What are proxies to efficiently estimate whether a file contains JSON?

Question

Not sure if this is the kind of question that SO allows, but better sorry than safe:

I have a script that should check whether a file is JSON, and replace it with a JSON version if it is not. The script is supposed to be fast. If I go with the brute-force solution:

try:
    json.loads(file)
except ValueError:
    replace(file)

... it is an order of magnitude too slow. Now in my specific situation, it's okay to have some false positives, so I decided to go with an imperfect solution for the sake of speed. For example:

if file.firstletter in ["{", "["]:
    replace(file)

(fwiw this is pseudocode. My question is language agnostic)

This will work in most cases, since most files that aren't JSON tend to start without brackets.

Another example I can think of is whether the file ends with a bracket.

What are other ways to efficiently estimate whether a file contains JSON?

Answer 1

You could look at the file extension (eg *.json ).

Other than that checking the starting/ending characters ( { , [ ) is a good heuristic to quickly rule out non-JSON files. But it can lead to false positives, so as the second step you should still try parsing the file as JSON to really know if it's JSON.