HTML pipeline
How Guchho treats an HTML document as a build entry point.
#An HTML file is a first-class entry point. Guchho parses the document, finds every reference to code, styles and assets, adds them to the graph, and re-emits the document with rewritten URLs.
What counts as a reference
#Guchho follows, among others:
<script src="../...">and inline<script type="module"><link rel="stylesheet" href="../..."><img src="../...">,<source src="../...">,srcset- asset references discovered inside JavaScript and CSS
Markup output
#HTML is re-emitted from the parsed tree, so the printer options describe the output rather than the file:
pretty(default on) - indent the tree so the result is readable and editable.minifyHtml(default off) - collapse the markup.
Neither touches content where whitespace is significant, so turning either on is always safe.
Hashed names and URL rewriting
#Output file names can carry a content hash via the naming templates
(entryNames, chunkNames, assetNames, e.g. assets/[name]-[hash][ext]).
Every URL that points at a renamed file is rewritten to match, so the document
keeps working without manual edits.