architecture.md

Architecture Overview

jxlint aims to be pretty simple, both in terms of usability and architecture.

The only required integration for your application is the Jxlint class. By creating an instance of it, and calling parseArgumentsAndDispatch, the logic for executing rules is handled.

What is that logic? First, we parse the command line arguments. If there are any fatal errors, we abort and display an informative error message. Otherwise, we build a ProgramOptions instance based on the command line parameters. This instance is passed to the Dispatcher, which is responsible for setting up and choosing the appropriate action based on the program options.

Assuming jxlint was asked to validate rules, it will then perform all the validations, each rule validating in its own thread. A list of errors is collected and passed to a Reporter class, which is responsible for reporting rule violations. The kind of reporter (i.e. HTML, XML, or CLI) created is based on the program options that were passed, with a default value of CLI. Every Reporter can be overridden by calling Reporters.setCustomReporter before calling Jxlint#parseArgumentsAndDispatch.

The packages are summarized below:

• com.selesse.jxlint.actions contains classes that perform actions that the Dispatcher calls on. An example might be printing out all the rules, or printing out the Markdown rule dump.

• com.selesse.jxlint.cli contains logic for parsing / handling the command line parameters passed to jxlint.

• com.selesse.jxlint.linter contains the logic for calling the functions to perform the validations.

• com.selesse.jxlint.model contains model information. Everything in here is mostly plain old Java objects. The important classes here are LintRule and LintError, the core of jxlint.

• com.selesse.jxlint.report contains classes related to reporting.

• com.selesse.jxlint.settings contains "settings" information, like the program name and the program version.