tappoz / ChemicalIngredients

A parser for strings containing information related to chemical ingredients. Using ANTLR4 for the parsing stage, the system makes sense of amounts (also as ranges), units of measure, ingredient description.

Chemical Ingredients parser

This Java project aims at parsing strings containing information related to chemical ingredients when experimenting your own recipes. Having fun!

The need is related to make sense of strings containing quantities (perhaps as ranges), units of measure, and ingredient names. There might be different scenarios when parsing these strings:

• 5 grams of calcium carbide we want to identify
  • 5.0 a Double object;
  • grams a validated unit of measure;
  • calcium carbide the name of this ingredient.
• 5 tablespoons (70 grams) of lye becomes:
  • 5.0 a Double, tablespoon;
  • 70.0 a Double, grams;
  • lye.
• 1 1/2 cups KNO3
  • 1.5 a Double object;
  • cup a validated unit of measure;
  • KNO3 the name of the ingredient.
• When a range is used e.g. 1 to 3 or 5 or 6 then a value representing the mean is provided.

The Context Free grammar with the parsing rules is generated with ANTLR4, there is some usage of regex as well. The Spring framework is used for dependency injection. The Logback framework is used for logging purposes. For the testing part Hamcrest and JUnit are used in combination with Spring Test for the integration tests, with Mockito for the unit tests. Some utility libraries are used to deal with string manipulation and collections manipulation (lists): Apache Commons and Google Guava. The Gson library is used in some cases to serialise the Java beans as JSON objects for debugging/logging purposes.

The BigDecimal class is used when dealing with divisions and double objects/types in order to keep the rounding precision. A Java enum class is used to validate the units of measure.

How to use it

The class containing the main method is: org.tappoz.run.ChemicalIngredientsRunner.

The project is built using Maven.

• In order to generate the Java source code regarding the ANTLR grammar use this command: mvn clean source:jar -e -X.
• To see the dependencies among different third party jars (e.g. Hamcrest used by JUnit) run: mvn dependency:tree -Dverbose.