conservancy_beancount/conservancy_beancount/cliutil.py

"""cliutil - Utilities for CLI tools"""
PKGNAME = 'conservancy_beancount'
LICENSE = """
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 argparse
import datetime
import enum
import io
import logging
import operator
import os
import pkg_resources
import re
import signal
import sys
import traceback
import types

from pathlib import Path

from . import data
from . import filters
from . import rtutil

from typing import (
    cast,
    Any,
    BinaryIO,
    Callable,
    IO,
    Iterable,
    NamedTuple,
    NoReturn,
    Optional,
    Sequence,
    TextIO,
    Type,
    Union,
)
from .beancount_types import (
    MetaKey,
)

OutputFile = Union[int, IO]

STDSTREAM_PATH = Path('-')
VERSION = pkg_resources.require(PKGNAME)[0].version

class ExceptHook:
    def __init__(self,
                 logger: Optional[logging.Logger]=None,
                 default_exitcode: int=3,
    ) -> None:
        if logger is None:
            logger = logging.getLogger()
        self.logger = logger
        self.default_exitcode = default_exitcode

    def __call__(self,
                 exc_type: Type[BaseException],
                 exc_value: BaseException,
                 exc_tb: types.TracebackType,
    ) -> NoReturn:
        exitcode = self.default_exitcode
        if isinstance(exc_value, KeyboardInterrupt):
            signal.signal(signal.SIGINT, signal.SIG_DFL)
            os.kill(0, signal.SIGINT)
            signal.pause()
        elif isinstance(exc_value, OSError):
            exitcode += 1
            msg = "I/O error: {e.filename}: {e.strerror}".format(e=exc_value)
        else:
            parts = [type(exc_value).__name__, *exc_value.args]
            msg = "internal " + ": ".join(parts)
        self.logger.critical(msg)
        self.logger.debug(
            ''.join(traceback.format_exception(exc_type, exc_value, exc_tb)),
        )
        raise SystemExit(exitcode)


class InfoAction(argparse.Action):
    def __call__(self,
                 parser: argparse.ArgumentParser,
                 namespace: argparse.Namespace,
                 values: Union[Sequence[Any], str, None]=None,
                 option_string: Optional[str]=None,
    ) -> NoReturn:
        if isinstance(self.const, str):
            info = self.const
            exitcode = 0
        else:
            info, exitcode = self.const
        print(info)
        raise SystemExit(exitcode)


class LogLevel(enum.IntEnum):
    DEBUG = logging.DEBUG
    INFO = logging.INFO
    WARNING = logging.WARNING
    ERROR = logging.ERROR
    CRITICAL = logging.CRITICAL
    WARN = WARNING
    ERR = ERROR
    CRIT = CRITICAL

    @classmethod
    def from_arg(cls, arg: str) -> int:
        try:
            return cls[arg.upper()].value
        except KeyError:
            raise ValueError(f"unknown loglevel {arg!r}") from None

    @classmethod
    def choices(cls) -> Iterable[str]:
        for level in sorted(cls, key=operator.attrgetter('value')):
            yield level.name.lower()


class SearchTerm(NamedTuple):
    """NamedTuple representing a user's metadata filter

    SearchTerm knows how to parse and store posting metadata filters provided
    by the user in `key=value` format. Reporting tools can use this to filter
    postings that match the user's criteria, to report on subsets of the books.

    Typical usage looks like::

      argument_parser.add_argument(
        'search_terms',
        type=SearchTerm.arg_parser(),
        …,
      )

      args = argument_parser.parse_args(…)
      for query in args.search_terms:
        postings = query.filter_postings(postings)
    """
    meta_key: MetaKey
    pattern: str

    @classmethod
    def arg_parser(cls,
                   default_key: Optional[str]=None,
                   ticket_default_key: Optional[str]=None,
    ) -> Callable[[str], 'SearchTerm']:
        """Build a SearchTerm parser

        This method returns a function that can parse strings in ``key=value``
        format and return a corresponding SearchTerm.

        If you specify a default key, then strings that just specify a ``value``
        will be parsed as if they said ``default_key=value``. Otherwise,
        parsing strings without a metadata key will raise a ValueError.

        If you specify a default key ticket links, then values in the format
        ``number``, ``rt:number``, or ``rt://ticket/number`` will be parsed as
        if they said ``ticket_default_key=value``.
        """
        if ticket_default_key is None:
            ticket_default_key = default_key
        def parse_search_term(arg: str) -> 'SearchTerm':
            key: Optional[str] = None
            if re.match(r'^[a-z][-\w]*=', arg):
                key, _, raw_link = arg.partition('=')
            else:
                raw_link = arg
            rt_ids = rtutil.RT.parse(raw_link)
            if rt_ids is None:
                rt_ids = rtutil.RT.parse('rt:' + raw_link)
            if rt_ids is None:
                if key is None:
                    key = default_key
                pattern = r'(?:^|\s){}(?:\s|$)'.format(re.escape(raw_link))
            else:
                ticket_id, attachment_id = rt_ids
                if key is None:
                    if attachment_id is None:
                        key = ticket_default_key
                    else:
                        key = default_key
                pattern = rtutil.RT.metadata_regexp(
                    ticket_id,
                    attachment_id,
                    first_link_only=key == 'rt-id' and attachment_id is None,
                )
            if key is None:
                raise ValueError(f"invalid search term {arg!r}: no metadata key")
            return cls(key, pattern)
        return parse_search_term

