pyrlp

A package for Recursive Length Prefix encoding and decoding

Contents

Quickstart

>>> import rlp
>>> from rlp.sedes import big_endian_int, text, List

>>> rlp.encode(1234)
b'\x82\x04\xd2'
>>> rlp.decode(b'\x82\x04\xd2', big_endian_int)
1234

>>> rlp.encode([1, [2, []]])
b'\xc4\x01\xc2\x02\xc0'
>>> list_sedes = List([big_endian_int, [big_endian_int, []]])
>>> rlp.decode(b'\xc4\x01\xc2\x02\xc0', list_sedes)
(1, (2, ()))

>>> class Tx(rlp.Serializable):
...     fields = [
...         ('from', text),
...         ('to', text),
...         ('amount', big_endian_int)
...     ]
...
>>> tx = Tx('me', 'you', 255)
>>> rlp.encode(tx)
b'\xc9\x82me\x83you\x81\xff'
>>> rlp.decode(b'\xc9\x82me\x83you\x81\xff', Tx) == tx
True

Tutorial

Basics

There are two types of fundamental items one can encode in RLP:

Strings of bytes

Lists of other items

In this package, byte strings are represented either as Python strings or as bytearrays. Lists can be any sequence, e.g. lists or tuples. To encode these kinds of objects, use rlp.encode():

>>> from rlp import encode
>>> encode('ethereum')
b'\x88ethereum'
>>> encode('')
b'\x80'
>>> encode('Lorem ipsum dolor sit amet, consetetur sadipscing elitr.')
b'\xb88Lorem ipsum dolor sit amet, consetetur sadipscing elitr.'
>>> encode([])
b'\xc0'
>>> encode(['this', ['is', ('a', ('nested', 'list', []))]])
b'\xd9\x84this\xd3\x82is\xcfa\xcd\x86nested\x84list\xc0'

Decoding is just as simple:

>>> from rlp import decode
>>> decode(b'\x88ethereum')
b'ethereum'
>>> decode(b'\x80')
b''
>>> decode(b'\xc0')
[]
>>> decode(b'\xd9\x84this\xd3\x82is\xcfa\xcd\x86nested\x84list\xc0')
[b'this', [b'is', [b'a', [b'nested', b'list', []]]]]

Now, what if we want to encode a different object, say, an integer? Let’s try:

>>> encode(1503)
b'\x82\x05\xdf'
>>> decode(b'\x82\x05\xdf')
b'\x05\xdf'

Oops, what happened? Encoding worked fine, but rlp.decode() refused to give an integer back. The reason is that RLP is typeless. It doesn’t know if the encoded data represents a number, a string, or a more complicated object. It only distinguishes between byte strings and lists. Therefore, pyrlp guesses how to serialize the object into a byte string (here, in big endian notation). When encoded however, the type information is lost and rlp.decode() returned the result in its most generic form, as a string. Thus, what we need to do is deserialize the result afterwards.

Sedes objects

Serialization and its couterpart, deserialization, is done by, what we call, sedes objects (borrowing from the word “codec”). For integers, the sedes rlp.sedes.big_endian_int is in charge. To decode our integer, we can pass this sedes to rlp.decode():

>>> from rlp.sedes import big_endian_int
>>> decode(b'\x82\x05\xdf', big_endian_int)
1503

For unicode strings, there’s the sedes rlp.sedes.binary, which uses UTF-8 to convert to and from byte strings:

>>> from rlp.sedes import binary
>>> encode(u'Ðapp')
b'\x85\xc3\x90app'
>>> decode(b'\x85\xc3\x90app', binary)
b'\xc3\x90app'
>>> print(decode(b'\x85\xc3\x90app', binary).decode('utf-8'))
Ðapp

Lists are a bit more difficult as they can contain arbitrarily complex combinations of types. Therefore, we need to create a sedes object specific for each list type. As base class for this we can use rlp.sedes.List:

>>> from rlp.sedes import List
>>> encode([5, 'fdsa', 0])
b'\xc7\x05\x84fdsa\x80'
>>> sedes = List([big_endian_int, binary, big_endian_int])
>>> decode(b'\xc7\x05\x84fdsa\x80', sedes)
(5, b'fdsa', 0)

