add support for multiple output formats

[dhth]

Add a dedicated output module that renders deterministic output in four formats: html, plain, markdown, and terminal.

Each renderer relies on a shared EventPresentation shape that describes an event in display terms. This keeps event wording, link targets, and format-specific metadata in one place while letting each renderer decide how to present the final output.

[coderabbitai]

📝 Walkthrough

Walkthrough

This PR introduces a multi-format output rendering system for the ghlog CLI. It adds --output and --html-template options to the run subcommand, allowing users to select between HTML (with four template choices), Markdown, plain text, or colored terminal output. The implementation lifts event data into a presentation layer of human-readable fragments and event kinds, routes all formatting through a centralized dispatcher, and refactors the run handler to use the new output system instead of inline per-event formatting. Four new HTML template assets provide different visual styles for HTML output.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title 'add support for multiple output formats' accurately and clearly summarizes the main objective of the changeset, which introduces four output format renderers.
Description check	✅ Passed	The description directly addresses the core change: adding a dedicated output module with four format renderers (html, plain, markdown, terminal) that share an EventPresentation abstraction.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 3

🧹 Nitpick comments (1)

src/domain/events.rs (1)

111-113: ⚡ Quick win

Remove DeleteEvent::ref_path() since it’s unused and yields misleading tree/{ref_name} paths for deleted refs.

src/output/presentation.rs renders deletions with text(delete.ref_name()) and the only .ref_path() call sites are push.ref_path() / create.ref_path()—so DeleteEvent::ref_path() is currently dead code and a future footgun.

🧹 Proposed fix

 impl DeleteEvent {
     pub fn ref_name(&self) -> &str {
         strip_git_ref_prefix(&self.git_ref)
     }
-
-    pub fn ref_path(&self) -> String {
-        format!("tree/{}", self.ref_name())
-    }
 }

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/domain/events.rs` around lines 111 - 113, Remove the unused and
misleading DeleteEvent::ref_path() method from the DeleteEvent implementation;
delete the entire fn ref_path(&self) -> String { ... } for DeleteEvent and
ensure no call sites rely on it (only PushEvent::ref_path() and
CreateEvent::ref_path() should remain); verify presentation code renders
deletions via delete.ref_name() so behavior is unchanged and no other types
expect a ref_path on DeleteEvent.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@src/cmds/run.rs`:
- Around line 17-19: The early return on events.is_empty() unconditionally skips
all rendering (including HTML), so change the logic to only short-circuit for
non-HTML outputs: instead of returning immediately when events.is_empty(), check
the selected output format (the variable representing the format, e.g., format
or output_format) and only return Ok(()) when events.is_empty() AND the format
is not HTML (e.g., Format::Html); otherwise continue into the HTML rendering
path so an empty document/shell is produced. Ensure you update the branch around
events.is_empty() where it currently returns to reference the format enum/flag
and preserve the existing render_html/render() call flow for HTML.

In `@src/output/markdown.rs`:
- Around line 24-27: The fragment renderer inserts raw part.text into Markdown
which can break formatting; implement a helper like escape_markdown(text: &str)
-> String that escapes Markdown-sensitive characters (at minimum: \, `, *, _, {,
}, [, ], (, ), #, +, -, ., !) and converts newlines to spaces or escaped
newlines, then call this helper inside render_fragment for both the link label
and the plain text branch (i.e., replace uses of part.text.clone() with
escape_markdown(&part.text)); ensure the function is referenced from
render_fragment and named so it’s easy to find.

In `@src/output/terminal.rs`:
- Around line 33-37: render_fragment currently injects raw fragment.text into
OSC/ANSI sequences, enabling control-sequence injection; add a sanitizer and use
it before any terminal rendering. Implement a helper (e.g.,
sanitize_for_terminal) that removes or escapes terminal-control characters (ESC
0x1B, BEL, C0 controls, and any OSC/ST sequences) or percent-encodes unsafe
bytes, then call it on fragment.text inside render_fragment (both the Some(url)
branch and None branch) so only the sanitized string is embedded in format! and
returned.

---

Nitpick comments:
In `@src/domain/events.rs`:
- Around line 111-113: Remove the unused and misleading DeleteEvent::ref_path()
method from the DeleteEvent implementation; delete the entire fn ref_path(&self)
-> String { ... } for DeleteEvent and ensure no call sites rely on it (only
PushEvent::ref_path() and CreateEvent::ref_path() should remain); verify
presentation code renders deletions via delete.ref_name() so behavior is
unchanged and no other types expect a ref_path on DeleteEvent.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: ee2f0f5c-f7e3-4bbf-8439-2db64a0ccd6b

📥 Commits

Reviewing files that changed from the base of the PR and between 477451b and 6a2bae8.

⛔ Files ignored due to path filters (1)

• Cargo.lock is excluded by !**/*.lock

📒 Files selected for processing (16)

• Cargo.toml
• src/cli.rs
• src/cmds/command.rs
• src/cmds/run.rs
• src/domain/events.rs
• src/main.rs
• src/output/assets/templates/editorial.html
• src/output/assets/templates/notebook.html
• src/output/assets/templates/terminal.html
• src/output/assets/templates/zine.html
• src/output/html.rs
• src/output/markdown.rs
• src/output/mod.rs
• src/output/plain.rs
• src/output/presentation.rs
• src/output/terminal.rs