Changeset - 80dace59b385
[Not reviewed]
1 0 1
Ben Sturmfels (bsturmfels) - 2 years ago 2022-02-22 21:18:52
ben@sturm.com.au
reconcile: Rename statement reconciler.
1 file changed with 7 insertions and 0 deletions:
0 comments (0 inline, 0 general)
conservancy_beancount/reconcile/statement_reconciler.py
Show inline comments
 
file renamed from conservancy_beancount/reconcile/prototype_amex_reconciler.py to conservancy_beancount/reconcile/statement_reconciler.py
...
 
@@ -197,85 +197,92 @@ 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 = ''
 
    # Sometimes we get unrelated numbers in the statement column with check-ids,
 
    # so we can't match based on the existence of a statement check-id.
 
    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)
 

	
 
    # We can delete the matched books trans, but seems not a good idea to delete
 
    # while iterating through statement_trans. Instead pushing onto a separate
 
    # list.
 
    remaining_statement_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
 

	
 

	
0 comments (0 inline, 0 general)