panamap 1.2.3

Python object mapper

panamap

Panamap is a Python object mapper. It is useful to avoid boilerplate code when copying data between objects with similar data, for example protobuf generated files and domain models.

Installation

Use the package manager pip to install panamap.

pip install panamap

Usage

Mapping primitive values

The most simple usage of panamap is to map primitive values:

from panamap import Mapper

mapper = Mapper()
print(mapper.map("123", int) + 1)
# 124

There is a set of standart primitive values converters.

Mapping classes with converter

The most straitforward usage of mapper is to set converter from class A to class B. Here is an example of such setup:

from panamap import Mapper

class A:
    def __init__(self, a_value: int):
        self.a_value = a_value


class B:
    def __init__(self, b_value: int):
        self.b_value = b_value

mapper = Mapper()
mapper.mapping(A, B) \
    .l_to_r_converter(lambda a: B(a.a_value)) \
    .register()

b = mapper.map(A(123), B)
print(b.b_value)
# 123

If there is limited set of values, enums are most common case, you can use utility function values_map:

from enum import Enum
from panamap import Mapper, values_map

class LangA(Enum):
    PYTHON = 1
    JAVA = 2
    CPP = 3


class LangB(Enum):
    PY = 1
    JAVA = 2
    CPP = 3


mapper = Mapper()
mapper.mapping(LangA, LangB) \
    .l_to_r_converter(values_map({
        LangA.PYTHON: LangB.PY,
        LangA.JAVA: LangB.JAVA,
        LangA.CPP: LangB.CPP,
    })) \
    .register()

print(mapper.map(LangA.PYTHON, LangB).name)
# PY

Mapping context

In some cases you need to pass some context to mapping operation. You can do it with context parameter in map method. Context is a dict and will be passed to converter as a second argument.

from panamap import Mapper


class A:
    def __init__(self, a_value: int):
        self.a_value = a_value


class B:
    def __init__(self, b_value: int, contextual: str):
        self.b_value = b_value
        self.contextual = contextual


mapper = Mapper()
mapper.mapping(A, B) \
    .l_to_r_converter(lambda a, ctx: B(a.a_value, ctx["contextual"])) \
    .register()

b = mapper.map(A(123), B, {"contextual": "contextual value"})
print(b.contextual)
# 'contextual value'

Mapping simple classes

To set up mapping call mapping function of mapper object. Each field pair can be bind with bidirectional function or separately with l_to_r and r_to_l if we only need one directional mapping or there is custom conversion on mapping.

Here are some examples:

from panamap import Mapper

class A:
    def __init__(self, a_value: int):
        self.a_value = a_value


class B:
    def __init__(self, b_value: int):
        self.b_value = b_value

mapper = Mapper()
mapper.mapping(A, B) \
    .l_to_r("a_value", "b_value") \
    .register()

b = mapper.map(A(123), B)
print(b.b_value)
# 123
# a = mapper.map(B(123), A) will raise MissingMappingException cause we didn't set any r_to_l map rules

bidirectional_mapper = Mapper()
bidirectional_mapper.mapping(A, B) \
    .bidirectional("a_value", "b_value") \
    .register()

b = bidirectional_mapper.map(A(123), B)
print(b.b_value)
# 123
a = bidirectional_mapper.map(B(123), A)
print(a.a_value)
# 123

shifting_mapper = Mapper()
shifting_mapper.mapping(A, B) \
    .l_to_r("a_value", "b_value", lambda a: a + 1) \
    .r_to_l("a_value", "b_value", lambda b: b - 1) \
    .register()

b = shifting_mapper.map(A(123), B)
print(b.b_value)
# 124
a = shifting_mapper.map(B(123), A)
print(a.a_value)
# 122

Mapping empty classes

Sometimes there is need to convert one empty class to another. For such case there is _empty versions of map config functions:

from panamap import Mapper

class A:
    pass

class B:
    pass

mapper = Mapper()
mapper.mapping(A, B) \
    .bidirectional_empty() \
    .register()

b = mapper.map(A(), B)
print(isinstance(b, B))
# True

Mapping nested fields

Panamap supports mapping of nested fields. To perform this mapping for nested fields classes must be set up.

from dataclasses import dataclass
from panamap import Mapper

@dataclass
class NestedA:
    value: str


@dataclass
class A:
    value: NestedA


@dataclass
class NestedB:
    value: str


@dataclass
class B:
    value: NestedB

mapper = Mapper()
mapper.mapping(A, B) \
    .map_matching() \
    .register()
mapper.mapping(NestedA, NestedB) \
    .map_matching() \
    .register()

b = mapper.map(A(NestedA("abc")), B)
print(isinstance(b.value, NestedB))
# True
print(b.value.value)
# abc

Mapping from and to dict

Panamap allow to set up mapping frm and to dict object. Here is an example:

from typing import List
from dataclasses import dataclass
from panamap import Mapper

@dataclass
class Nested:
    value: str


@dataclass
class A:
    nested: Nested
    list_of_nested: List[Nested]

mapper = Mapper()
mapper.mapping(A, dict) \
    .map_matching() \
    .register()
mapper.mapping(Nested, dict) \
    .map_matching() \
    .register()

a = mapper.map(
    {
        "nested": {
            "value": "abc",
        },
        "list_of_nested": [
            {"value": "def",},
            {"value": "xyz",},
        ]
    },
    A,
)
print(a)
# A(nested=Nested(value='abc'), list_of_nested=[Nested(value='def'), Nested(value='xyz')])

Mapping protobuf generated classes

To map protobuf generated classes use separate module panamap-proto.

Contributing

Contributing described in separate document.