spaghettimap 0.5.0

Map Pydantic models safely, using JMESPath

spaghettimap

A Python library for pydantic model-to-model conversion powered by JMESPath.

Features

• Map any pydantic BaseModel to another using a declarative schema
• Schema values can be JMESPath expressions, Python callables, or a dict combining both with an optional transform
• Full support for all JMESPath built-in functions (length, sort, max_by, contains, join, keys, to_string, …)
• Custom JMESPath functions via jmespath.functions.Functions subclass
• Filter expressions, multi-select hash/list, pipe expressions, wildcards, and or-expressions
• Solid error handling with ConfigurationError, MappingError, and FieldMappingError – all with clear, field-specific messages
• Fail-fast config checks for invalid JMESPath expressions and schema fields missing from the target model
• Pydantic validators (@field_validator, @model_validator) and type coercion run on the target model automatically
• map_many() for batch conversion of model lists

Installation

pip install spaghettimap
# or with uv
uv add spaghettimap

Quick Start

from pydantic import BaseModel
from spaghettimap import Mapper, MappingConfig

class Source(BaseModel):
    first_name: str
    last_name: str
    contact: dict  # {"email": "...", "phone": "..."}
    tags: list[dict]  # [{"name": "...", "weight": 1.0}]

class Target(BaseModel):
    full_name: str
    email: str
    tag_count: int
    upper_name: str

mapper = Mapper()
mapper.add_config(
    MappingConfig(
        from_type=Source,
        to_type=Target,
        schema={
            # Python callable
            "full_name": lambda d: f"{d['first_name']} {d['last_name']}",
            # Nested JMESPath expression
            "email": "contact.email",
            # JMESPath built-in function
            "tag_count": "length(tags)",
            # JMESPath expression + Python transform
            "upper_name": {"expression": "first_name", "transform": str.upper},
        },
    )
)

result: Target = mapper.map(source_instance, Target)

Schema Value Types

Type	Description	Example
str	JMESPath expression	"contact.email", "tags[*].name", "length(tags)"
Callable[[dict], Any]	Python function receiving the full source dict	lambda d: d["x"] + d["y"]
dict	{"expression": str|Callable, "transform": Callable}	{"expression": "price", "transform": lambda p: f"£{p:.2f}"}

Custom JMESPath Functions

import jmespath.functions
from spaghettimap import Mapper, MappingConfig

class MyFunctions(jmespath.functions.Functions):
    @jmespath.functions.signature({"types": ["string"]})
    def _func_upper(self, value: str) -> str:
        return value.upper()

mapper.add_config(
    MappingConfig(
        from_type=Source,
        to_type=Target,
        schema={"name": "upper(first_name)"},
        custom_functions=MyFunctions(),
    )
)

Batch Mapping

results: list[Target] = mapper.map_many(source_list, Target)

Error Hierarchy

SpaghettimapMapperError
├── ConfigurationError   – invalid config (bad types, missing keys, unregistered pair)
└── MappingError         – runtime mapping failure
    └── FieldMappingError – failure for a specific field (has .field attribute)