Unsurprisingly, it is also possible to nest rlp.List objects:

>>> inner = List([binary, binary])
>>> outer = List([inner, inner, inner])
>>> decode(encode(['asdf', 'fdsa']), inner)
(b'asdf', b'fdsa')
>>> decode(encode([['a1', 'a2'], ['b1', 'b2'], ['c1', 'c2']]), outer)
((b'a1', b'a2'), (b'b1', b'b2'), (b'c1', b'c2'))

What Sedes Objects Actually Are

We saw how to use sedes objects, but what exactly are they? They are characterized by providing the following three member functions:

serializable(obj)

serialize(obj)

deserialize(serial)

The latter two are used to convert between a Python object and its representation as byte strings or sequences. The former one may be called by rlp.encode() to infer which sedes object to use for a given object (see Sedes Inference).

For basic types, the sedes object is usually a module (e.g. rlp.sedes.big_endian_int and rlp.sedes.binary). Instances of rlp.sedes.List provide the sedes interface too, as well as the class rlp.Serializable which is discussed in the following section.

Encoding Custom Objects

Often, we want to encode our own objects in RLP. Examples from the Ethereum world are transactions, blocks or anything send over the Wire. With pyrlp, this is as easy as subclassing rlp.Serializable:

>>> import rlp
>>> class Transaction(rlp.Serializable):
...    fields = (
...        ('sender', binary),
...        ('receiver', binary),
...        ('amount', big_endian_int)
...    )

The class attribute fields is a sequence of 2-tuples defining the field names and the corresponding sedes. For each name an instance attribute is created, that can conveniently be initialized with __init__():

>>> tx1 = Transaction(b'me', b'you', 255)
>>> tx2 = Transaction(amount=255, sender=b'you', receiver=b'me')
>>> tx1.amount
255

At serialization, the field names are dropped and the object is converted to a list, where the provided sedes objects are used to serialize the object attributes:

>>> Transaction.serialize(tx1)
[b'me', b'you', b'\xff']
>>> tx1 == Transaction.deserialize([b'me', b'you', b'\xff'])
True

As we can see, each subclass of rlp.Serializable implements the sedes responsible for its instances. Therefore, we can use rlp.encode() and rlp.decode() as expected:

>>> encode(tx1)
b'\xc9\x82me\x83you\x81\xff'
>>> decode(b'\xc9\x82me\x83you\x81\xff', Transaction) == tx1
True

Sedes Inference

As we have seen, rlp.encode() (or, rather, rlp.infer_sedes()) tries to guess a sedes capable of serializing the object before encoding. In this process, it follows the following steps:

Check if the object’s class is a sedes object (like every subclass of rlp.Serializable). If so, its class is the sedes.
Check if one of the entries in rlp.sedes.sedes_list can serialize the object (via serializable(obj)). If so, this is the sedes.
Check if the object is a sequence. If so, build a rlp.sedes.List by recursively infering a sedes for each of its elements.
If none of these steps was successful, sedes inference has failed.

If you have build your own basic sedes (e.g. for dicts or floats), you might want to hook in at step 2 and add it to rlp.sedes.sedes_list, whereby it will be automatically be used by rlp.encode().

API Reference

Functions

rlp.encode(obj, sedes=None, infer_serializer=True, cache=True)

Encode a Python object in RLP format.

By default, the object is serialized in a suitable way first (using rlp.infer_sedes()) and then encoded. Serialization can be explicitly suppressed by setting infer_serializer to False and not passing an alternative as sedes.

If obj has an attribute _cached_rlp (as, notably, rlp.Serializable) and its value is not None, this value is returned bypassing serialization and encoding, unless sedes is given (as the cache is assumed to refer to the standard serialization which can be replaced by specifying sedes).

If obj is a rlp.Serializable and cache is true, the result of the encoding will be stored in _cached_rlp if it is empty.

Parameters:

sedes – an object implementing a function serialize(obj) which will be used to serialize obj before encoding, or None to use the infered one (if any)
infer_serializer – if True an appropriate serializer will be selected using rlp.infer_sedes() to serialize obj before encoding
cache – cache the return value in obj._cached_rlp if possible (default True)

Returns:

the RLP encoded item

Raises:

rlp.EncodingError in the rather unlikely case that the item is too big to encode (will not happen)

Raises:

rlp.SerializationError if the serialization fails

rlp.decode(rlp, sedes=None, strict=True, recursive_cache=False, **kwargs)

Decode an RLP encoded object.

