Changelog

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

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Unreleased

.parse! will be removed in a future release. Use .parse instead, and validate the result using #valid? and #errors.

Nodaire::Indental and Nodaire::Tablatal now implement Enumerable, which allows you to iterate over the results with #each, #map etc.
These classes also implement [], so you can reference Indental data by category name, and Tablatal data by row index.
Convert Tablatal to JSON with Nodaire::Tablatal#to_json.

Support for Indental keys with blank values.
Two string normalization methods are now public: Nodaire.squeeze and Nodaire.symbolize.

Category names and keys are converted to upper case, except when symbolize_names is specified.
Any sequence of whitespace is converted into a space character. Punctuation is left intact.
symbolize_names converts any sequence of non-alphanumeric characters to _.

Strings are used for keys by default. To use symbols, pass the symbolize_names argument.
Category names and keys are normalised when detecting duplicates, and when converting them to symbols. This involves converting to lowercase and replacing each sequence of whitespace/underscores/dashes with _.
Indental parser detects when the input is wrapped in a JS template string.
Require Ruby >= 2.5.0. (Previously there was no minimum set, but it used a feature that required Ruby >= 2.6.0.)

.parse silently ignores duplicate keys and other errors, instead of raising an exception. This is paired with a new .parse! method which does raise an exception.