def read_transactions_from_csv(f: TextIO, standardize_statement_record: Callable) -> list:

    reader = csv.DictReader(f)

    return sort_records([standardize_statement_record(row, reader.line_num) for row in reader])

def standardize_amex_record(row: Dict, line: int) -> Dict:

    """Turn an AMEX CSV row into a standard dict format representing a transaction."""

    return {

        'date': datetime.datetime.strptime(row['Date'], '%m/%d/%Y').date(),

        'amount': -1 * decimal.Decimal(row['Amount']),

        # Descriptions have too much noise, so taking just the start

        # significantly assists the fuzzy matching.

        'payee': remove_payee_junk(row['Description'] or '')[:20],

        'check_id': '',

        'line': line,

    }

def standardize_fr_record(row: Dict, line: int) -> Dict:

    return {

        'date': datetime.datetime.strptime(row['Date'], '%m/%d/%Y').date(),

        'amount': decimal.Decimal(row['Amount']),

        'payee': remove_payee_junk(row['Detail'] or '')[:20],

        'check_id': row['Serial Num'].lstrip('0'),

        'line': line,

    }

def standardize_beancount_record(row) -> Dict:  # type: ignore[no-untyped-def]

    """Turn a Beancount query result row into a standard dict representing a transaction."""

    return {

        'date': row.date,

        'amount': row.number_cost_position,

        'payee': remove_payee_junk(f'{row.payee or ""} {row.entity or ""} {row.narration or ""}'),

        'check_id': str(row.check_id or ''),

        'filename': row.filename,

        'line': row.line,

        'bank_statement': row.bank_statement,

    }

def format_record(records: list[dict]) -> str:

    if len(records) == 1:

        record = records[0]

        if record['payee'] and record['check_id']:

            output = f"{record['date'].isoformat()}: {record['amount']:11,.2f} {record['payee'][:25]} #{record['check_id']}".ljust(59)

        elif record['payee']:

            output = f"{record['date'].isoformat()}: {record['amount']:11,.2f} {record['payee'][:35]}".ljust(59)

        else:

            output = f"{record['date'].isoformat()}: {record['amount']:11,.2f} #{record['check_id']}".ljust(59)

        return output

    else:

        raise NotImplementedError

def sort_records(records: List) -> List:

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

def first_word_exact_match(a, b):

    if len(a) == 0 or len(b) == 0:

        return 0

    first_a = a.split()[0].strip()

    first_b = b.split()[0].strip()

    if first_a.casefold() == first_b.casefold():

        return min(1.0, 0.2 * len(first_a))

    else:

        return 0;

def payee_match(a, b):

    fuzzy_match = fuzz.token_set_ratio(a, b) / 100.00

    first_word_match = first_word_exact_match(a, b)

    return max(fuzzy_match, first_word_match)

def records_match(r1: Dict, r2: Dict) -> Tuple[bool, str]:

    """Do these records represent the same transaction?"""

    date_score = date_proximity(r1['date'], r2['date'])

    if r1['date'] == r2['date']:

        date_message = ''

    elif date_score > 0.0:

        diff = abs((r1['date'] - r2['date']).days)

        date_message = f'+/- {diff} days'

    else:

        date_message = 'date mismatch'

    if r1['amount'] == r2['amount']:

        amount_score, amount_message = 2.0, ''

    else:

        amount_score, amount_message = 0.0, 'amount mismatch'

    # We never consider payee if there's a check_id in the books.

    check_message = ''

    payee_message = ''

    if r2['check_id']:

        payee_score = 0.0

        if r1['check_id'] and r2['check_id'] and r1['check_id'] == r2['check_id']:

            check_score = 1.0

        else:

            check_message = 'check-id mismatch'

            check_score = 0.0

    else:

        check_score = 0.0

        payee_score = payee_match(r1['payee'], r2['payee'])

        if payee_score > 0.8:

            payee_message = ''

        elif payee_score > 0.4:

            payee_message = 'partial payee match'

        else:

            payee_message = 'payee mismatch'

    overall_score = (date_score + amount_score + check_score + payee_score) / 4

    overall_message = [m for m in [date_message, amount_message, check_message, payee_message] if m]

    return overall_score, overall_message

def match_statement_and_books(statement_trans: list, books_trans: list):

    """

    Runs through all the statement transactions to find a matching transaction

    in the books. If found, the books transaction is marked off so that it can

    only be matched once. Some transactions will be matched, some will be on the

    statement but not the books and some on the books but not the statement.

