vinzenz/hyperhive
vinzenz/hyperhive
Watch
0
0
Fork
You've already forked hyperhive
 0
Code Activity

review classifier: plain table cells, no backslash escapes

Browse source
This commit is contained in:
atlas 2026-10-16 21:31:48 +02:00 • committed by mara
commit a2dca32bb8
3 changed files with 57 additions and 18 deletions
Download patch file Download diff file Expand all files Collapse all files

60
swarm-controller/src/review_classifier/review.rs
View file

@ -842,26 +842,37 @@ fn overall_line(overall: Overall) -> &'static str {
}
}
/// `text` as one table cell's inline text: on one line, with every character
/// that could end the cell, start markup, math (`$`) or raw HTML escaped, and a zero-width
/// space after every `@`, `#`, `!`, `://` and `www.`, so it carries no live
/// mention, issue or PR reference, or link. The forge links issue and PR
/// numbers and bare URLs itself, so the escapes alone don't stop those.
/// `text` as one table cell's inline text: on one line, with a backslash before
/// every character that could end the cell, start markup, math (`$`) or raw
/// HTML, and a zero-width space after every `@`, `#`, `!`, `://` and `www.`, so
/// it carries no live mention, issue or PR reference, or link. The forge links
/// issue and PR numbers and bare URLs itself, so the escapes alone don't stop
/// those; `#` and `!` get only the zero-width space.
///
/// `_` is escaped except between two alphanumerics, where it can't open
/// emphasis (`ci_metrics.rs`). Other punctuation, such as `(`, `)`, `.` and
/// `:`, passes through: it starts no markup inside a table cell, and the `[`
/// that would make `(` a link target is escaped.
fn cell(text: &str) -> String {
let mut out = String::with_capacity(text.len());
for (i, word) in text.split_whitespace().enumerate() {
if i > 0 {
out.push(' ');
}
for c in word.chars() {
let chars: Vec<char> = word.chars().collect();
for (j, &c) in chars.iter().enumerate() {
match c {
'\\' | '|' | '`' | '*' | '_' | '~' | '[' | ']' | '(' | ')' | '<' | '>' | '&'
| '$' => {
'_' if j > 0
&& chars[j - 1].is_alphanumeric()
&& chars.get(j + 1).is_some_and(|n| n.is_alphanumeric()) =>
{
out.push(c);
}
'\\' | '|' | '`' | '*' | '_' | '~' | '[' | ']' | '<' | '>' | '&' | '$' => {
out.push('\\');
out.push(c);
}
'#' | '!' => {
out.push('\\');
out.push(c);
out.push('\u{200b}');
}
@ -1932,7 +1943,7 @@ mod tests {
.count();
assert_eq!(unescaped, 5, "{line}");
assert!(
line.contains("\\!\u{200b}\\[\\]") && line.contains("\\<b\\>"),
line.contains("!\u{200b}\\[\\]") && line.contains("\\<b\\>"),
"{line}"
);
}
@ -1965,8 +1976,33 @@ mod tests {
let text = cell("see https://evil.example/x, WWW.evil.example, #1, o/r#3 and !4");
assert_eq!(
text,
"see https://\u{200b}evil.example/x, WWW.\u{200b}evil.example, \\#\u{200b}1, \
o/r\\#\u{200b}3 and \\!\u{200b}4"
"see https://\u{200b}evil.example/x, WWW.\u{200b}evil.example, #\u{200b}1, \
o/r#\u{200b}3 and !\u{200b}4"
);
}
#[test]
fn a_plain_reason_reads_without_backslashes() {
assert_eq!(
cell("an added #[allow(clippy::foo)] in ci_metrics.rs (see main_loop_v2.rs)"),
"an added #\u{200b}\\[allow(clippy::foo)\\] in ci_metrics.rs (see main_loop_v2.rs)"
);
}
#[test]
fn an_underscore_escapes_unless_it_sits_between_alphanumerics() {
// Presence control: the edge underscores are escaped.
assert_eq!(
cell("_em_ __b__ a_ _b a__b snake_case"),
"\\_em\\_ \\_\\_b\\_\\_ a\\_ \\_b a\\_\\_b snake_case"
);
}
#[test]
fn a_reason_with_markup_characters_keeps_them_inert() {
assert_eq!(
cell(r"a\b `c` *d* ~e~ [f] <g> &h; i|j $k"),
r"a\\b \`c\` \*d\* \~e\~ \[f\] \<g\> \&h; i\|j \$k"
);
}