This document defines the Ubiquitous Language for the Webmix project. Use these terms consistently in code, tests, feature files, and documentation.
The process of combining multiple extracted pages into a single, cohesive Markdown document.
- Synonyms: Compilation, Merging.
- Output: The "Webmix" file.
The mode of operation where the tool processes a complete website mirror (downloaded via wget) in one go.
- Context:
webmix.shscript.
The phase where the system identifies which files or pages belong to the target website.
- Mechanism: Currently filesystem walking (
os.walk) on a local mirror. - Artifact: A list of file paths or a tree structure.
The process of parsing an HTML file and converting its main content into Markdown.
- Tool:
trafilatura. - Features: Link rewriting, image preservation (alt text).
- Goal: Remove navigation, ads, footers, and boilerplate.
Alternative text provided for images, used for accessibility and preserved during extraction to maintain context in the Markdown output.
The transformation of internal hyperlinks (pointing to other HTML files) into internal Markdown anchors (pointing to sections within the aggregated file).
- Goal: Offline navigability.
- Format:
[Link Text](#target-anchor) (see: Target Title).
A directory containing a copy of a website's HTML files, preserving the original directory structure.
- Source: Usually created by
wget --mirror.
The specific heuristic used during Extraction to identify the "meat" of an article and discard the "shell" (nav, sidebar).
The process of creating the Table of Contents and Directory Structure visualization from the discovered files.
- Output: A hierarchical tree view reflecting the directory structure.
The hierarchical list of links generated at the beginning of the aggregated file, reflecting the structure of the Local Mirror.
The final single Markdown file produced by the Aggregation process.
- Structure: File Summary -> Directory Structure -> Files (Content).