hpos-data/scripts/user_collection.py

import os
import firebase_admin
from firebase_admin import credentials
from firebase_admin import firestore
from google.cloud.firestore_v1.base_query import FieldFilter
import pandas as pd
#import datetime
import sys
from datetime import datetime
import matplotlib.pyplot as plt

cred = credentials.Certificate(os.getcwd() + '/keys/hpos-prod-firebase-adminsdk.json')  # Replace with your own service account key path
firebase_admin.initialize_app(cred)

db = firestore.client()

patient_collection = db.collection("patientData")
test_collection = db.collection("testData")
start_date = sys.argv[1]
end_date = sys.argv[2]

query = patient_collection #.where(filter=FieldFilter("createdAt", ">=", start_date)).where(filter=FieldFilter("createdAt", "<", end_date))
patient_docs = query.stream()

data = []
for patient_doc in patient_docs:
    patient_data = patient_doc.to_dict()
    data.append(patient_data)

df = pd.DataFrame(data)
print(df)
print(df.size)

print("duplicates", len(df['_id']) - len(df['_id'].drop_duplicates()))

df = df.sort_values('bloodGroup', na_position='first', ascending=False).drop_duplicates('_id').sort_index()

output_filename =  f'users_{datetime.today().strftime("%d_%m_%Y_%H_%M")}.xlsx'
df.to_excel(output_filename, index=False)

downloads_dir = os.path.join(os.path.expanduser("~"), "Downloads")
output_path = os.path.join(downloads_dir, output_filename)
df.to_excel(output_path, index=False)

print(f"Data saved to '{output_path}'")

firebase_admin.delete_app(firebase_admin.get_app())