If the deserialized result obj has an attribute _cached_rlp (e.g. if sedes is a subclass of rlp.Serializable) it will be set to rlp, which will improve performance on subsequent rlp.encode() calls. Bear in mind however that obj needs to make sure that this value is updated whenever one of its fields changes or prevent such changes entirely (rlp.sedes.Serializable does the latter).

Parameters:

sedes – an object implementing a function deserialize(code) which will be applied after decoding, or None if no deserialization should be performed
**kwargs – additional keyword arguments that will be passed to the deserializer
strict – if false inputs that are longer than necessary don’t cause an exception

Returns:

the decoded and maybe deserialized Python object

Raises:

rlp.DecodingError if the input string does not end after the root item and strict is true

Raises:

rlp.DeserializationError if the deserialization fails

rlp.decode_lazy(rlp, sedes=None, **sedes_kwargs)

Decode an RLP encoded object in a lazy fashion.

If the encoded object is a bytestring, this function acts similar to rlp.decode(). If it is a list however, a LazyList is returned instead. This object will decode the string lazily, avoiding both horizontal and vertical traversing as much as possible.

The way sedes is applied depends on the decoded object: If it is a string sedes deserializes it as a whole; if it is a list, each element is deserialized individually. In both cases, sedes_kwargs are passed on. Note that, if a deserializer is used, only “horizontal” but not “vertical lazyness” can be preserved.

Parameters:

rlp – the RLP string to decode
sedes – an object implementing a method deserialize(code) which is used as described above, or None if no deserialization should be performed
**sedes_kwargs – additional keyword arguments that will be passed to the deserializers

Returns:

either the already decoded and deserialized object (if encoded as a string) or an instance of rlp.LazyList

class rlp.LazyList(rlp, start, end, sedes=None, **sedes_kwargs)

A RLP encoded list which decodes itself when necessary.

Both indexing with positive indices and iterating are supported. Getting the length with len() is possible as well but requires full horizontal encoding.

Parameters:

rlp – the rlp string in which the list is encoded
start – the position of the first payload byte of the encoded list
end – the position of the last payload byte of the encoded list
sedes – a sedes object which deserializes each element of the list, or None for no deserialization
**sedes_kwargs – keyword arguments which will be passed on to the deserializer

rlp.infer_sedes(obj)

Try to find a sedes objects suitable for a given Python object.

The sedes objects considered are obj’s class, big_endian_int and binary. If obj is a sequence, a rlp.sedes.List will be constructed recursively.

Parameters:: obj – the python object for which to find a sedes object
Raises:: TypeError if no appropriate sedes could be found

Sedes Objects

rlp.sedes.raw: A sedes object that does nothing. Thus, it can serialize everything that can be directly encoded in RLP (nested lists of strings). This sedes can be used as a placeholder when deserializing larger structures.

class rlp.sedes.Binary(min_length=None, max_length=None, allow_empty=False)

A sedes object for binary data of certain length.

Parameters:

min_length – the minimal length in bytes or None for no lower limit
max_length – the maximal length in bytes or None for no upper limit
allow_empty – if true, empty strings are considered valid even if a minimum length is required otherwise

classmethod fixed_length(length, allow_empty=False): Create a sedes for binary data with exactly length bytes.

rlp.sedes.binary: A sedes object for binary data of arbitrary length (an instance of rlp.sedes.Binary with default arguments).

class rlp.sedes.Boolean: A sedes for booleans

rlp.sedes.boolean: A sedes object for boolean types.

class rlp.sedes.Text(min_length=None, max_length=None, allow_empty=False, encoding='utf8')

A sedes object for encoded text data of certain length.

Parameters:

min_length – the minimal length in encoded characters or None for no lower limit
max_length – the maximal length in encoded characters or None for no upper limit
allow_empty – if true, empty strings are considered valid even if a minimum length is required otherwise

classmethod fixed_length(length, allow_empty=False): Create a sedes for text data with exactly length encoded characters.

rlp.sedes.text: A sedes object for utf encoded text data of arbitrary length (an instance of rlp.sedes.Text with default arguments).

class rlp.sedes.BigEndianInt(length=None)

A sedes for big endian integers.

