Skip to content

Latest commit

 

History

History
106 lines (88 loc) · 3.5 KB

README.md

File metadata and controls

106 lines (88 loc) · 3.5 KB
Raw

brine

brine is a light-weight parser for Gherkin written in Java.

Intentions

brine is intended to produce a object representation of a Gherkin .feature file consistent with the implementation provided by cucumber/gherkin, with the exception that brine allows tags to include spaces. brine also enforces some additional rules, namely:

  • Only one tag is allowed per line
  • A Feature must contain at least one Scenario/Scenario Outline
  • An examples table must contain at least one data row (i.e. only a header row is not allowed)

brine uses a slightly simplified grammar for Gherkin that produces a simplified object model. Additionally, brine produces a single object (Feature) that captures all known information about a .feature, including its location.

Unlike cucumber/gherkin, brine produces no intermediate abstract syntax tree (AST); rather, it generates a Feature directly. Also note that brine does not associate tags hierarchically, i.e. a Scenario contains only the tags explicitly placed on the represented Scenario, and does not include tags on the Feature.

How to Use

The main entry point for the brine library is the PickleJar, which can produce a Feature from a File, Reader, or URI.

The various cure methods are fail-fast and will throw an exception if anything goes wrong. The cureCollectErrors methods accept a List<String> which will be populated with any errors encountered by the tokenizer or parser. The methods return the Feature representing the Gherkin, ignoring any invalid lines.

Exception Types

TokenizerException

thrown when a line is encountered that the tokenizer cannot interpret, e.g. a examples/table line without a terminating |

ParserException

thrown when a line is encountered that was properly tokenized, but violates the Gherkin grammar, e.g. a description line between two steps

Example

Take the example .feature below:

@NormalTag
@Meta("this tag has spaces")
Feature: README Feature
 
  Scenario: README Scenario
  A simple explanation of this scenario.
  Just a little more information.
 
    Given some initial condition
    When a user performs an action
    Then there is some result

Invoking PickleJar.cure() on the above produces a Feature with the below json representation:

{
   "absoluteLocation":"C:/experiment/brine/src/main/resources/readme_example.feature",
   "relativeLocation":"src/main/resources/readme_example.feature",
   "line":3,
   "tags":[
      "@NormalTag",
      "@Meta(\"this tag has spaces\")"
   ],
   "name":"README Feature",
   "scenarios":[
      {
         "line":5,
         "name":"README Scenario",
         "description":[
            "A simple explanation of this scenario.",
            "Just a little more information."
         ],
         "steps":[
            {
               "line":9,
               "keyword":"Given",
               "step":"some initial condition"
            },
            {
               "line":10,
               "keyword":"When",
               "step":"a user performs an action"
            },
            {
               "line":11,
               "keyword":"Then",
               "step":"there is some result"
            }
         ]
      }
   ]
}

Acknowledgements

brine is based heavily on the cucumber/gherkin implementation. Like that implementation, brine employs berp, an extensible parser generator.