#!/usr/bin/env python3
"""
Foundation Accounts pipeline — Step 3: convert bulk_runner responses → CSV.

Reads `foundation_vendors_responses.json`, and (optionally) merges the Main
Client Contact block (columns M–S) from the source spreadsheet by account name.
That contact/address block has no ows-account endpoint, so it is carried here
for manual follow-up rather than sent during account creation.

Basic (vendor_id/uuid/name/link only):
    uv run python src/pipelines/foundation_accounts/3_postprocess.py <responses.json>

With contact block:
    uv run python src/pipelines/foundation_accounts/3_postprocess.py <responses.json> \\
        --runner-csv data/input/foundation_vendors.csv \\
        --source-csv "data/prepare/Foundation Accounts - Filled in Orchard Spreadsheet.csv"
"""

import argparse
import csv
import json
import sys

# Source column indices for the Main Client Contact block (M–S).
CONTACT_COLS: dict[str, int] = {
    'main_client_contact_name': 12,
    'main_client_contact_email': 13,
    'street_address': 14,
    'city': 15,
    'contact_country': 16,
    'zipcode': 17,
    'phone': 18,
}

OUTPUT_COLS_BASE = ['name', 'vendor_id', 'vendor_uuid', 'link']
OUTPUT_COLS_FULL = OUTPUT_COLS_BASE + list(CONTACT_COLS.keys())


def load_runner_csv(path: str) -> dict[int, str]:
    """Row num (1-based) → account name."""
    with open(path, newline='', encoding='utf-8-sig') as f:
        rows = list(csv.DictReader(f))
    return {i + 1: row['name'].strip() for i, row in enumerate(rows)}


def load_source_contacts(path: str) -> dict[str, dict[str, str]]:
    """Normalized account name → Main Client Contact block (columns M–S)."""
    with open(path, newline='', encoding='utf-8-sig') as f:
        rows = list(csv.reader(f))

    result: dict[str, dict[str, str]] = {}
    for row in rows[1:]:  # row 0 is the header
        if not row or not row[0].strip():
            continue
        name = row[0].strip()
        result[name.lower()] = {field: (row[idx].strip() if len(row) > idx else '') for field, idx in CONTACT_COLS.items()}
    return result


def main() -> None:
    parser = argparse.ArgumentParser(description='Foundation Accounts Step 3: responses JSON → CSV')
    parser.add_argument('responses', help='Path to foundation_vendors_responses.json from bulk_runner.py')
    parser.add_argument('--runner-csv', help='foundation_vendors.csv used for the run (row → name mapping)')
    parser.add_argument('--source-csv', help='Source spreadsheet (for the Main Client Contact block, columns M–S)')
    args = parser.parse_args()

    full_mode = bool(args.runner_csv or args.source_csv)
    if full_mode and not (args.runner_csv and args.source_csv):
        print('Error: --runner-csv and --source-csv must be used together', file=sys.stderr)
        sys.exit(1)

    output_path = args.responses.replace('.json', '_ids.csv')

    with open(args.responses) as f:
        data = json.load(f)

    row_to_name = load_runner_csv(args.runner_csv) if full_mode else {}
    contacts = load_source_contacts(args.source_csv) if full_mode else {}
    output_cols = OUTPUT_COLS_FULL if full_mode else OUTPUT_COLS_BASE

    unmatched: list[str] = []
    with open(output_path, 'w', newline='') as f:
        writer = csv.DictWriter(f, fieldnames=output_cols)
        writer.writeheader()
        for entry in data:
            r = entry.get('response') or {}
            row_num = entry.get('row')
            vendor_id = r.get('vendor_id')

            # Prefer the runner CSV row → name mapping over the response name (handles drift).
            name = row_to_name.get(row_num, r.get('name', '')).strip()

            out_row: dict[str, object] = {
                'name': name,
                'vendor_id': vendor_id,
                'vendor_uuid': r.get('vendor_uuid'),
                'link': f'https://oa.theorchard.com/cont_mgmt/view_vendor.php?vendor_id={vendor_id}',
            }
            if full_mode:
                contact = contacts.get(name.lower())
                if contact is None:
                    unmatched.append(f'row={row_num} name={name!r}')
                    contact = {field: '' for field in CONTACT_COLS}
                out_row.update(contact)
            writer.writerow(out_row)

    print(f'Written {len(data)} rows to {output_path}')
    if unmatched:
        print(f'WARNING: {len(unmatched)} rows not matched in source CSV:')
        for m in unmatched:
            print(f'  {m}')


if __name__ == '__main__':
    main()