A simple Python utility to download and save Firestore collections to CSV file

README.md

exportCSV.py

A simple utility download and save Firestore collections as CSV file.

Setup and usage

• In the Firebase console, open Settings > Service Accounts.
• Click Generate New Private Key, then confirm by clicking Generate Key.
• Save the JSON file with name service-account-file.json
• Install Firebase Admin SDK:

pip install firebase-admin

• Run ./exportCSV.sh [document name]

Command line options

• --help, -h Displays a help message

• --output OUTPUT, -o OUTPUT Sets an output file to export to. '[collection]-export-[time].csv' is used if this is left out

exportCSV.py

import argparse
import csv
from datetime import datetime
from os import PathLike
from pathlib import Path
from typing import Generator, Tuple, Union

import firebase_admin
from firebase_admin import firestore


def _generate_filename(collection_name) -> str:
    return f"{collection_name}-{datetime.now().strftime('%Y-%m-%d-%H-%M-%S')}.csv"

def export(
    file: Union[PathLike, str],
    collection: firestore.firestore.CollectionReference
    ) -> Generator[Tuple[str, int], None, None]:
    """
    Export a collection to given file.
    This is a generator function which yields a tuple each time a row is written to the csv file.
    The first element of the tuple is the id of last document written and the second is rows
    written to the CSV file so far.

    Note: To ensure the CSV file gets written to completely, the generator must be exhausted 
    """
    db = firestore.client()
    documents = collection.stream()

    with open(file, "w") as fp:
        try:
            first = next(documents)
        except StopIteration:
            return  # no rows
        
        first_obj = first.to_dict()

        writer = csv.DictWriter(fp, first_obj.keys())
        writer.writeheader()
        writer.writerow(first_obj)
        yield first.id, 1

        for count, document in enumerate(documents, 2):
            writer.writerow(document.to_dict())
            yield document.id, count

def _interactive_export(
    file: Union[PathLike, str],
    collection: firestore.firestore.CollectionReference
    ):

    print(f'Exporting collection "{collection.id}" to file "{file}"')

    for doc_name, doc_count in export(file, collection):
        print(f'Retrieved document "{doc_name}" ({doc_count})')

if __name__ == "__main__":
    firebase_admin.initialize_app()

    argument_parser = argparse.ArgumentParser("exportCSV")
    argument_parser.add_argument("collection", help="The collection to export")
    argument_parser.add_argument(
        "--output",
        "-o",
        required=False,
        help=(
            "Output file to export to. "
            "'[collection]-export-[time].csv' is used if this is left out"
        )
    )

    args = argument_parser.parse_args()
    
    file = Path(args.output or _generate_filename(args.collection))

    db = firestore.client()
    collection = db.collection(args.collection)

    _interactive_export(file, collection)

exportCSV.sh

#!/bin/bash

export GOOGLE_APPLICATION_CREDENTIALS="service-account-file.json"
venv/bin/python exportCSV.py $@