import sys
import json

# Check if the minimum number of arguments is provided
if len(sys.argv) < 3:
    print("Usage: python script.py <input_file> <output_file> [timestamp_threshold]", file=sys.stderr)
    sys.exit(1)

# Assign command-line arguments to variables
input_file_path = sys.argv[1]
output_file_path = sys.argv[2]

# Optional timestamp threshold
timestamp_threshold = None
if len(sys.argv) >= 4:
    try:
        timestamp_threshold = int(sys.argv[3])
    except ValueError:
        print("Timestamp threshold must be an integer", file=sys.stderr)
        sys.exit(1)

# Initialize a list to hold valid JSON objects
json_objects = []

# Initialize counters
skipped_invalid = 0
skipped_below_threshold = 0

# Open the input file and read lines
with open(input_file_path, 'r') as input_file:
    for line in input_file:
        try:
            # Try to parse the line as JSON
            json_object = json.loads(line)
        except json.JSONDecodeError:
            skipped_invalid += 1
            continue

        timestamp = json_object.get('Timestamp', None)
        if timestamp is None or not isinstance(timestamp, int):
            skipped_invalid += 1
            continue

        # Skip records below or equal to the threshold
        if timestamp_threshold is not None and timestamp <= timestamp_threshold:
            skipped_below_threshold += 1
            continue

        # Add valid object
        json_objects.append(json_object)

# Sort the list of JSON objects by the 'Timestamp' field
json_objects_sorted = sorted(json_objects, key=lambda x: x['Timestamp'])

# Remove duplicates by 'Timestamp' using a dictionary to preserve order
json_objects_unique = {}
for obj in json_objects_sorted:
    json_objects_unique[obj['Timestamp']] = obj

# Write the unique JSON objects to the output file
with open(output_file_path, 'w') as output_file:
    for obj in json_objects_unique.values():
        json_line = json.dumps(obj)
        output_file.write(json_line + '\n')

# Print summary to stderr
total_written = len(json_objects_unique)
print(f"Processing complete. {total_written} records written to {output_file_path}", file=sys.stderr)
print(f"Skipped {skipped_invalid} records: invalid JSON or missing/invalid Timestamp", file=sys.stderr)
print(f"Skipped {skipped_below_threshold} records: Timestamp ≤ threshold ({timestamp_threshold})", file=sys.stderr)