    def filter_postings(self, postings: Iterable[data.Posting]) -> Iterable[data.Posting]:
        return filters.filter_meta_match(
            postings, self.meta_key, re.compile(self.pattern),
        )


def add_loglevel_argument(parser: argparse.ArgumentParser,
                          default: LogLevel=LogLevel.INFO) -> argparse.Action:
    return parser.add_argument(
        '--loglevel',
        metavar='LEVEL',
        default=default.value,
        type=LogLevel.from_arg,
        help="Show logs at this level and above."
        f" Specify one of {', '.join(LogLevel.choices())}."
        f" Default {default.name.lower()}.",
    )

def add_version_argument(parser: argparse.ArgumentParser) -> argparse.Action:
    progname = parser.prog or sys.argv[0]
    return parser.add_argument(
        '--version', '--copyright', '--license',
        action=InfoAction,
        nargs=0,
        const=f"{progname} version {VERSION}\n{LICENSE}",
        help="Show program version and license information",
    )

def date_arg(arg: str) -> datetime.date:
    return datetime.datetime.strptime(arg, '%Y-%m-%d').date()

def year_or_date_arg(arg: str) -> Union[int, datetime.date]:
    """Get either a date or a year (int) from an argument string

    This is a useful argument type for arguments that will be passed into
    Books loader methods which can accept either a fiscal year or a full date.
    """
    try:
        year = int(arg, 10)
    except ValueError:
        ok = False
    else:
        ok = datetime.MINYEAR <= year <= datetime.MAXYEAR
    if ok:
        return year
    else:
        return date_arg(arg)

def make_entry_point(mod_name: str, prog_name: str=sys.argv[0]) -> Callable[[], int]:
    """Create an entry_point function for a tool

    The returned function is suitable for use as an entry_point in setup.py.
    It sets up the root logger and excepthook, then calls the module's main
    function.
    """
    def entry_point():  # type:ignore
        prog_mod = sys.modules[mod_name]
        setup_logger()
        prog_mod.logger = logging.getLogger(prog_name)
        sys.excepthook = ExceptHook(prog_mod.logger)
        return prog_mod.main()
    return entry_point

def setup_logger(logger: Union[str, logging.Logger]='',
                 stream: TextIO=sys.stderr,
                 fmt: str='%(name)s: %(levelname)s: %(message)s',
) -> logging.Logger:
    """Set up a logger with a StreamHandler with the given format"""
    if isinstance(logger, str):
        logger = logging.getLogger(logger)
    formatter = logging.Formatter(fmt)
    handler = logging.StreamHandler(stream)
    handler.setFormatter(formatter)
    logger.addHandler(handler)
    return logger

def set_loglevel(logger: logging.Logger, loglevel: int=logging.INFO) -> None:
    """Set the loglevel for a tool or module

    If the given logger is not under a hierarchy, this function sets the
    loglevel for the root logger, along with some specific levels for libraries
    used by reporting tools. Otherwise, it's the same as
    ``logger.setLevel(loglevel)``.
    """
    if '.' not in logger.name:
        logger = logging.getLogger()
        if loglevel <= logging.DEBUG:
            # At the debug level, the rt module logs the full body of every
            # request and response. That's too much.
            logging.getLogger('rt.rt').setLevel(logging.INFO)
    logger.setLevel(loglevel)

def bytes_output(path: Optional[Path]=None,
                 default: OutputFile=sys.stdout,
                 mode: str='w',
) -> BinaryIO:
    """Get a file-like object suitable for binary output

    If ``path`` is ``None`` or ``-``, returns a file-like object backed by
    ``default``. If ``default`` is a file descriptor or text IO object, this
    method returns a file-like object that writes to the same place.

    Otherwise, returns ``path.open(mode)``.
    """
    mode = f'{mode}b'
    if path is None or path == STDSTREAM_PATH:
        if isinstance(default, int):
            retval = open(default, mode)
        elif isinstance(default, TextIO):
            retval = default.buffer
        else:
            retval = default
    else:
        retval = path.open(mode)
    return cast(BinaryIO, retval)

def text_output(path: Optional[Path]=None,
                default: OutputFile=sys.stdout,
                mode: str='w',
                encoding: Optional[str]=None,
) -> TextIO:
    """Get a file-like object suitable for text output

    If ``path`` is ``None`` or ``-``, returns a file-like object backed by
    ``default``. If ``default`` is a file descriptor or binary IO object, this
    method returns a file-like object that writes to the same place.

    Otherwise, returns ``path.open(mode)``.
    """
    if path is None or path == STDSTREAM_PATH:
        if isinstance(default, int):
            retval = open(default, mode, encoding=encoding)
        elif isinstance(default, BinaryIO):
            retval = io.TextIOWrapper(default, encoding=encoding)
        else:
            retval = default
    else:
        retval = path.open(mode, encoding=encoding)
    return cast(TextIO, retval)