import os
import sys
import glob
import datetime
import json

import parse_pdf
import train_detail
import validate

# List all PDF files in the given directory
def list_pdf_files(directory):
    pdf_files = glob.glob(os.path.join(directory, '*.pdf'))
    return pdf_files

# Extracts date from PDF filename and returns its content
def get_schedule_card_data(filepath):
    filename = os.path.basename(filepath)
    date_str = filename[:4]
    date_object = datetime.datetime.now()
    if len(date_str) < 4:
        raise ValueError("Filename must have at least four characters")
        return None

    month_str = date_str[:2]
    day_str = date_str[2:]

    try:
        month = int(month_str)
        day = int(day_str)

        year = datetime.datetime.now().year
        date_object = datetime.datetime(year=year, month=month, day=day)
    except ValueError as e:
        print(f"Error parsing date: {e}")
        return None

    schedule_card_data = {
        "schedule_date": date_object,
        "schedule_data": parse_pdf.parse_pdf_file(filepath)
    }

    return schedule_card_data


# Loop through data and remove duplicate codes.

# Check for existing codes via OwlBoard API.
# Validate existing codes and submit issue if not correct.

# Use the train list to search for stopping pattern of any absent codes.


# Format file and commit to git

def main():
    # Check for arguments
    if len(sys.argv) > 1:
        directory = sys.argv[1]
    else:
        directory = os.getcwd()

    if not os.path.isdir(directory):
        print(f"Error: '{directory}' is not a valid directory")
        return

    pdf_files = list_pdf_files(directory)
    if len(pdf_files) == 0:
        print(f"Error: '{directory}' contains no PDF files")
        return
    else:
        print(f"Found {len(pdf_files)} PDF files")

    schedule_cards = []
    for pdf_file in pdf_files:
        schedule_cards.append(get_schedule_card_data(pdf_file))

    # Iterate over schedule cards and fetch train data
    trains = []
    existing_codes = []
    for schedule_card in schedule_cards:
        date = schedule_card['schedule_date']
        for schedule in schedule_card['schedule_data']:
            
            ## Skip over schedules which have no PIS Code
            if schedule['pis_code'] is None:
                continue
            
            ## Check if PIS code already processed, if so skip it.  Else add it to 'existing_codes'
            if schedule['pis_code'] in existing_codes:
                continue

            existing_codes.append(schedule['pis_code'])

            full_schedule = train_detail.find_gw_trains_by_headcode(schedule['headcode'], date)
            train = {
                'schedule_date': date,
                'schedule_headcode': schedule['headcode'],
                'schedule_location': schedule['location'],
                'schedule_time1': schedule['time0'],
                'schedule_time2': schedule['time1'],
                'schedule_pis_code': schedule['pis_code'],
                'timetable_entry': full_schedule,
            }

            ## Filter out values missing schedule details or with matching PIS Codes
            if not full_schedule:
                continue # Skip if timetable entry was not found

            if len(train['timetable_entry']) == 1 and train['schedule_pis_code'] == train['timetable_entry'][0].get('pis', {}).get('code'):
                continue # Skip if only one possible timetable entry exists, and PIS code matches OwlBoards suggestion

            trains.append(train)

    organised_for_processing = train_detail.organise_trains(trains)

    auto_matched = validate.filter_timetable_entries(organised_for_processing)


    # print(trains)
    out = open("output.txt", "w")
    out.write(json.dumps(auto_matched, indent=4, default=str))
    out.close()
    
if __name__ == "__main__":
    main()