History

### 1.0.1
* Moved EBNF stuff to related ebnf gem, and use it to generate parser tables.
* Parser and lexer moved to generic EBNF modules.

### 1.0.0
* 1.0 release

### 0.3.1
* Sync with rdf-spec 0.3.11 changes due to respec updates.

### 0.3.0
* Use Ruby native EBNF parser. Still requires cwm to create first/follow rules.
* Updated to latest specification and EBNF rules.

### 0.1.1
* Improvements to allow reader & writer to be sub-classed for rdf-trig.
* Improved error recovery when parsing.

### 0.1.0
* Updates for RDF.rb 0.3.4.
* Format detection.

### 0.0.5
* Update turtle and writer specs to use latest version of Turtle specs.
* Use Spira Manifest w/list to order tests.
* Change debug to use blocks to reduce computation overhead when not doing debugging, results in a reasonable performance boost.

### 0.0.4
* Make an un-defined empty prefix an error, rather than treating it as <>.
* Replace remaining uses of SPARQL with RDF::Turtle or RDF::LL1

### 0.0.3
 * Completed RDF 1.1 Turtle based on https://www.w3.org/TR/2011/WD-turtle-20110809/
    * Reader
    * Writer
 * Issues:
    * IRI lexical representations
    * PNAMES not unescaped, should they be?
    * Assume undefined empty prefix is synonym for base
    * Can a list be used on it's own? Used in Turtle example.