fix(render): fix code block rendering and quote escaping

Two issues fixed:

1. Language-less code blocks (``` without lang) were not accumulating
   text content. The guard `code_block_lang.is_some()` was false for
   them, so content fell through to regular text rendering.

   Fix: Add `in_code_block` flag to track code block state separately
   from language presence.

2. Single quotes in code blocks were being HTML-escaped as ',
   breaking CSP headers like 'self' in documentation.

   Fix: Create code_escape/code_escape_into in escape.rs that only
   escapes <, >, & (required to prevent HTML tag injection) but
   preserves quotes (safe inside <pre><code> content).

Rationale for code_escape:
- < and > MUST be escaped to prevent browser interpreting code as HTML
- & MUST be escaped to prevent HTML entity interpretation
- Quotes are safe inside element content (no attribute context)

Also:
- Add test for unlabeled code block quote preservation

All 71 tests pass.
This commit is contained in:
Timothy DeHerrera
2026-02-05 17:26:17 -07:00
parent 1e5ed28788
commit 8df8aa434f
3 changed files with 55 additions and 11 deletions

View File

@@ -25,6 +25,27 @@ pub fn html_escape_into(out: &mut String, s: &str) {
}
}
/// Escape characters for safe embedding in code blocks.
///
/// Only escapes `&`, `<`, `>` — quotes are safe inside `<pre><code>`.
pub fn code_escape(s: &str) -> String {
let mut result = String::with_capacity(s.len());
code_escape_into(&mut result, s);
result
}
/// Escape code block characters into an existing string.
pub fn code_escape_into(out: &mut String, s: &str) {
for c in s.chars() {
match c {
'&' => out.push_str("&amp;"),
'<' => out.push_str("&lt;"),
'>' => out.push_str("&gt;"),
_ => out.push(c),
}
}
}
/// Escape XML special characters for safe embedding in XML documents.
///
/// Escapes: `&`, `<`, `>`, `"`, `'`

View File

@@ -8,7 +8,7 @@ use std::collections::HashMap;
use std::sync::LazyLock;
use std::time::Duration;
use crate::escape::{html_escape, html_escape_into};
use crate::escape::{code_escape, code_escape_into};
use ropey::RopeSlice;
use tree_house::highlighter::{Highlight, HighlightEvent, Highlighter};
use tree_house::{
@@ -565,14 +565,14 @@ pub fn highlight_code(lang: Language, source: &str) -> String {
// Check if we have a config for this language
if !loader.configs.contains_key(&lang) {
return html_escape(source);
return code_escape(source);
}
// Parse the syntax tree
let rope = RopeSlice::from(source);
let syntax = match Syntax::new(rope, lang.to_th_language(), Duration::from_secs(5), loader) {
Ok(s) => s,
Err(_) => return html_escape(source),
Err(_) => return code_escape(source),
};
// Create highlighter and render
@@ -595,7 +595,7 @@ fn render_html<'a>(source: &str, mut highlighter: Highlighter<'a, 'a, SukrLoader
let end = next_pos as usize;
if start < source.len() {
let text = &source[start..end.min(source.len())];
html_escape_into(&mut html, text);
code_escape_into(&mut html, text);
}
}
@@ -672,7 +672,7 @@ mod tests {
#[test]
fn test_html_escape() {
let escaped = html_escape("<script>alert('xss')</script>");
let escaped = code_escape("<script>alert('xss')</script>");
assert!(!escaped.contains('<'));
assert!(escaped.contains("&lt;"));
}

View File

@@ -1,6 +1,6 @@
//! Markdown to HTML rendering via pulldown-cmark with syntax highlighting.
use crate::escape::html_escape;
use crate::escape::{code_escape, html_escape};
use crate::highlight::{highlight_code, Language};
use pulldown_cmark::{CodeBlockKind, Event, HeadingLevel, Options, Parser, Tag, TagEnd};
use serde::Serialize;
@@ -30,6 +30,7 @@ pub fn markdown_to_html(markdown: &str) -> (String, Vec<Anchor>) {
let mut anchors = Vec::new();
let mut code_block_lang: Option<String> = None;
let mut code_block_content = String::new();
let mut in_code_block = false;
// Image alt text accumulation state
let mut image_alt_content: Option<String> = None;
@@ -54,9 +55,10 @@ pub fn markdown_to_html(markdown: &str) -> (String, Vec<Anchor>) {
}
CodeBlockKind::Indented => None,
};
in_code_block = true;
code_block_content.clear();
}
Event::Text(text) if code_block_lang.is_some() => {
Event::Text(text) if in_code_block => {
// Accumulate code block content
code_block_content.push_str(&text);
}
@@ -100,13 +102,14 @@ pub fn markdown_to_html(markdown: &str) -> (String, Vec<Anchor>) {
} else {
html_output.push('>');
}
html_output.push_str(&html_escape(&code_block_content));
html_output.push_str(&code_escape(&code_block_content));
}
html_output.push_str("</code></pre>\n");
}
code_block_lang = None;
in_code_block = false;
code_block_content.clear();
}
Event::Text(text) if heading_level.is_some() => {
@@ -283,12 +286,12 @@ fn start_tag_to_html(tag: &Tag) -> String {
dest_url, title, ..
} => {
if title.is_empty() {
format!("<a href=\"{}\">", html_escape(&dest_url))
format!("<a href=\"{}\">", html_escape(dest_url))
} else {
format!(
"<a href=\"{}\" title=\"{}\">",
html_escape(&dest_url),
html_escape(&title)
html_escape(dest_url),
html_escape(title)
)
}
}
@@ -496,4 +499,24 @@ Config details.
"special chars in src should be escaped"
);
}
#[test]
fn test_unlabeled_code_block_preserves_quotes() {
// Code block without language specifier should preserve quotes
let md = "```\nContent-Security-Policy: default-src 'self';\n```";
let (html, _) = markdown_to_html(md);
// Should be inside <pre><code>
assert!(html.contains("<pre><code>"), "should have code block");
// Quotes should NOT be escaped (only <, >, & need escaping in code)
assert!(
html.contains("'self'"),
"single quotes should be preserved in code blocks"
);
// Should NOT have escaped quotes
assert!(
!html.contains("&#39;"),
"quotes should not be HTML-escaped in code blocks"
);
}
}