Files
@ a2ee9c73fe38
Branch filter:
Location: NPO-Accounting/conservancy_beancount/conservancy_beancount/data.py
a2ee9c73fe38
15.9 KiB
text/x-python
ranges: Start module.
The ledger report wants to use this functionality, so make it available in a
higher-level module.
I took the opportunity to clean up a lot of the surrounding type
declarations. It is less flexible, since it relies on the static list of
types in RangeT, but I don't think the other method actually worked at all
except by cheating with generic Any.
The ledger report wants to use this functionality, so make it available in a
higher-level module.
I took the opportunity to clean up a lot of the surrounding type
declarations. It is less flexible, since it relies on the static list of
types in RangeT, but I don't think the other method actually worked at all
except by cheating with generic Any.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 | """Enhanced Beancount data structures for Conservancy
The classes in this module are interface-compatible with Beancount's core data
structures, and provide additional business logic that we want to use
throughout Conservancy tools.
"""
# Copyright © 2020 Brett Smith
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Affero General Public License for more details.
#
# You should have received a copy of the GNU Affero General Public License
# along with this program. If not, see <https://www.gnu.org/licenses/>.
import collections
import datetime
import decimal
import functools
import re
from beancount.core import account as bc_account
from beancount.core import amount as bc_amount
from beancount.core import convert as bc_convert
from beancount.core import position as bc_position
from typing import (
cast,
overload,
Callable,
Hashable,
Iterable,
Iterator,
MutableMapping,
Optional,
Sequence,
TypeVar,
Union,
)
from .beancount_types import (
Directive,
MetaKey,
MetaValue,
Posting as BasePosting,
Transaction,
)
DecimalCompat = Union[decimal.Decimal, int]
LINK_METADATA = frozenset([
'approval',
'check',
'contract',
'invoice',
'purchase-order',
'receipt',
'rt-id',
'statement',
])
class Account(str):
"""Account name string
This is a string that names an account, like Assets:Bank:Checking
or Income:Donations. This class provides additional methods for common
account name parsing and queries.
"""
__slots__ = ()
SEP = bc_account.sep
def is_cash_equivalent(self) -> bool:
return (
self.is_under('Assets:') is not None
and self.is_under('Assets:Prepaid', 'Assets:Receivable') is None
)
def is_checking(self) -> bool:
return self.is_cash_equivalent() and ':Check' in self
def is_credit_card(self) -> bool:
return self.is_under('Liabilities:CreditCard') is not None
def is_opening_equity(self) -> bool:
return self.is_under('Equity:Funds', 'Equity:OpeningBalance') is not None
def is_under(self, *acct_seq: str) -> Optional[str]:
"""Return a match if this account is "under" a part of the hierarchy
Pass in any number of account name strings as arguments. If this
account is under one of those strings in the account hierarchy, the
first matching string will be returned. Otherwise, None is returned.
You can use the return value of this method as a boolean if you don't
care which account string is matched.
An account is considered to be under itself:
Account('Expenses:Tax').is_under('Expenses:Tax') # returns 'Expenses:Tax'
To do a "strictly under" search, end your search strings with colons:
Account('Expenses:Tax').is_under('Expenses:Tax:') # returns None
Account('Expenses:Tax').is_under('Expenses:') # returns 'Expenses:'
This method does check that all the account boundaries match:
Account('Expenses:Tax').is_under('Exp') # returns None
"""
for prefix in acct_seq:
if self.startswith(prefix) and (
prefix.endswith(self.SEP)
or self == prefix
or self[len(prefix)] == self.SEP
):
return prefix
return None
def _find_part_slice(self, index: int) -> slice:
if index < 0:
raise ValueError(f"bad part index {index!r}")
start = 0
for _ in range(index):
try:
start = self.index(self.SEP, start) + 1
except ValueError:
raise IndexError("part index {index!r} out of range") from None
try:
stop = self.index(self.SEP, start + 1)
except ValueError:
stop = len(self)
return slice(start, stop)
def count_parts(self) -> int:
return self.count(self.SEP) + 1
@overload
def slice_parts(self, start: None=None, stop: None=None) -> Sequence[str]: ...
@overload
def slice_parts(self, start: slice, stop: None=None) -> Sequence[str]: ...
@overload
def slice_parts(self, start: int, stop: int) -> Sequence[str]: ...
@overload
def slice_parts(self, start: int, stop: None=None) -> str: ...
def slice_parts(self,
start: Optional[Union[int, slice]]=None,
stop: Optional[int]=None,
) -> Sequence[str]:
"""Slice the account parts like they were a list
Given a single index, return that part of the account name as a string.
Otherwise, return a list of part names sliced according to the arguments.
