conservancy_beancount/conservancy_beancount/plugin/core.py

"""Base classes for plugin checks"""
# 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 abc
import datetime
import re

from .. import config as configmod
from .. import data
from .. import errors as errormod

from typing import (
    Any,
    Dict,
    FrozenSet,
    Generic,
    Iterable,
    Iterator,
    Mapping,
    Optional,
    Sequence,
    Type,
    TypeVar,
)
from ..beancount_types import (
    Account,
    Directive,
    MetaKey,
    MetaValue,
    MetaValueEnum,
    Transaction,
)

### CONSTANTS

# I expect these will become configurable in the future, which is why I'm
# keeping them outside of a class, but for now constants will do.
DEFAULT_START_DATE: datetime.date = datetime.date(2019, 3, 1)
# The default stop date leaves a little room after so it's easy to test
# dates past the far end of the range.
DEFAULT_STOP_DATE: datetime.date = datetime.date(datetime.MAXYEAR, 1, 1)

### TYPE DEFINITIONS

HookName = str

Entry = TypeVar('Entry', bound=Directive)
class Hook(Generic[Entry], metaclass=abc.ABCMeta):
    DIRECTIVE: Type[Directive]
    HOOK_GROUPS: FrozenSet[HookName] = frozenset()

    def __init__(self, config: configmod.Config) -> None:
        pass
        # Subclasses that need configuration should override __init__ to check
        # and store it.

    @abc.abstractmethod
    def run(self, entry: Entry) -> errormod.Iter: ...


### HELPER CLASSES

class LessComparable(metaclass=abc.ABCMeta):
    @abc.abstractmethod
    def __le__(self, other: Any) -> bool: ...

    @abc.abstractmethod
    def __lt__(self, other: Any) -> bool: ...


CT = TypeVar('CT', bound=LessComparable)
class _GenericRange(Generic[CT]):
    """Convenience class to check whether a value is within a range.

    `foo in generic_range` is equivalent to `start <= foo < stop`.
    Since we have multiple user-configurable ranges, having the check
    encapsulated in an object helps implement the check consistently, and
    makes it easier for subclasses to override.
    """

    def __init__(self, start: CT, stop: CT) -> None:
        self.start = start
        self.stop = stop

    def __repr__(self) -> str:
        return "{clsname}({self.start!r}, {self.stop!r})".format(
            clsname=type(self).__name__,
            self=self,
        )

    def __contains__(self, item: CT) -> bool:
        return self.start <= item < self.stop


class MetadataEnum:
    """Map acceptable metadata values to their normalized forms.

    When a piece of metadata uses a set of allowed values, use this class to
    define them. You can also specify aliases that hooks will normalize to
    the primary values.
    """

    def __init__(self,
                 key: MetaKey,
                 standard_values: Iterable[MetaValueEnum],
                 aliases_map: Optional[Mapping[MetaValueEnum, MetaValueEnum]]=None,
    ) -> None:
        """Specify allowed values and aliases for this metadata.

        Arguments:

        * key: The name of the metadata key that uses this enum.
        * standard_values: A sequence of strings that enumerate the standard
          values for this metadata.
        * aliases_map: A mapping of strings to strings. The keys are
          additional allowed metadata values. The values are standard values
          that each key will evaluate to. The code asserts that all values are
          in standard_values.
        """
        self.key = key
        self._stdvalues = frozenset(standard_values)
        self._aliases: Dict[MetaValueEnum, MetaValueEnum] = dict(aliases_map or ())
        assert self._stdvalues.issuperset(self._aliases.values())
        self._aliases.update((v, v) for v in standard_values)

    def __repr__(self) -> str:
        return "{}<{}>".format(type(self).__name__, self.key)

    def __contains__(self, key: MetaValueEnum) -> bool:
        """Returns true if `key` is a standard value or alias."""
        return key in self._aliases

    def __getitem__(self, key: MetaValueEnum) -> MetaValueEnum:
        """Return the standard value for `key`.