Parameters:: l – the size of the serialized representation in bytes or None to use the shortest possible one

rlp.sedes.big_endian_int: A sedes object for integers encoded in big endian without any leading zeros (an instance of rlp.sedes.BigEndianInt with default arguments).

class rlp.sedes.List(elements=None, strict=True)

A sedes for lists, implemented as a list of other sedes objects.

Parameters:: strict – If true (de)serializing lists that have a length not matching the sedes length will result in an error. If false (de)serialization will stop as soon as either one of the lists runs out of elements.

class rlp.sedes.CountableList(element_sedes, max_length=None)

A sedes for lists of arbitrary length.

Parameters:

element_sedes – when (de-)serializing a list, this sedes will be applied to all of its elements
max_length – maximum number of allowed elements, or None for no limit

class rlp.Serializable(*args, **kwargs): The base class for serializable objects.

Exceptions

exception rlp.RLPException: Base class for exceptions raised by this package.

exception rlp.EncodingError(message, obj)

Exception raised if encoding fails.

Variables:: obj – the object that could not be encoded

exception rlp.DecodingError(message, rlp)

Exception raised if decoding fails.

Variables:: rlp – the RLP string that could not be decoded

exception rlp.SerializationError(message, obj)

Exception raised if serialization fails.

Variables:: obj – the object that could not be serialized

exception rlp.DeserializationError(message, serial)

Exception raised if deserialization fails.

Variables:: serial – the decoded RLP string that could not be deserialized

rlp package

Subpackages

rlp.sedes package

Submodules

rlp.sedes.big_endian_int module

class rlp.sedes.big_endian_int.BigEndianInt(length=None)

Bases: object

A sedes for big endian integers.

Parameters:: l – the size of the serialized representation in bytes or None to use the shortest possible one

deserialize(serial)

serialize(obj)

rlp.sedes.binary module

class rlp.sedes.binary.Binary(min_length=None, max_length=None, allow_empty=False)

Bases: object

A sedes object for binary data of certain length.

Parameters:

min_length – the minimal length in bytes or None for no lower limit
max_length – the maximal length in bytes or None for no upper limit
allow_empty – if true, empty strings are considered valid even if a minimum length is required otherwise

deserialize(serial)

classmethod fixed_length(length, allow_empty=False): Create a sedes for binary data with exactly length bytes.

is_valid_length(length)

classmethod is_valid_type(obj)

serialize(obj)

rlp.sedes.boolean module

class rlp.sedes.boolean.Boolean

Bases: object

A sedes for booleans

deserialize(serial)

serialize(obj)

rlp.sedes.lists module

Module for sedes objects that use lists as serialization format.

class rlp.sedes.lists.CountableList(element_sedes, max_length=None)

Bases: object

A sedes for lists of arbitrary length.

Parameters:

element_sedes – when (de-)serializing a list, this sedes will be applied to all of its elements
max_length – maximum number of allowed elements, or None for no limit

deserialize(serial)

serialize(obj)

class rlp.sedes.lists.List(elements=None, strict=True)

Bases: list

A sedes for lists, implemented as a list of other sedes objects.

Parameters:: strict – If true (de)serializing lists that have a length not matching the sedes length will result in an error. If false (de)serialization will stop as soon as either one of the lists runs out of elements.

deserialize(serial)

serialize(obj)

rlp.sedes.lists.is_sedes(obj)

Check if obj is a sedes object.

A sedes object is characterized by having the methods serialize(obj) and deserialize(serial).

rlp.sedes.lists.is_sequence(obj): Check if obj is a sequence, but not a string or bytes.

rlp.sedes.raw module

A sedes that does nothing. Thus, everything that can be directly encoded by RLP is serializable. This sedes can be used as a placeholder when deserializing larger structures.

rlp.sedes.raw.deserialize(serial)

rlp.sedes.raw.serializable(obj)

rlp.sedes.raw.serialize(obj)

rlp.sedes.serializable module

class rlp.sedes.serializable.BaseChangeset(obj, changes=None)

Bases: object

build_rlp()

close()

commit()

open()

class rlp.sedes.serializable.BaseSerializable(*args, **kwargs)

Bases: Sequence

as_dict()

build_changeset(*args, **kwargs)

copy(*args, **kwargs)

classmethod deserialize(serial, **extra_kwargs)

classmethod serialize(obj)

rlp.sedes.serializable.Changeset(obj, changes)

class rlp.sedes.serializable.ChangesetField(field)

Bases: object

field = None

class rlp.sedes.serializable.ChangesetState(value)

Bases: Enum

An enumeration.

