SchemaBind

SchemaBind reduces repetitive dictionary lookup boilerplate by exposing existing fields as readable Python attributes.

Instead of:

first_name = row["first_name"]
email = row["email_address"]
phone = row.get("phone_number")

Use:

from schemabind import bind

customer = bind(row)

customer.first_name
customer.email_address
customer.phone_number

SchemaBind is currently focused on dictionaries, including rows produced by csv.DictReader.

It does not create schemas, validate data, transform values, or infer new fields. It simply provides a cleaner way to access fields that already exist.

Example

import csv
from schemabind import bind

with open("samples/csv/customer.csv") as f:
    rows = list(csv.DictReader(f))

customer = bind(rows[0])

print(customer.first_name)
print(customer.email_address)
print(customer.phone_number)

Missing attributes raise AttributeError instead of silently returning None, which helps catch typos in field names.

Field Normalization

SchemaBind supports simple field-name normalization for attribute access.

For example:

customer = bind({"First Name": "Kaladin"})

print(customer.first_name)

The original dictionary keys are preserved for dictionary-style helpers:

customer.keys()
customer.get("First Name")

If multiple fields normalize to the same attribute name, SchemaBind raises ValueError instead of guessing which field to use.

For example:

bind({
    "First Name": "Dalinar",
    "first_name": "Shallan",
})

Both fields would normalize to first_name, so the binding is rejected as ambiguous.