    """

    matches = []

    # We need a realised list and should be a copy so we can safely delete

    # items.

    books_trans = list(books_trans)

    for r1 in statement_trans:

        best_match_score = 0

        best_match_index = None

        best_match_note = ''

        matches_found = 0

        for i, r2 in enumerate(books_trans):

            score, note = records_match(r1, r2)

            if score >= 0.5 and score >= best_match_score:

                matches_found += 1

                best_match_score = score

                best_match_index = i

                best_match_note = note

        if best_match_score > 0.5 and matches_found == 1 and 'check-id mismatch' not in best_match_note or best_match_score > 0.8:

            if best_match_score <= 0.8:

                best_match_note.append('only one decent match')

            matches.append(([r1], [books_trans[best_match_index]], best_match_note))

            del books_trans[best_match_index]

        else:

            matches.append(([r1], [], ['no match']))

    for r2 in books_trans:

        matches.append(([], [r2], ['no match']))

    return matches

def format_matches(matches, csv_statement: str, show_reconciled_matches):

    match_output = []

    for r1, r2, note in matches:

        note = ', '.join(note)

        note = ': ' + note if note else note

        if r1 and r2:

            if show_reconciled_matches:

                match_output.append([r1[0]['date'], f'{format_record(r1)}  →  {format_record(r2)}  ✓ Matched{note}'])

        elif r1:

            match_output.append([r1[0]['date'], f'{format_record(r1)}  →  {" ":^59}  ✗ NOT IN BOOKS ({os.path.basename(csv_statement)}:{r1[0]["line"]})'])

        else:

            match_output.append([r2[0]['date'], f'{" ":^59}  →  {format_record(r2)}  ✗ NOT ON STATEMENT ({os.path.basename(r2[0]["filename"])}:{r2[0]["line"]})'])

    return match_output

def date_proximity(d1, d2):

    diff = abs((d1 - d2).days)

    if diff > 60:

        return 0

    else:

        return 1.0 - (diff / 60.0)

def metadata_for_match(match, statement_filename, csv_filename):

    # Can we really ever have multiple statement entries? Probably not.

    statement_filename = get_repo_relative_path(statement_filename)

    csv_filename = get_repo_relative_path(csv_filename)

    metadata = []

    statement_entries, books_entries, _ = match

    for books_entry in books_entries:

        for statement_entry in statement_entries:

            if not books_entry['bank_statement']:

                metadata.append((books_entry['filename'], books_entry['line'], f'    bank-statement: {statement_filename}'))

                metadata.append((books_entry['filename'], books_entry['line'], f'    bank-statement-csv: {csv_filename}:{statement_entry["line"]}'))

    return metadata

# TODO: Is there a way to pull the side-effecting code out of this function?

def write_metadata_to_books(metadata_to_apply: List[Tuple[str, int, str]]) -> None:

    """Insert reconciliation metadata in the books files.

    Takes a list of edits to make as tuples of form (filename, lineno, metadata):

    [

        ('2021/main.beancount', 4245, '    bank-statement: statement.pdf'),

        ('2021/main.beancount', 1057, '    bank-statement: statement.pdf'),

        ('2021/payroll.beancount', 257, '    bank-statement: statement.pdf'),

        ...,

    ]

    """

    file_contents: dict[str, list] = {}

    file_offsets: dict[str, int] = collections.defaultdict(int)

    # Load each books file into memory and insert the relevant metadata lines.

    # Line numbers change as we do this, so we keep track of the offset for each

    # file. Changes must be sorted by line number first or else the offsets will

    # break because we're jumping around making edits.

    for filename, line, metadata in sorted(metadata_to_apply):

        if filename not in file_contents:

            with open(filename, 'r') as f:

                file_contents[filename] = f.readlines()

        # Insert is inefficient, but fast enough for now in practise.

        file_contents[filename].insert(line + file_offsets[filename], metadata.rstrip() + '\n')

        file_offsets[filename] += 1

    # Writes each updated file back to disk.

    for filename, contents in file_contents.items():

        with open(filename, 'w') as f:

            f.writelines(contents)

            print(f'Wrote {filename}.')

def get_repo_relative_path(path):

    return os.path.relpath(path, start=os.getenv('CONSERVANCY_REPOSITORY'))