Source code for lena.context.context

""":class:`Context` provides a better representation for context."""
import functools
import json

import lena.core
import lena.flow


[docs]class Context(dict): """Dictionary with easy-to-read formatting. :class:`Context` provides a better representation for context. Example: >>> from lena.context import Context >>> c = Context({"1": 1, "2": {"3": 4}}) >>> print(c) # doctest: +NORMALIZE_WHITESPACE { "1": 1, "2": { "3": 4 } } """ def __init__(self, d=None, formatter=None): """Initialize from a dictionary *d* (empty by default). Representation is defined by the *formatter*. That must be a callable accepting a dictionary and returning a string. The default is ``json.dumps``. All public attributes of a :class:`Context` can be got or set using dot notation (for example, *context["data_path"]* is equal to *context.data_path*). Only one level of nesting is accessible using dot notation. Tip --- `JSON <https://docs.python.org/3/library/json.html>`_ and Python representations are different. In particular, JSON *True* is written as lowercase *true*. To convert JSON back to Python, use ``json.loads(string)``. If *formatter* is given but is not callable, :exc:`.LenaTypeError` is raised. If the attribute to be got is missing, :exc:`.LenaAttributeError` is raised. An attempt to get a private attribute raises :exc:`AttributeError`. """ if d is None: d = {} super(Context, self).__init__(d) if formatter is not None: if not callable(formatter): raise lena.core.LenaTypeError( "formatter must be callable, " "{} given".format(formatter) ) self._formatter = formatter else: self._formatter = functools.partial(json.dumps, sort_keys=True, indent=4) # same, but doesn't allow pickling # lambda s: json.dumps(s, sort_keys=True, indent=4) # formatter should be private, # otherwise it'll mess with other attributes # self._formatter = pprint.PrettyPrinter(indent=1)
[docs] def __call__(self, value): """Convert *value*'s context to :class:`Context` on the fly. If the *value* is a *(data, context)* pair, convert its context part to :class:`Context`. If the *value* doesn't contain a context, it is created as an empty :class:`Context`. When a :class:`Context` is used as a sequence element, its initialization argument *d* has no effect on the produced values. """ data, context = lena.flow.get_data_context(value) return (data, Context(context))
def __getattr__(self, name): # we don't implement getting nested attributes, # because that would require creating proxy objects # -- maybe in the future if needed. if name.startswith('_'): # this is not LenaAttributeError, # as it wouldn't be so for other Lena classes # that don't implement __getattr__ . # See comment for Variable. raise AttributeError(name) try: return self[name] except KeyError: raise lena.core.LenaAttributeError( "{} missing".format(name) ) def __repr__(self): return self._formatter(self) def __setattr__(self, attr, value): if attr in ["_formatter"]: # from https://stackoverflow.com/a/17020163/952234 super(Context, self).__setattr__(attr, value) elif attr.startswith('_'): raise AttributeError(attr) else: self[attr] = value