Changelog

All notable changes to this project will be documented in this file.

For more information about changelogs, check Keep a Changelog and Vandamme.

3.3.0 - 7/19/26 Sun

[ENHANCEMENT] Tagged 2-tuples ({:ok, payload} / {:error, payload} — any {atom, payload} shape) are now transformed by every Digger operation: the payload is transformed, the tag is left untouched, and tuples nested inside maps and lists are handled transparently. Any other tuple shape (3-tuples, non-atom first element, {}) passes through unchanged, consistent with the library's existing passthrough convention. flatten/unflatten treat all tuples as leaf values per their documented non-map contract.

[ENHANCEMENT] New Digger.flatten/2 collapses a nested map into a single-level map of separator-joined string paths (%{a: %{b: 1}} → %{"a.b" => 1}), and new Digger.unflatten/2 rebuilds the nesting — the exact inverse for the maps flatten produces. The separator is configurable (separator:, default "."). Structs (calendar types included), lists, and empty maps are treated as leaf values and never descended into; non-map input passes through unchanged, like every other Digger operation.

[BUG FIX] Fixed deprecated negative-step String.slice/2 ranges in the internal camelize path (five call sites), which emitted a runtime deprecation warning on every camel_case call under Elixir 1.19+. Output is unchanged; a regression test now asserts camel_case writes nothing to stderr.
[DOCS] hexdocs now renders the README as the landing page with the CHANGELOG and a new "Recipes" sidebar group (ex_doc docs: config was previously absent), including a Phoenix/JS camelCase↔snake_case bridge guide composed from existing functions.
[DOCS] README rewritten to lead with the library's differentiators (zero runtime deps, atom-table-safe atomize(existing: true), key and value transforms, calendar-struct passthrough) and a copy-paste bridge recipe; dead Inch CI / CodeClimate badges removed, hexdocs badge added.
[DOCS] All seven protocol functions now carry iex> examples run as doctests on every test build; every example was verified against the library before being documented.

[ENHANCEMENT] Atomizer module now supports an existing: true option to atomize via String.to_existing_atom/1 instead of String.to_atom/1, so atomizing untrusted/external input can no longer grow the atom table unbounded; strings with no matching atom pass through unchanged rather than raising or minting a new one. Default behavior is unchanged.

[breaking change] - Stringifier module API updated to allow maps nested within lists to have their keys (and values) "stringified".
[breaking change] - CamelCaser module API updated to allow maps nested within lists to have their keys (and values) "camelized".
[breaking change] - Atomizer module API updated to allow maps nested within lists to have their keys (and values) "atomized".
[breaking change] - Dasher module API updated to allow maps nested within lists to have their keys (and values) "dasherized".
[breaking change] - LowerCaser module API updated to allow maps nested within lists to have their keys (and values) "lower-cased first".
[breaking change] - SnakeCaser module API updated to allow maps nested within lists to have their keys (and values) "snake-cased".
[breaking change] - UpperCaser module API updated to allow maps nested within lists to have the first letter of keys (and values) upcased

[BUG FIX] - Stringifier module updated to allow an atom not to be atomized, with respect to maps

[breaking change] - Atomizer module updated to apply #atomize to lists (including char lists).
[breaking change] - CamelCaser module updated to apply #camel_case to lists
[breaking change] - Dasher module updated to apply #dasherize to lists
[breaking change] - LowerCaser module updated to apply #lowercase_first to lists
[breaking change] - SnakeCaser module updated to apply #snake_case to lists
[breaking change] - Stringifier module updated to apply #stringify to lists
[breaking change] - UpperCaser module updated to apply #upcase_first to lists

[breaking change] Atomizer#atomize now expects a symbol :atomize instead of a string if you want to "atomize" a data type
[breaking change] Stringifier#stringify now expects a symobl :stringify instead of a string if you want to "stringify" a data type

[ENHANCEMENT] Atomizer module updated to use protocols
[ENHANCEMENT] Stringifier module updated to use protocols
[ENHANCEMENT] Floating point keys are now handled by the various protocols (see tests)

[FEATURE] CamelCaser module to camelize nested (and non-nested) map keys
[FEATURE] UpperCase module to upcase the first letter of nested (and non-nested) map keys
[FEATURE] LowerCase module to lowercase the first letter of nested (and non-nested) map keys
[FEATURE] Dasher module to dasherize the underscored nested (and non-nested) map keys

[FEATURE] Module to convert nested (and non-nested) map keys to "stringified" keys