term.py_codec_impl module

Module implements encoder and decoder from ETF (Erlang External Term Format) used by the network distribution layer.

term.py_codec_impl.binary_to_term(data: bytes, options: dict = None) -> (<built-in function any>, <class 'bytes'>)

Strip 131 header and unpack if the data was compressed.

Parameters:

• data -- The incoming encoded data with the 131 byte
• options --
  • "atom": "str" | "bytes" | "Atom" (default "Atom"). Returns atoms as strings, as bytes or as atom.Atom objects.
  • "byte_string": "str" | "bytes" (default "str"). Returns 8-bit strings as Python str or bytes.

Raises:

PyCodecError -- when the tag is not 131, when compressed data is incomplete or corrupted

Returns:

Remaining unconsumed bytes

term.py_codec_impl.binary_to_term_2(data: bytes, options: dict = None) -> (<built-in function any>, <class 'bytes'>)

Proceed decoding after leading tag has been checked and removed.

Erlang lists are decoded into term.List object, whose elements_ field contains the data, tail_ field has the optional tail and a helper function exists to assist with extracting an unicode string.

Atoms are decoded to Atom or optionally to bytes or to strings. Pids are decoded into Pid. Refs decoded to and Reference. Maps are decoded into Python dict. Binaries are decoded into bytes object and bitstrings into a pair of (bytes, last_byte_bits:int).

Parameters:

• options --

  dict(str, _);

  • "atom": "str" | "bytes" | "Atom"; default "Atom". Returns atoms as strings, as bytes or as atom.Atom class objects

  • "byte_string": "str" | "bytes" (default "str"). Returns 8-bit strings as Python str or bytes.
• data -- Bytes containing encoded term without 131 header

Returns:

Tuple[Value, Tail: bytes] The function consumes as much data as possible and returns the tail. Tail can be used again to parse another term if there was any.

Raises:

PyCodecError(str) -- on various errors

term.py_codec_impl.term_to_binary(val, opt: typing.Union[NoneType, dict] = None) → bytes

Prepend the 131 header byte to encoded data. :param opt: None or dict of options: "encode_hook" is a callable which

will return representation for unknown object types. Returning None will be encoded as such and becomes Atom('undefined').

term.py_codec_impl.term_to_binary_2(val, encode_hook: typing.Union[typing.Callable, NoneType]) → bytes

Erlang lists are decoded into term.List object, whose elements_ field contains the data, tail_ field has the optional tail and a helper function exists to assist with extracting an unicode string.

Parameters:

• encode_hook -- None or a callable which will represent an unknown object as an Erlang term before encoding. Returning None will be encoded as such and becomes Atom('undefined').
• val -- Almost any Python value

Returns:

bytes object with encoded data, but without a 131 header byte.

exception term.py_codec_impl.PyCodecError

Bases: Exception

term.py_codec_impl.generic_serialize_object(obj, cd: set = None)

Given an arbitraty Python object creates a tuple (ClassName, {Fields}). A fair effort is made to avoid infinite recursion on cyclic objects. :param obj: Arbitrary object to encode :param cd: A set with ids of object, for cycle detection :return: A pair of result: (ClassName :: bytes(), Fields :: {bytes(), _})

or None, and a CycleDetect value