import subprocess
import json
import time
import csv
from datetime import datetime
import argparse
import base64

# Parse command-line arguments for data logging parameters
parser = argparse.ArgumentParser(description='Parameters for data logging')
parser.add_argument('-mj', '--maxjobs', type=int, required=True, help="Max Jobs parameter value")
parser.add_argument('-c', '--churn', type=int, required=True, help="Churn parameter value")
parser.add_argument('-i', '--interval', type=int, required=True, help="Interval parameter value")
parser.add_argument('-u', '--username', type=str, required=True, help="Username for authentication")
parser.add_argument('-p', '--password', type=str, required=True, help="Password for authentication")
parser.add_argument('-db', '--filterByDatabase', type=str, required=False, help="Filter by specific database ID (optional)")
parser.add_argument('-s', '--serverURL', type=str, default="http://localhost:5984/", help="Server URL (optional, default is http://localhost:5984/)")
args = parser.parse_args()

# Encode the username and password into base64
authentication_token = base64.b64encode(f"{args.username}:{args.password}".encode()).decode()

# Send a curl request to get scheduler data from CouchDB
cmd = ["curl", "--location", f'{args.serverURL}_scheduler/docs', 
        "--header", f'Authorization: Basic {authentication_token}']

while True:
    result = subprocess.run(cmd, stdout=subprocess.PIPE)
    data = json.loads(result.stdout)
    
    with open('output.csv', 'a', newline='') as f:
        writer = csv.writer(f)
        for doc in data['docs']:
            doc_id = doc['doc_id']
            if args.filterByDatabase:
                if doc_id != args.filterByDatabase:
                    continue
            else:
                print("doc_id is", doc_id)

            docs_written = doc['info']['docs_written']
            start_time = datetime.strptime(doc['start_time'], "%Y-%m-%dT%H:%M:%SZ")
            time_elapsed = -(datetime.now() - start_time).total_seconds()
            speed = docs_written/time_elapsed

            # Log the speed of document processing
            print("Speed is", speed, "documents per second")
            writer.writerow([datetime.now(), doc_id, docs_written, speed, args.maxjobs, args.churn, args.interval])

    # Adjust time between iterations here
    time.sleep(15)