Package org.machanism.machai.gw.reviewer

package org.machanism.machai.gw.reviewer

Provides the file-format-specific reviewer implementations used to discover and normalize embedded @guidance instructions for the Ghostwriter documentation pipeline.

The package defines the Reviewer service provider interface and its format adapters. A reviewer understands the comment or document syntax of a supported format, determines whether the file contains a guidance tag, and returns a prompt fragment containing the context required by the downstream guidance processor. Implementations return null when a file does not contain applicable guidance and may throw IOException when its content cannot be read.

The available reviewers are:

• JavaReviewer, which handles Java source files, including package-level guidance in package-info.java;
• MarkdownReviewer, which handles Markdown files with guidance in HTML comments;
• HtmlReviewer, which handles HTML, HTM, and XML comment syntax;
• PythonReviewer, which handles Python line comments and triple-quoted guidance blocks;
• TypeScriptReviewer, which handles TypeScript line and block comments;
• PumlReviewer, which handles PlantUML files; and
• TextReviewer, which handles files named @guidance.txt.

Each implementation exposes its supported extensions through Reviewer.getSupportedFileExtensions(). A caller can use those extensions to select a reviewer, then invoke Reviewer.perform(java.io.File, java.io.File). The first argument is the project root and the second is the file being reviewed; the root is used to provide stable project-relative path context in the result.

For example, a caller can try the reviewer associated with a file's extension and forward only applicable results:

 Reviewer reviewer = new JavaReviewer();
 String prompt = reviewer.perform(projectDirectory, sourceFile);
 if (prompt != null) {
     guidanceProcessor.process(prompt);
 }
 

Reviewers are intentionally format-specific: they identify guidance according to the syntax they support rather than treating every file as plain text. This keeps extraction rules isolated and allows additional formats to be introduced by adding another Reviewer implementation.

Class	Description
HtmlReviewer	Reviewer implementation for HTML and XML files (.html, .htm, .xml).
JavaReviewer	Reviewer implementation for Java source files (.java).
MarkdownReviewer	Reviewer implementation for Markdown files (.md).
PumlReviewer	Reviewer implementation for PlantUML files (.puml).
PythonReviewer	Reviewer implementation for Python source files (.py).
Reviewer	Service-provider interface (SPI) for components that scan project files and extract @guidance instructions for downstream processing.
TextReviewer	Reviewer implementation for generic guidance text files.
TypeScriptReviewer	Reviewer implementation for TypeScript source files (.ts).