This is a simple but complete example of using ANTLR 4 as the basis for developing a translator app. Using the app or any of the ANTLR 4 tools requires the user to have a local copy of version 4.9.2 of the ANTLR 4 jar file which is availble here: https://www.antlr.org/download.html

You will also need a local copy of the the JRE or Java Runtime environment in order to run the executable. Your machine may already have this. if so your machine will recognize the "Java" command. If not it can be download from here: https://www.java.com/en/download/manual.jsp

Once you have the ANTLR 4 jar file and the Java command locally you can run the mini.jar executable directly on all the operating systems I'm familiar with; the xjar Windows batch file shows how I do iton Windows 10 using a command line when in a directory with all the necessary files.

The braille being backtranslated is a very tiny subset of contracted English braille. (It is assumed the user is familiar with braille.) The input is the 34-character subset of ASCII braille which represents letters, a comma, a period (full stop), the capital letter indicator, and the symbols '&', '='. '('. "!' ')' for five contractions. The last five always represent the contracted items. They are the only contractions; the letters are always backtranslated as letters, not as alphabetic wordsigns. (Note that the as far as entering letters in ASCII braille, the lexer doesn't make a distinction between capital and small letters.) Looking at the rules in miniLexer.g4 might be helpful as far as understanding the mini-braille system.

The developer creates 3 items: a lexer grammar, a parser grmmmar, and a translator. The ANTLR tool, which is written in Java, converts the two gammars (files with .g4 extennsions) to code in the users's choice of programming language with Java being the default which is what is done here. The ANTLR tool also generates a template (here a Java class file) for a translator based on the parser. The developer then manually extends the template and adds supporting code. The simple translator here, which is self-contained, is miniBack.java. Mayne.java, based on samples from ANTLR 4 documentation, couples the lexer, parser, and translator.

A introductory three-page description of the three-phase process starts on this wiki page:

https://github.com/SusanJ/BasicUEB/wiki/4.1--An-Introduction-to-Parsing-and-Lexical-Analysis