- Use Path.expand(Path.join(app_root, path)) to correctly resolve both
app files (lib/...) and blogex files (../blogex/...)
- Previously Path.join alone did not resolve '..' components, causing
all source viewers to show the wrong file content
- Sort source files by mtime (newest first) and take only the first 2
- This limits the page scope for easier testing/investigation
- Temporary limit; will be 10 most recently changed files later
- Updated tests to work with the 2-file limit
- Install highlight.js via npm with 8 language definitions
- SourceViewer hook applies syntax highlighting on mount/update
- Atom One Dark theme for dark mode, Atom One Light for light mode
- Add hooks directory, package.json, and package-lock.json
- Add demo document and screenshots
- Add rodney-docker.md documentation
- Ignore .rodney/ chrome data directory
And particular for this project. I am now reading more markdown, making
smaller steps. So the installation instructions went to the README.
Hmm. having just written about init scripts, the installation may be
best moved to an init script in the skill directory.
also added go to install showboat and rodney for the next steps.
microprints work in qwan tracker, but not in firehose. some of the
rendering is in the project, maybe the library should provide sample
webpages.
- Drop redundant :ok return in _validate_links/2 (blog.ex)
- Remove dead HTML link regex from extract_links/1 (body is raw markdown)
- Rename slug_slug_end/1 to slug_end/1
- Simplify parse_blog_link/1 to return {blog_id, slug}, removing
parse_query_fragment/1 and dead case branches
Manually refactor multi-line conn = post/put(conn, ...) patterns
across user_session, user_settings, and user_registration controller
tests. Rename shadowed conn to response using pipeline operator.
Also add plans/multi-line-conn-refactoring.md spec for future
Elixir-based tooling to handle these patterns automatically.