        Raises KeyError if `key` is not a known value or alias.
        """
        return self._aliases[key]

    def __iter__(self) -> Iterator[MetaValueEnum]:
        """Iterate over standard values."""
        return iter(self._stdvalues)

    def get(self,
            key: MetaValueEnum,
            default_key: Optional[MetaValueEnum]=None,
    ) -> Optional[MetaValueEnum]:
        """Return self[key], or a default fallback if that doesn't exist.

        default_key is another key to look up, *not* a default value to return.
        This helps ensure you always get a standard value.
        """
        try:
            return self[key]
        except KeyError:
            if default_key is None:
                return None
            else:
                return self[default_key]


### HOOK SUBCLASSES

class TransactionHook(Hook[Transaction]):
    DIRECTIVE = Transaction
    TXN_DATE_RANGE: _GenericRange = _GenericRange(DEFAULT_START_DATE, DEFAULT_STOP_DATE)

    def _run_on_txn(self, txn: Transaction) -> bool:
        """Check whether we should run on a given transaction

        This method implements our usual checks for whether or not a hook
        should run on a given transaction. It's here for subclasses to use in
        their own implementations. See _PostingHook below for an example.
        """
        return (
            txn.flag != '!'
            and txn.date in self.TXN_DATE_RANGE
            and not data.is_opening_balance_txn(txn)
        )


class _PostingHook(TransactionHook, metaclass=abc.ABCMeta):
    def __init_subclass__(cls) -> None:
        cls.HOOK_GROUPS = cls.HOOK_GROUPS.union(['posting'])

    def _run_on_post(self, txn: Transaction, post: data.Posting) -> bool:
        return True

    def run(self, txn: Transaction) -> errormod.Iter:
        if self._run_on_txn(txn):
            for post in data.Posting.from_txn(txn):
                if self._run_on_post(txn, post):
                    yield from self.post_run(txn, post)

    @abc.abstractmethod
    def post_run(self, txn: Transaction, post: data.Posting) -> errormod.Iter: ...


class _NormalizePostingMetadataHook(_PostingHook):
    """Base class to normalize posting metadata from an enum."""
    # This class provides basic functionality to filter postings, normalize
    # metadata values, and set default values.
    METADATA_KEY: MetaKey
    VALUES_ENUM: MetadataEnum

    def __init_subclass__(cls) -> None:
        super().__init_subclass__()
        cls.METADATA_KEY = cls.VALUES_ENUM.key
        cls.HOOK_GROUPS = cls.HOOK_GROUPS.union(['metadata', cls.METADATA_KEY])

    # If the posting does not specify METADATA_KEY, the hook calls
    # _default_value to get a default. This method should either return
    # a value string from METADATA_ENUM, or else raise InvalidMetadataError.
    # This base implementation does the latter.
    def _default_value(self, txn: Transaction, post: data.Posting) -> MetaValueEnum:
        raise errormod.InvalidMetadataError(txn, self.METADATA_KEY, None, post)

    def post_run(self, txn: Transaction, post: data.Posting) -> errormod.Iter:
        source_value = post.meta.get(self.METADATA_KEY)
        set_value = source_value
        error: Optional[errormod.Error] = None
        if source_value is None:
            try:
                set_value = self._default_value(txn, post)
            except errormod.Error as error_:
                error = error_
        else:
            try:
                set_value = self.VALUES_ENUM[source_value]
            except KeyError:
                error = errormod.InvalidMetadataError(
                    txn, self.METADATA_KEY, source_value, post,
                )
        if error is None:
            post.meta[self.METADATA_KEY] = set_value
        else:
            yield error


class _RequireLinksPostingMetadataHook(_PostingHook):
    """Base class to require that posting metadata include links"""
    # This base class confirms that a posting's metadata has one or more links
    # under one of the metadata keys listed in CHECKED_METADATA.
    # Most subclasses only need to define CHECKED_METADATA and _run_on_post.
    CHECKED_METADATA: Sequence[MetaKey]

    def __init_subclass__(cls) -> None:
        super().__init_subclass__()
        cls.HOOK_GROUPS = cls.HOOK_GROUPS.union(cls.CHECKED_METADATA).union('metadata')

    def _check_metadata(self,
                        txn: Transaction,
                        post: data.Posting,
                        keys: Sequence[MetaKey],
    ) -> Iterator[errormod.InvalidMetadataError]:
        have_docs = False
        for key in keys:
            try:
                links = post.meta.get_links(key)
            except TypeError as error:
                yield errormod.InvalidMetadataError(txn, key, post.meta[key], post)
            else:
                have_docs = have_docs or any(links)
        if not have_docs:
            yield errormod.InvalidMetadataError(txn, '/'.join(keys), None, post)

    def post_run(self, txn: Transaction, post: data.Posting) -> errormod.Iter:
        return self._check_metadata(txn, post, self.CHECKED_METADATA)