE books (id int, author json);
INSERT INTO books VALUES (1, null),
  (2, '{ "first_name": "Mary" }'),
  (3, '{ "address": { "street_name": "19 Red Avenue" } }'),
  (4, '{ "address": null }');

With the behaviour of the -> operator, we are able to do this without an error (it’ll return null):

SELECT author->'address'->'street_name' FROM books where id = 1;

But if we do this, we’ll get an error:

SELECT author->'address'->'street_name' FROM books where id = 4;
ERROR:  cannot extract element from a scalar

Using a CASE expression

We can use a case expression to bail out in case the field turns out to be a null:

      when (author->>'address') IS NULL then null
      else (author->'address'->>'street_name')
  'No street name') AS author_street_name
FROM books
WHERE id = 4;

Note that we’re using the ->> operator to check for the null. This coerces the return value as text, which responds appropriately to the IS NULL check. We’ve also used the coalesce function to provide a default value. The result:

 id | author_street_name
  4 | No street name

This technique will work for any level of nesting, because when used in this way the case expression essentially ‘short circuits’ and will stop evaulating when it has a match. There are some caveats around this with regards to constant expressions, but these shouldn’t apply in most cases.

Using a function

Alternatively, you may find it more convenient (and less verbose) to create your own function. The function below, json_fetch, is a simple implementation which will safely traverse and return a nested object without errors.

CREATE OR REPLACE FUNCTION json_fetch(object json, variadic nodes text[])
RETURNS json AS $$
  result json := object;
  k text;
  foreach k in array nodes loop
    if (result ->> k) is null then
      result := null;
    end if;

    result := result -> k;
  end loop;

  return result;
$$ LANGUAGE plpgsql;

To use it, you pass it the object you’re working with (i.e. the author field) and then one parameter for each nested field. Note that the function will return the found object (unless it’s null) as a json type, so you’ll usually need to cast it to text or another suitable type when you use it.

    json_fetch(author, 'address', 'street_name')::text, 'No address'
  ) AS street_name
FROM books;
 id |   street_name
  1 | No address
  2 | No address
  3 | "19 Red Avenue"
  4 | No address

Hope that helps!

blog comments powered by Disqus