Firestore export schema structure to json

Question

I've been looking on the net and I'm not sure, it is possible to export the Firestore schema structure to a json?

Something similar to create a DDL in which are the instructions to create the table of a DB.

I just need the schema not the data.

Answer 1

As the comments mention, Firestore doesn't really follow a schema.

But we can still try to understand the structure of the data.

Here's what I did:

Step 1: export your data.

On the Firebase console, go to -> project settings -> Service account -> Generate new private key -> save it as exportedDB.json Run npx -p node-firestore-import-export firestore-export -a exportedDB.json -b original.json

Step 2: remove unnecessary information to make the file much smaller.

Here's a script I used to cut mine from 7MB to 10kB while still maintaining the structure. Essentially it just drops sibling documents and truncates long strings:

import json

def truncate_strings(json_obj):
    if isinstance(json_obj, dict):
        # check if object has more than one key that is alphanumeric and 20 characters long
        long_keys = [k for k in json_obj.keys() if isinstance(k, str) and k.isalnum() and len(k) == 20]
        if len(long_keys) > 1:
            # drop all keys except the first one
            first_key = long_keys[0]
            return {first_key: truncate_strings(json_obj[first_key])}
        else:
            # truncate all keys and values recursively
            return {k[:30]: truncate_strings(v) for k, v in json_obj.items()}
    elif isinstance(json_obj, list):
        if len(json_obj) > 5:
            truncated_list = [truncate_strings(elem) for elem in json_obj[:5]]
        else:
            truncated_list = [truncate_strings(elem) for elem in json_obj]
        return truncated_list
    elif isinstance(json_obj, str) and len(json_obj) > 30:
        return json_obj[:30]
    else:
        return json_obj

# Read in the original JSON file
with open('original.json', 'r') as f:
    original_json = json.load(f)

# Truncate any strings longer than 30 characters, drop array elements beyond the first 5,
# and drop object keys beyond the first one if there are multiple long keys
truncated_json = truncate_strings(original_json)

# Write the truncated JSON to a new file
with open('truncated.json', 'w') as f:
    json.dump(truncated_json, f)

Step 3: Ask your favorite LLM, such as Chat-GPT, to infer the schema for you:

Please infer the database schema from this json: [paste the contents of the truncated file]