def format_multirecord(r1s: List[dict], r2s: List[dict], note: str) -> List[list]:

    """Generates output lines for one statement:multiple books transaction match."""

    assert len(r1s) == 1

    assert len(r2s) > 1

    match_output = []

    match_output.append([r1s[0]['date'], f'{format_record(r1s[0])}  →  {format_record(r2s[0])}  ✓ Matched{note}'])

    for r2 in r2s[1:]:

        match_output.append([r1s[0]['date'], f'{r1s[0]["date"].isoformat()}:             ↳                                    →  {format_record(r2)}  ✓ Matched{note}'])

    return match_output

def sort_records(records: List) -> List:

    return sorted(records, key=lambda x: (x['date'], x['amount']))

def first_word_exact_match(a: str, b: str) -> float:

    """Score a payee match based first word.

    We get a whole lot of good matches this way. Helps in the

    situation where the first word or two of a transaction description

    is useful and the rest is garbage.

    """

def parse_arguments(argv: List[str]) -> argparse.Namespace:

    parser = argparse.ArgumentParser(prog=PROGNAME, description='Reconciliation helper')

    cliutil.add_version_argument(parser)

    cliutil.add_loglevel_argument(parser)

    parser.add_argument('--beancount-file', required=True, type=parse_path)

    parser.add_argument('--csv-statement', required=True, type=parse_repo_relative_path)

    parser.add_argument('--bank-statement', required=True, type=parse_repo_relative_path)

    parser.add_argument('--account', required=True, help='eg. Liabilities:CreditCard:AMEX')

    # parser.add_argument('--report-group-regex')

    parser.add_argument('--show-reconciled-matches', action='store_true')

    parser.add_argument('--non-interactive', action='store_true', help="Don't prompt to write to the books")    # parser.add_argument('--statement-balance', type=parse_decimal_with_separator, required=True, help="A.K.A \"cleared balance\" taken from the end of the period on the PDF statement. Required because CSV statements don't include final or running totals")

    args = parser.parse_args(args=argv)

    return args

def totals(matches: List[Tuple[List, List, List]]) -> Tuple[decimal.Decimal, decimal.Decimal, decimal.Decimal]:

    """Calculate the totals of transactions matched/not-matched."""

    total_matched = decimal.Decimal(0)

    total_missing_from_books = decimal.Decimal(0)

    total_missing_from_statement = decimal.Decimal(0)

    for statement_entries, books_entries, _ in matches:

        if statement_entries and books_entries:

            total_matched += sum(c['amount'] for c in statement_entries)

        sample = f.read(200)

        # Validate should return true/false and a message.

        validate_csv(sample)

        f.seek(0)

        # TODO: Needs a custom read_transactions_from_csv for each of AMEX and

        # FR since AMEX has a header row and FR doesn't.

        statement_trans = read_csv(f)

    # Dates are taken from the beginning/end of the statement.

    begin_date = statement_trans[0]['date']

    end_date = statement_trans[-1]['date']

    # Query for the Beancount books data for this above period.

    #

    # There are pros and cons for using Beancount's in-memory entries

    # list directly and also for using Beancount Query Language (BQL)

    # to get a list of transactions? Using BQL because it's

    # convenient, but we don't have access to the full transaction

    # entry objects. Feels a bit strange that these approaches are so

    # disconnected.

    #

    # beancount.query.query_compile.compile() and

    # beancount.query.query_execute.filter_entries() look useful in this respect,

    # but I'm not clear on how to use compile(). An example would help.

import datetime

import decimal

import io

import os

import tempfile

import textwrap

from conservancy_beancount.reconcile.statement_reconciler import (

    date_proximity,

    format_output,

    match_statement_and_books,

    metadata_for_match,

    payee_match,

    read_amex_csv,

    read_fr_csv,

    remove_duplicate_words,

    remove_payee_junk,

    subset_match,

    totals,

    write_metadata_to_books,

)

# These data structures represent individual transactions as taken from the

# statement ("S") or the books ("B").

# Statement transaction examples.

S1 = {

    'date': datetime.date(2022, 1, 1),

    'amount': decimal.Decimal('10.00'),

            'line': 3,

        },

    ]

    assert read_fr_csv(io.StringIO(CSV)) == expected

def test_format_output():

    statement = [S1]

    books = [B1]

    matches, _, _ = match_statement_and_books(statement, books)

    output = format_output(matches, datetime.date(2022, 1, 1), datetime.date(2022, 2, 1), 'test.csv', True)

    assert '2022-01-01:       10.00 Patreon         / Patreon   / 12345  →  2022-01-01:       10.00 Patreon                              ✓ Matched' in output