conservancy_beancount/conservancy_beancount/data.py

"""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 decimal
import operator

from beancount.core import account as bc_account

from typing import (
    cast,
    Callable,
    Iterable,
    Iterator,
    MutableMapping,
    Optional,
    Sequence,
    Union,
)

from .beancount_types import (
    MetaKey,
    MetaValue,
    Posting as BasePosting,
    Transaction,
)

DecimalCompat = Union[decimal.Decimal, int]

LINK_METADATA = frozenset([
    'approval',
    'check',
    'contract',
    'invoice',
    'purchase-order',
    'receipt',
    '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


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',)

    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__,
            ))


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)


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 `meta` field is a PostingMeta object
    """
    __slots__ = ()

    account: Account
    # 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: Metadata  # type:ignore[assignment]

    def _compare_amount(self,
                        op: Callable[[decimal.Decimal], decimal.Decimal],
                        threshold: DecimalCompat,
                        default: Optional[bool],
    ) -> Optional[bool]:
        if self.units.number is None:
            return default
        else:
            return op(self.units.number) > threshold

    def is_credit(self,
                  threshold: DecimalCompat=0,
                  default: Optional[bool]=None,
    ) -> Optional[bool]:
        return self._compare_amount(operator.pos, threshold, default)

    def is_debit(self,
                  threshold: DecimalCompat=0,
                  default: Optional[bool]=None,
    ) -> Optional[bool]:
        return self._compare_amount(operator.neg, threshold, default)

    def is_payment(self,
                  threshold: DecimalCompat=0,
                  default: Optional[bool]=None,
    ) -> Optional[bool]:
        return self.account.is_cash_equivalent() and self.is_debit(threshold, default)


def balance_of(txn: Transaction,
               *preds: Callable[[Account], Optional[bool]],
               default: Optional[DecimalCompat]=None,
) -> Optional[decimal.Decimal]:
    """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.

    If any of the postings have no amount, returns default.
    """
    retval = decimal.Decimal(0)
    for post in txn.postings:
        acct = Account(post.account)
        if any(p(acct) for p in preds):
            if post.units.number is None:
                return None if default is None else decimal.Decimal(default)
            else:
                retval += post.units.number
    return retval

def iter_postings(txn: Transaction) -> Iterator[Posting]:
    """Yield an enhanced Posting object for every posting in the transaction"""
    for index, source in enumerate(txn.postings):
        yield Posting(
            Account(source.account),
            *source[1:5],
            # see rationale above about Posting.meta
            PostingMeta(txn, index, source), # type:ignore[arg-type]
        )