engine: use sets for slide-range expansion, added slide selection and repeated-word offsets (#1756)

crates/engine/src/cmd/file.rs

@@ -397,6 +397,7 @@ pub fn pdf_options(p: &Value, count: usize) -> Result<deckcraft_pdf::PdfOptions>
 pub fn parse_slide_range(text: &str, count: usize) -> Result<Vec<usize>> {
     let invalid = || bad("file.export", format!("“{text}” isn't a valid slide range: use numbers and ranges such as 1-3, 5"));
     let mut out: Vec<usize> = vec![];
+    let mut seen: std::collections::HashSet<usize> = std::collections::HashSet::new();
     for part in text.split([',', ';']).map(str::trim).filter(|p| !p.is_empty()) {
         let (a, b) = match part.split_once('-') {
             Some((a, b)) => (a.trim(), b.trim()),
@@ -412,7 +413,7 @@ pub fn parse_slide_range(text: &str, count: usize) -> Result<Vec<usize>> {
             return Err(bad("file.export", format!("the presentation has {count} slides; “{part}” is outside them")));
         }
         for i in lo - 1..hi {
-            if !out.contains(&i) {
+            if seen.insert(i) {
                 out.push(i);
             }
         }
@@ -776,6 +777,16 @@ mod tests {
 
     use super::{parse_slide_range, write_file};
 
+    #[test]
+    fn broad_and_repeated_slide_ranges_expand_once() {
+        // 200 copies of a 100,000-slide range: each index once, in order, without a scan per index.
+        let text = vec!["1-100000"; 200].join(",");
+        let v = parse_slide_range(&text, 100_000).unwrap();
+        assert_eq!(v.len(), 100_000);
+        assert!(v.windows(2).all(|w| w[0] < w[1]));
+        assert_eq!(parse_slide_range("3, 1-3, 2", 3).unwrap(), vec![2, 0, 1]);
+    }
+
     #[test]
     fn slide_range_rejects_plus_sign() {
         assert!(parse_slide_range("+2", 3).is_err());

crates/engine/src/cmd/review.rs

@@ -367,7 +367,7 @@ fn spelling(s: &mut Session, p: &Value) -> Result<Value> {
             }
             out.push(v);
         }
-        let rep: Vec<usize> = crate::spell::repeated(para, &default).iter().map(|w| w.start).collect();
+        let rep: std::collections::HashSet<usize> = crate::spell::repeated(para, &default).iter().map(|w| w.start).collect();
         for w in crate::spell::paragraph_words(para, &default) {
             if ignore.contains(&w.text.to_lowercase()) || rep.contains(&w.start) {
                 continue;

crates/engine/src/cmd/slide.rs

@@ -408,8 +408,9 @@ fn select_slides(s: &mut Session, p: &Value) -> Result<Value> {
     };
     s.select(|doc, sel| {
         if add {
+            let mut have: std::collections::HashSet<_> = sel.slides.iter().copied().collect();
             for i in &ids {
-                if !sel.slides.contains(i) {
+                if have.insert(*i) {
                     sel.slides.push(*i);
                 }
             }