compact, efficient, extensible binary serialization format
Project description
extprot: compact, efficient, extensible binary serialization format
This is a python implementation of the 'extprot' serialization scheme, the
details of which are described at:
http://eigenclass.org/R2/writings/extprot-extensible-protocols-intro
Similar to Google's Protocol Buffers and Apache Thrift, extprot allows the
definition of structured data "messages". Messages are essentially a set
of typed key-value pairs that can be efficiently serialized to/from a
compact binary format, and are defined in a language-neutral "protocol" file.
Here's a simple example of an extprot message:
message person = {
id: int;
name: string;
emails: [ string ]
}
Here the 'person' message contains three fields: 'id' is an integer, 'name'
is a string, and 'emails' is a list of strings. Such protocol descriptions
are compiled into a set of Python classes that can be manipulated using
standard syntax and idioms. If the above protocol is recorded in the file
"person.proto", here's a simple example of how it might be used:
>>> extprot.import_protocol("person.proto",globals())
>>> p1 = person(1,"Guido")
>>> print p1.emails # fields use a sensible default if possible
[]
>>> p1.emails.append("guido@python.org")
>>> p1.emails.append(7) # all fields are dynamically typechecked
Traceback (mosts recent call last):
...
ValueError: not a valid String: 7
>>> print repr(p1.to_string())
'\x01\x1f\x03\x00\x02\x03\x05Guido\x05\x13\x01\x03\x10guido@python.org'
>>> print person.from_string(p1.to_string()).name
'Guido'
>>>
Extprot compares favourably to related serialization technologies:
* powerful type system; strongly-typed tuples and lists, tagged disjoint
unions, parametric polymorphism.
* self-delimiting data; all serialized messages indicate their length,
allowing easy streaming and skipping of messages.
* self-describing data; the 'skeleton' of a message can be reconstructed
without having the protocol definition.
* compact binary format; comparable to protocol-buffers/thrift, but with
some overhead due to self-delimiting nature.
These features combine to make extprot strongly extensible, often allowing
messages to maintain backward *and* forward compatibility across protocol
extensions that include: adding fields to a message, adding elements to a
tuple, adding cases to a disjoint union, and promoting a primitive type into
a tuple, list or union.
The function extprot.import_protocol() will dynamically load a protocol file
and convert it into the corresponding python class structure. This is quite
convenient while developing a protocol since it avoids an extra compilation
step, but it does add some startup overhead and requires the pyparsing module.
To compile a protocol definition into python sourcecode for the corresponding
class definitions, use the function extprot.compile_protocol() or pipe the file
through extprot/compiler.py like so:
$ cat mydefs.proto | python extprot/compiler.py > mydefs.py
This is a python implementation of the 'extprot' serialization scheme, the
details of which are described at:
http://eigenclass.org/R2/writings/extprot-extensible-protocols-intro
Similar to Google's Protocol Buffers and Apache Thrift, extprot allows the
definition of structured data "messages". Messages are essentially a set
of typed key-value pairs that can be efficiently serialized to/from a
compact binary format, and are defined in a language-neutral "protocol" file.
Here's a simple example of an extprot message:
message person = {
id: int;
name: string;
emails: [ string ]
}
Here the 'person' message contains three fields: 'id' is an integer, 'name'
is a string, and 'emails' is a list of strings. Such protocol descriptions
are compiled into a set of Python classes that can be manipulated using
standard syntax and idioms. If the above protocol is recorded in the file
"person.proto", here's a simple example of how it might be used:
>>> extprot.import_protocol("person.proto",globals())
>>> p1 = person(1,"Guido")
>>> print p1.emails # fields use a sensible default if possible
[]
>>> p1.emails.append("guido@python.org")
>>> p1.emails.append(7) # all fields are dynamically typechecked
Traceback (mosts recent call last):
...
ValueError: not a valid String: 7
>>> print repr(p1.to_string())
'\x01\x1f\x03\x00\x02\x03\x05Guido\x05\x13\x01\x03\x10guido@python.org'
>>> print person.from_string(p1.to_string()).name
'Guido'
>>>
Extprot compares favourably to related serialization technologies:
* powerful type system; strongly-typed tuples and lists, tagged disjoint
unions, parametric polymorphism.
* self-delimiting data; all serialized messages indicate their length,
allowing easy streaming and skipping of messages.
* self-describing data; the 'skeleton' of a message can be reconstructed
without having the protocol definition.
* compact binary format; comparable to protocol-buffers/thrift, but with
some overhead due to self-delimiting nature.
These features combine to make extprot strongly extensible, often allowing
messages to maintain backward *and* forward compatibility across protocol
extensions that include: adding fields to a message, adding elements to a
tuple, adding cases to a disjoint union, and promoting a primitive type into
a tuple, list or union.
The function extprot.import_protocol() will dynamically load a protocol file
and convert it into the corresponding python class structure. This is quite
convenient while developing a protocol since it avoids an extra compilation
step, but it does add some startup overhead and requires the pyparsing module.
To compile a protocol definition into python sourcecode for the corresponding
class definitions, use the function extprot.compile_protocol() or pipe the file
through extprot/compiler.py like so:
$ cat mydefs.proto | python extprot/compiler.py > mydefs.py
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
extprot-0.2.4.tar.gz
(23.7 kB
view hashes)