escode 1.0.0

ESCODE binary serialization

Fast Serializing and Deserializing of Python types.

escode is a Python C-Extension which quickly serializes and deserializes arbitrary base types into binary data. The serialization has also been designed to be small. This is useful for storage and network applications. Finally, the module provides index serialization for a tuple of data - where the serialization maintains sorting order against all other tuples containing the same types in the same order (or prefixes).

import escode

data = {
    "id": 1234,
    "name":"Akhil",
    "usernames": ["akhil", "awable"],
    "location": {"lat": 28.7041, "lng":77.1025}
}

blob = escode.encode(data)
db.put(<id>,..., blob)

...

dbdata = escode.decode(db.get(<id>))
assert dbdata == data

Index encoding is useful for applications where a multi-attribute index must be maintained in a single binary field, while maintaining the sort order of the original tuples.

tuples = [
    (1234, 'Akhil', 13.014492753623188),
    (831, 'Rohit', 1.3086614173228346),
    (844, 'Poornima', 1.2375366568914956),
    (723, 'Mark', 2.0716332378223496),
    (874, 'John', 4.065116279069767),
    (94, 'Takkar', 0.09456740442655935)
]

tuple_encodings = [
    (t, escode.encode_index(t))
    for t in tuples]

# sorting by tuples is the same as sorting by the encodings
assert (sorted(tuple_encodings, key=lambda tup_enc: tup_enc[0]) ==
        sorted(tuple_encodings, key=lambda tup_enc: tup_enc[1]))

A quick note on implementation: Index order of tuples can be tricky since one must maintain tuple element boundaries which compare lower than any data. i.e. ('a','z') < ('aa', 'z') but 'az' > 'aaz' This is accomplished in escode by using '\x00\x00' as the boundary, and escaping \x00s in the tuple elements themselves, but compressing consecutive \x00s to preserve space.