CLOSED = 'CLOSED'

INITIALIZED = 'INITIALIZED'

OPEN = 'OPEN'

class rlp.sedes.serializable.MetaBase

Bases: object

field_attrs = None

field_names = None

fields = None

sedes = None

class rlp.sedes.serializable.Serializable(*args, **kwargs)

Bases: BaseSerializable

The base class for serializable objects.

class rlp.sedes.serializable.SerializableBase(name, bases, attrs): Bases: ABCMeta

rlp.sedes.serializable.make_immutable(value)

rlp.sedes.serializable.merge_args_to_kwargs(args, kwargs, arg_names, allow_missing=False)

rlp.sedes.serializable.merge_kwargs_to_args(args, kwargs, arg_names, allow_missing=False)

rlp.sedes.serializable.validate_args_and_kwargs(args, kwargs, arg_names, allow_missing=False)

rlp.sedes.text module

class rlp.sedes.text.Text(min_length=None, max_length=None, allow_empty=False, encoding='utf8')

Bases: object

A sedes object for encoded text data of certain length.

Parameters:

min_length – the minimal length in encoded characters or None for no lower limit
max_length – the maximal length in encoded characters or None for no upper limit
allow_empty – if true, empty strings are considered valid even if a minimum length is required otherwise

deserialize(serial)

classmethod fixed_length(length, allow_empty=False): Create a sedes for text data with exactly length encoded characters.

is_valid_length(length)

classmethod is_valid_type(obj)

serialize(obj)

Module contents

Submodules

rlp.atomic module

class rlp.atomic.Atomic

Bases: object

ABC for objects that can be RLP encoded as is.

rlp.codec module

rlp.codec.consume_item(rlp, start)

Read an item from an RLP string.

Parameters:

rlp – the rlp string to read from
start – the position at which to start reading

Returns:

a tuple (item, per_item_rlp, end), where item is the read item, per_item_rlp is a list containing the RLP encoding of each item and end is the position of the first unprocessed byte

rlp.codec.consume_length_prefix(rlp, start)

Read a length prefix from an RLP string.

Parameters:

rlp – the rlp byte string to read from
start – the position at which to start reading

Returns:

a tuple (prefix, type, length, end), where type is either str or list depending on the type of the following payload, length is the length of the payload in bytes, and end is the position of the first payload byte in the rlp string

rlp.codec.consume_payload(rlp, prefix, start, type_, length)

Read the payload of an item from an RLP string.

Parameters:

rlp – the rlp string to read from
type – the type of the payload (bytes or list)
start – the position at which to start reading
length – the length of the payload in bytes

Returns:

a tuple (item, per_item_rlp, end), where item is the read item, per_item_rlp is a list containing the RLP encoding of each item and end is the position of the first unprocessed byte

rlp.codec.decode(rlp, sedes=None, strict=True, recursive_cache=False, **kwargs)

Decode an RLP encoded object.

If the deserialized result obj has an attribute _cached_rlp (e.g. if sedes is a subclass of rlp.Serializable) it will be set to rlp, which will improve performance on subsequent rlp.encode() calls. Bear in mind however that obj needs to make sure that this value is updated whenever one of its fields changes or prevent such changes entirely (rlp.sedes.Serializable does the latter).

Parameters:

sedes – an object implementing a function deserialize(code) which will be applied after decoding, or None if no deserialization should be performed
**kwargs – additional keyword arguments that will be passed to the deserializer
strict – if false inputs that are longer than necessary don’t cause an exception

Returns:

the decoded and maybe deserialized Python object

Raises:

rlp.DecodingError if the input string does not end after the root item and strict is true

Raises:

rlp.DeserializationError if the deserialization fails

rlp.codec.decode_raw(item, strict, _)

rlp.codec.encode(obj, sedes=None, infer_serializer=True, cache=True)

Encode a Python object in RLP format.

