import argparse
import csv
import gzip
import sys


def domains(path):
    with gzip.open(path, "rt", encoding="utf-8") as source:
        for line in source:
            domain = line.strip().lower().rstrip(".")
            if domain:
                yield domain

import datetime
import sqlite3

parser = argparse.ArgumentParser(description="Import a dated domain snapshot into SQLite")
parser.add_argument("file")
parser.add_argument("--database", required=True)
parser.add_argument("--snapshot-date", required=True)
args = parser.parse_args()
try:
    date = datetime.date.fromisoformat(args.snapshot_date).isoformat()
except ValueError:
    parser.error("snapshot-date must be YYYY-MM-DD")
connection = sqlite3.connect(args.database)
try:
    with connection:
        connection.execute("CREATE TABLE IF NOT EXISTS snapshots "
                           "(domain TEXT, observed_date TEXT, "
                           "PRIMARY KEY (domain, observed_date))")
        batch = []
        for domain in domains(args.file):
            batch.append((domain, date))
            if len(batch) == 1000:
                connection.executemany("INSERT OR IGNORE INTO snapshots VALUES (?, ?)", batch)
                batch.clear()
        connection.executemany("INSERT OR IGNORE INTO snapshots VALUES (?, ?)", batch)
    print("Rows for date:", connection.execute(
        "SELECT COUNT(*) FROM snapshots WHERE observed_date = ?", (date,)).fetchone()[0])
finally:
    connection.close()