General Idea

Syntactic coloring uses 2 paradigms which are used on different levels:

On the first levels, the context being very important, especially with the absence of quotation marks for strings, it is necessary to use an automaton (with states and transitions) in order to color these levels.
For the following levels, the context being less important, we can simply locate keywords, and color them.

The final rules will be available in the highlightingRules folder especially for tests and for the syntaxModesRules.ts file

File roles

highlightingRules/highlighting-*.json this file contains the final rules that have been generated for syntax highlighting.
definition.py definition of constants (space, start of line, etc...).
rules_automaton.py definitions of the coloring automata for the first levels.
list_keywords.py keyword lists for levels where context is no longer important.
rules_list.py file that allows to generate the syntax coloring rules from the file list_keywords.py
generate-rules-highlighting.py This file generates the rules for each level, and for each language, from the rules_automaton.py and rules_list.py

If you want to modify the syntax highlighting on a level lower than 3, you have to modify the associated function which creates the highlighting automaton in the file rules_automaton.py
If you want to modify the syntax highlighting on the following levels, you have to modify the KEYWORDS and NUMBER variables in the list_keywords.py file, at the desired level.

To regenerate the json files containing the regexes, it is the generate-rules-highlighting.py file that you have to execute, it will allow you to see in the highlighting-*.json files the new regexes
If you want to see the changes in Hedy, you will also have to regenerate the javascript. This is done by running the ./build-tools/heroku/generate-typescript file