By default, the object is serialized in a suitable way first (using rlp.infer_sedes()) and then encoded. Serialization can be explicitly suppressed by setting infer_serializer to False and not passing an alternative as sedes.

If obj has an attribute _cached_rlp (as, notably, rlp.Serializable) and its value is not None, this value is returned bypassing serialization and encoding, unless sedes is given (as the cache is assumed to refer to the standard serialization which can be replaced by specifying sedes).

If obj is a rlp.Serializable and cache is true, the result of the encoding will be stored in _cached_rlp if it is empty.

Parameters:

sedes – an object implementing a function serialize(obj) which will be used to serialize obj before encoding, or None to use the infered one (if any)
infer_serializer – if True an appropriate serializer will be selected using rlp.infer_sedes() to serialize obj before encoding
cache – cache the return value in obj._cached_rlp if possible (default True)

Returns:

the RLP encoded item

Raises:

rlp.EncodingError in the rather unlikely case that the item is too big to encode (will not happen)

Raises:

rlp.SerializationError if the serialization fails

rlp.codec.encode_raw(item): RLP encode (a nested sequence of) Atomics.

rlp.codec.infer_sedes(obj)

Try to find a sedes objects suitable for a given Python object.

The sedes objects considered are obj’s class, big_endian_int and binary. If obj is a sequence, a rlp.sedes.List will be constructed recursively.

Parameters:: obj – the python object for which to find a sedes object
Raises:: TypeError if no appropriate sedes could be found

rlp.codec.length_prefix(length, offset)

Construct the prefix to lists or strings denoting their length.

Parameters:

length – the length of the item in bytes
offset – 0x80 when encoding raw bytes, 0xc0 when encoding a list

rlp.exceptions module

exception rlp.exceptions.DecodingError(message, rlp)

Bases: RLPException

Exception raised if decoding fails.

Variables:: rlp – the RLP string that could not be decoded

exception rlp.exceptions.DeserializationError(message, serial)

Bases: RLPException

Exception raised if deserialization fails.

Variables:: serial – the decoded RLP string that could not be deserialized

exception rlp.exceptions.EncodingError(message, obj)

Bases: RLPException

Exception raised if encoding fails.

Variables:: obj – the object that could not be encoded

exception rlp.exceptions.ListDeserializationError(message=None, serial=None, element_exception=None, index=None)

Bases: DeserializationError

Exception raised if deserialization by a sedes.List fails.

Variables:

element_exception – the exception that occurred during the deserialization of one of the elements, or None if the error is unrelated to a specific element
index – the index in the list that produced the error or None if the error is unrelated to a specific element

exception rlp.exceptions.ListSerializationError(message=None, obj=None, element_exception=None, index=None)

Bases: SerializationError

Exception raised if serialization by a sedes.List fails.

Variables:

element_exception – the exception that occurred during the serialization of one of the elements, or None if the error is unrelated to a specific element
index – the index in the list that produced the error or None if the error is unrelated to a specific element

exception rlp.exceptions.ObjectDeserializationError(message=None, serial=None, sedes=None, list_exception=None)

Bases: DeserializationError

Exception raised if deserialization by a sedes.Serializable fails.

Variables:

sedes – the sedes.Serializable that failed
list_exception – exception raised by the underlying list sedes, or None if no such exception has been raised
field – name of the field of the object that produced the error, or None if no field responsible for the error

exception rlp.exceptions.ObjectSerializationError(message=None, obj=None, sedes=None, list_exception=None)

Bases: SerializationError

Exception raised if serialization of a sedes.Serializable object fails.

Variables:

sedes – the sedes.Serializable that failed
list_exception – exception raised by the underlying list sedes, or None if no such exception has been raised
field – name of the field of the object that produced the error, or None if no field responsible for the error

exception rlp.exceptions.RLPException

Bases: Exception

Base class for exceptions raised by this package.

exception rlp.exceptions.SerializationError(message, obj)

Bases: RLPException

Exception raised if serialization fails.

Variables:: obj – the object that could not be serialized

rlp.lazy module

class rlp.lazy.LazyList(rlp, start, end, sedes=None, **sedes_kwargs)

Bases: Sequence

A RLP encoded list which decodes itself when necessary.