"""
if start is None:
part_slice = slice(None)
elif isinstance(start, slice):
part_slice = start
elif stop is None:
return self[self._find_part_slice(start)]
else:
part_slice = slice(start, stop)
return self.split(self.SEP)[part_slice]
def root_part(self, count: int=1) -> str:
"""Return the first part(s) of the account name as a string"""
try:
stop = self._find_part_slice(count - 1).stop
except IndexError:
return self
else:
return self[:stop]
class Amount(bc_amount.Amount):
"""Beancount amount after processing
Beancount's native Amount class declares number to be Optional[Decimal],
because the number is None when Beancount first parses a posting that does
not have an amount, because the user wants it to be automatically balanced.
As part of the loading process, Beancount replaces those None numbers
with the calculated amount, so it will always be a Decimal. This class
overrides the type declaration accordingly, so the type checker knows
that our code doesn't have to consider the possibility that number is
None.
"""
number: decimal.Decimal
# beancount.core._Amount is the plain namedtuple.
# beancore.core.Amount adds instance methods to it.
# b.c.Amount.__New__ calls `b.c._Amount.__new__`, which confuses type
# checking. See <https://github.com/python/mypy/issues/1279>.
# It works fine if you use super(), which is better practice anyway.
# So we override __new__ just to call _Amount.__new__ this way.
def __new__(cls, number: decimal.Decimal, currency: str) -> 'Amount':
return super(bc_amount.Amount, Amount).__new__(cls, number, currency)
class Metadata(MutableMapping[MetaKey, MetaValue]):
"""Transaction or posting metadata
This class wraps a Beancount metadata dictionary with additional methods
for common parsing and query tasks.
"""
__slots__ = ('meta',)
_HUMAN_NAMES: MutableMapping[MetaKey, str] = {
# Initialize this dict with special cases.
# We use it as a cache for other metadata names as they're queried.
'check-id': 'Check Number',
'paypal-id': 'PayPal ID',
'rt-id': 'Ticket',
}
def __init__(self, source: MutableMapping[MetaKey, MetaValue]) -> None:
self.meta = source
def __iter__(self) -> Iterator[MetaKey]:
return iter(self.meta)
def __len__(self) -> int:
return len(self.meta)
def __getitem__(self, key: MetaKey) -> MetaValue:
return self.meta[key]
def __setitem__(self, key: MetaKey, value: MetaValue) -> None:
self.meta[key] = value
def __delitem__(self, key: MetaKey) -> None:
del self.meta[key]
def get_links(self, key: MetaKey) -> Sequence[str]:
try:
value = self.meta[key]
except KeyError:
return ()
if isinstance(value, str):
return value.split()
else:
raise TypeError("{} metadata is a {}, not str".format(
key, type(value).__name__,
))
@overload
def first_link(self, key: MetaKey, default: None=None) -> Optional[str]: ...
@overload
def first_link(self, key: MetaKey, default: str) -> str: ...
def first_link(self, key: MetaKey, default: Optional[str]=None) -> Optional[str]:
try:
return self.get_links(key)[0]
except (IndexError, TypeError):
return default
@classmethod
def human_name(cls, key: MetaKey) -> str:
"""Return the "human" version of a metadata name
This is usually the metadata key with punctuation replaced with spaces,
and then titlecased, with a few special cases. The return value is
suitable for using in reports.
"""
try:
retval = cls._HUMAN_NAMES[key]
except KeyError:
retval = key.replace('-', ' ').title()
retval = re.sub(r'\bId$', 'ID', retval)
cls._HUMAN_NAMES[key] = retval
return retval
class PostingMeta(Metadata):
"""Combined access to posting metadata with its parent transaction metadata
This lets you access posting metadata through a single dict-like object.
If you try to look up metadata that doesn't exist on the posting, it will
look for the value in the parent transaction metadata instead.
You can set and delete metadata as well. Changes only affect the metadata
of the posting, never the transaction. Changes are propagated to the
underlying Beancount data structures.
Functionally, you can think of this as identical to:
collections.ChainMap(post.meta, txn.meta)
Under the hood, this class does a little extra work to avoid creating
posting metadata if it doesn't have to.
"""
__slots__ = ('txn', 'index', 'post')
def __init__(self,
txn: Transaction,
index: int,
post: Optional[BasePosting]=None,
) -> None:
if post is None:
post = txn.postings[index]
self.txn = txn
self.index = index
self.post = post
if post.meta is None:
self.meta = self.txn.meta
else:
self.meta = collections.ChainMap(post.meta, txn.meta)
def __getitem__(self, key: MetaKey) -> MetaValue:
try:
return super().__getitem__(key)
except KeyError:
if key == 'entity' and self.txn.payee is not None:
return self.txn.payee
else:
raise
def __setitem__(self, key: MetaKey, value: MetaValue) -> None:
if self.post.meta is None:
self.post = self.post._replace(meta={key: value})
self.txn.postings[self.index] = self.post
# mypy complains that self.post.meta could be None, but we know
# from two lines up that it's not.
self.meta = collections.ChainMap(self.post.meta, self.txn.meta) # type:ignore[arg-type]
else:
super().__setitem__(key, value)
def __delitem__(self, key: MetaKey) -> None:
if self.post.meta is None:
raise KeyError(key)
else:
super().__delitem__(key)
# This is arguably cheating a litttle bit, but I'd argue the date of
# the parent transaction still qualifies as posting metadata, and
# it's something we want to access so often it's good to have it
# within easy reach.
@property
def date(self) -> datetime.date:
return self.txn.date
class Posting(BasePosting):
"""Enhanced Posting objects
This class is a subclass of Beancount's native Posting class where
specific fields are replaced with enhanced versions:
* The `account` field is an Account object
* The `units` field is our Amount object (which simply declares that the
number is always a Decimal—see that docstring for details)
* The `meta` field is a PostingMeta object
"""
__slots__ = ()
account: Account
units: Amount
cost: Optional[bc_position.Cost]
# mypy correctly complains that our MutableMapping is not compatible
# with Beancount's meta type declaration of Optional[Dict]. IMO
# Beancount's type declaration is a smidge too specific: I think its type
# declaration should also use MutableMapping, because it would be very
# unusual for code to specifically require a Dict over that.
# If it did, this declaration would pass without issue.
meta: PostingMeta # type:ignore[assignment]
@classmethod
def from_beancount(cls,
txn: Transaction,
index: int,
post: Optional[BasePosting]=None,
) -> 'Posting':
if post is None:
post = txn.postings[index]
return cls(
Account(post.account),
*post[1:5],
# see rationale above about Posting.meta
PostingMeta(txn, index, post), # type:ignore[arg-type]
)
@classmethod
def from_txn(cls, txn: Transaction) -> Iterator['Posting']:
"""Yield an enhanced Posting object for every posting in the transaction"""
for index, post in enumerate(txn.postings):
yield cls.from_beancount(txn, index, post)
@classmethod
def from_entries(cls, entries: Iterable[Directive]) -> Iterator['Posting']:
"""Yield an enhanced Posting object for every posting in these entries"""
for entry in entries:
# Because Beancount's own Transaction class isn't type-checkable,
# we can't statically check this. Might as well rely on duck
# typing while we're at it: just try to yield postings from
# everything, and ignore entries that lack a postings attribute.
try:
yield from cls.from_txn(entry) # type:ignore[arg-type]
except AttributeError:
pass
_KT = TypeVar('_KT', bound=Hashable)
_VT = TypeVar('_VT')
class _SizedDict(collections.OrderedDict, MutableMapping[_KT, _VT]):
def __init__(self, maxsize: int=128) -> None:
self.maxsize = maxsize
super().__init__()
def __setitem__(self, key: _KT, value: _VT) -> None:
super().__setitem__(key, value)
for _ in range(self.maxsize, len(self)):
self.popitem(last=False)
def balance_of(txn: Transaction,
*preds: Callable[[Account], Optional[bool]],
) -> Amount:
"""Return the balance of specified postings in a transaction.
Given a transaction and a series of account predicates, balance_of
returns the balance of the amounts of all postings with accounts that
match any of the predicates.
balance_of uses the "weight" of each posting, so the return value will
use the currency of the postings' cost when available.
"""
match_posts = [post for post in Posting.from_txn(txn)
if any(pred(post.account) for pred in preds)]
number = decimal.Decimal(0)
if not match_posts:
currency = ''
else:
weights: Sequence[Amount] = [
bc_convert.get_weight(post) for post in match_posts
]
number = sum((wt.number for wt in weights), number)
currency = weights[0].currency
return Amount(number, currency)
_opening_balance_cache: MutableMapping[str, bool] = _SizedDict()
def is_opening_balance_txn(txn: Transaction) -> bool:
key = '\0'.join(
f'{post.account}={post.units}' for post in txn.postings
)
try:
return _opening_balance_cache[key]
except KeyError:
pass
opening_equity = balance_of(txn, Account.is_opening_equity)
if not opening_equity.currency:
retval = False
else:
rest = balance_of(txn, lambda acct: not acct.is_opening_equity())
if not rest.currency:
retval = False
else:
retval = abs(opening_equity.number + rest.number) < decimal.Decimal('.01')
_opening_balance_cache[key] = retval
return retval
|