Both indexing with positive indices and iterating are supported. Getting the length with len() is possible as well but requires full horizontal encoding.

Parameters:

rlp – the rlp string in which the list is encoded
start – the position of the first payload byte of the encoded list
end – the position of the last payload byte of the encoded list
sedes – a sedes object which deserializes each element of the list, or None for no deserialization
**sedes_kwargs – keyword arguments which will be passed on to the deserializer

next()

rlp.lazy.consume_item_lazy(rlp, start)

Read an item from an RLP string lazily.

If the length prefix announces a string, the string is read; if it announces a list, a LazyList is created.

Parameters:

rlp – the rlp string to read from
start – the position at which to start reading

Returns:

a tuple (item, end) where item is the read string or a LazyList and end is the position of the first unprocessed byte.

rlp.lazy.decode_lazy(rlp, sedes=None, **sedes_kwargs)

Decode an RLP encoded object in a lazy fashion.

If the encoded object is a bytestring, this function acts similar to rlp.decode(). If it is a list however, a LazyList is returned instead. This object will decode the string lazily, avoiding both horizontal and vertical traversing as much as possible.

The way sedes is applied depends on the decoded object: If it is a string sedes deserializes it as a whole; if it is a list, each element is deserialized individually. In both cases, sedes_kwargs are passed on. Note that, if a deserializer is used, only “horizontal” but not “vertical lazyness” can be preserved.

Parameters:

rlp – the RLP string to decode
sedes – an object implementing a method deserialize(code) which is used as described above, or None if no deserialization should be performed
**sedes_kwargs – additional keyword arguments that will be passed to the deserializers

Returns:

either the already decoded and deserialized object (if encoded as a string) or an instance of rlp.LazyList

rlp.lazy.peek(rlp, index, sedes=None)

Get a specific element from an rlp encoded nested list.

This function uses rlp.decode_lazy() and, thus, decodes only the necessary parts of the string.

Usage example:

>>> import rlp
>>> rlpdata = rlp.encode([1, 2, [3, [4, 5]]])
>>> rlp.peek(rlpdata, 0, rlp.sedes.big_endian_int)
1
>>> rlp.peek(rlpdata, [2, 0], rlp.sedes.big_endian_int)
3

Parameters:

rlp – the rlp string
index – the index of the element to peek at (can be a list for nested data)
sedes – a sedes used to deserialize the peeked at object, or None if no deserialization should be performed

Raises:

IndexError if index is invalid (out of range or too many levels)

rlp.utils module

Module contents

pyrlp v4.0.1 (2024-04-24)

Internal Changes - for pyrlp Contributors

Add python 3.12 support, rust-backend now works with python 3.11 and 3.12 (#150)

Miscellaneous Changes

#151

pyrlp v4.0.0 (2023-11-29)

Features

repr() now returns an evaluatable string, like MyRLPObj(my_int_field=1, my_str_field="a_str") (#117)

Internal Changes - for pyrlp Contributors

Convert .format strings to f-strings (#144)
Add autoflake linting and move config to pyproject.toml (#145)

Release Notes

0.4.8

Implement Serializable.make_mutable and rlp.sedes.make_mutable API.
Add mutable flag to Serializable.deserialize to allow deserialization into mutable objects.

pyrlp

Contents

Quickstart

Tutorial

Basics

Sedes objects

What Sedes Objects Actually Are

Encoding Custom Objects

Sedes Inference

Further Reading

API Reference

Functions

Sedes Objects

Exceptions

rlp package

Subpackages

rlp.sedes package

Submodules

rlp.sedes.big_endian_int module

rlp.sedes.binary module

rlp.sedes.boolean module

rlp.sedes.lists module

rlp.sedes.raw module

rlp.sedes.serializable module

rlp.sedes.text module

Module contents

Submodules

rlp.atomic module

rlp.codec module

rlp.exceptions module

rlp.lazy module

rlp.utils module

Module contents

pyrlp v4.0.1 (2024-04-24)

Internal Changes - for pyrlp Contributors

Miscellaneous Changes

pyrlp v4.0.0 (2023-11-29)

Features

Internal Changes - for pyrlp Contributors

Release Notes

0.4.8

Indices and tables