Analysis ToolPak: stop building a report once it outgrows the output limit

Report builders filled their whole output map before write_report
checked its size, so Descriptive Statistics grouped by rows over a
tall, nearly empty column allocated about 16 entries for each of up
to a million variables before refusing the report.

Grid now tracks its extent as values are stored and keeps nothing
once the report's rows x columns pass MAX_OUTPUT_CELLS; write_report
refuses such a report. extent() no longer rescans every entry.

crates/engine/src/cmd/analysis.rs

@@ -272,11 +272,25 @@ struct Grid {
     cells: BTreeMap<(u32, u32), Value>,
     /// Cells shown as percentages.
     pct: Vec<(u32, u32)>,
+    /// Rows and columns spanned by `cells`.
+    rows: u32,
+    cols: u32,
+    /// A cell was dropped because the report outgrew [`MAX_OUTPUT_CELLS`]: the report is refused.
+    too_large: bool,
 }
 
 impl Grid {
+    /// Stores a value. Past [`MAX_OUTPUT_CELLS`] (the most a report writes) nothing more is kept,
+    /// so a builder that would produce millions of entries stops allocating them, and
+    /// [`write_report`] refuses the report.
     fn put(&mut self, r: u32, c: u32, v: Value) {
         if !v.is_empty() {
+            let (rows, cols) = (self.rows.max(r.saturating_add(1)), self.cols.max(c.saturating_add(1)));
+            if self.too_large || u64::from(rows) * u64::from(cols) > MAX_OUTPUT_CELLS {
+                self.too_large = true;
+                return;
+            }
+            (self.rows, self.cols) = (rows, cols);
             self.cells.insert((r, c), v);
         }
     }
@@ -296,9 +310,7 @@ impl Grid {
     }
     /// (rows, cols) spanned.
     fn extent(&self) -> (u32, u32) {
-        let rows = self.cells.keys().map(|(r, _)| r + 1).max().unwrap_or(0);
-        let cols = self.cells.keys().map(|(_, c)| c + 1).max().unwrap_or(0);
-        (rows, cols)
+        (self.rows, self.cols)
     }
 }
 
@@ -1749,7 +1761,7 @@ fn sampling(s: &Session, p: &Json, seed: Option<u64>) -> Result<Grid> {
 fn write_report(s: &mut Session, p: &Json, tool: &str, g: Grid) -> Result<Json> {
     let (rows, cols) = g.extent();
     let (rows, cols) = (rows.max(1), cols.max(1));
-    if u64::from(rows) * u64::from(cols) > MAX_OUTPUT_CELLS {
+    if g.too_large || u64::from(rows) * u64::from(cols) > MAX_OUTPUT_CELLS {
         return Err(bad(CMD, "the report is too large"));
     }
     let d = s.doc()?;
@@ -1914,6 +1926,28 @@ mod tests {
         x
     }
 
+    #[test]
+    fn a_report_stops_growing_past_the_output_limit() {
+        // Entries past the limit are not kept, and the report is refused.
+        let mut g = super::Grid::default();
+        g.num(1999, 1999, 1.0);
+        assert!(!g.too_large);
+        g.num(0, 2000, 1.0);
+        assert!(g.too_large);
+        assert_eq!(g.cells.len(), 1);
+        assert_eq!(g.extent(), (2000, 2000));
+        // Row-grouped statistics of a tall, almost empty column: one variable per row would make
+        // millions of report entries; the report is refused instead of built.
+        let mut s = s();
+        s.execute("range.setValues", json!({"range": "A1", "values": [[1]]})).unwrap();
+        s.execute("range.setValues", json!({"range": "A1000000", "values": [[2]]})).unwrap();
+        let err = s
+            .execute("data.analysisToolPak", json!({"tool": "descriptive", "range": "A1:A1000000", "groupedBy": "rows", "output": "C1"}))
+            .expect_err("report too large");
+        assert!(err.to_string().contains("too large"), "{err}");
+        assert_eq!(v(&s, "C1"), Value::Empty);
+    }
+
     #[test]
     fn forecast_standard_errors_match_the_full_window_rescan() {
         // #1043: the running sum must give what re-adding every window gave, gaps included.

crates/engine/src/cmd/print/tests.rs

@@ -695,79 +695,6 @@ fn bands_unit() {
     assert!(w.len() > 2);
 }
 
-/// The wrapper `wrap_lines` replaced, re-measuring strings, as the reference for its lines.
-fn wrap_lines_remeasuring(text: &str, font: Font, size: f32, width: f32) -> Vec<String> {
-    let mut out = Vec::new();
-    for para in text.split('\n') {
-        let mut line = String::new();
-        for word in para.split(' ') {
-            let cand = if line.is_empty() { word.to_string() } else { format!("{line} {word}") };
-            if line.is_empty() || gridcraft_pdf::text_width(font, size, &cand) <= width {
-                line = cand;
-            } else {
-                out.push(std::mem::take(&mut line));
-                line = word.to_string();
-            }
-            while gridcraft_pdf::text_width(font, size, &line) > width && line.chars().count() > 1 {
-                let chars: Vec<char> = line.chars().collect();
-                let (mut lo, mut hi) = (1usize, chars.len());
-                while lo < hi {
-                    let mid = lo + (hi - lo).div_ceil(2);
-                    let cut: String = chars[..mid].iter().collect();
-                    if gridcraft_pdf::text_width(font, size, &cut) <= width {
-                        lo = mid;
-                    } else {
-                        hi = mid - 1;
-                    }
-                }
-                if lo == chars.len() {
-                    break;
-                }
-                out.push(chars[..lo].iter().collect());
-                line = chars[lo..].iter().collect();
-            }
-            if out.len() > 2000 {
-                return out;
-            }
-        }
-        out.push(line);
-    }
-    out
-}
-
-#[test]
-fn wrap_lines_matches_the_remeasuring_wrapper() {
-    let texts = [
-        "the quick brown fox jumps",
-        "Supercalifragilistic",
-        "a  b   c\n\nverylongwordthatneedsbreaking short\nend ",
-        "Wide WWWWWWWWWW words, iiiiiiii narrow ones; mixed MMMMiiiiMMMM",
-        "€ ü ß — unicode Ünïcödé text",
-        "",
-        " ",
-    ];
-    for text in texts {
-        for font in [Font::Helvetica, Font::HelveticaBold, Font::Courier] {
-            for (size, width) in [(10.0, 50.0), (10.0, 20.0), (11.0, 1.0), (8.0, 300.0), (12.0, 0.0)] {
-                assert_eq!(wrap_lines(text, font, size, width), wrap_lines_remeasuring(text, font, size, width), "{text:?} {font:?} {size} {width}");
-            }
-        }
-    }
-}
-
-#[test]
-fn wrapping_a_long_word_in_a_narrow_box_stops_at_the_line_limit() {
-    // One character per line: the old wrapper re-measured and copied the rest of the word for
-    // each of the 100,000 lines before checking the limit.
-    let word = "w".repeat(100_000);
-    let lines = wrap_lines(&word, Font::Helvetica, 10.0, 1.0);
-    assert!(lines.len() <= 2002, "{}", lines.len());
-    assert!(lines.iter().all(|l| l == "w"));
-    // A long paragraph in a very wide box is one line.
-    let para = vec!["word"; 50_000].join(" ");
-    assert_eq!(wrap_lines(&para, Font::Helvetica, 10.0, 1.0e9), vec![para.clone()]);
-}
-
 #[test]
 fn center_across_spans_following_empty_center_across_cells() {
     let mut styles = StyleTable::default();

crates/engine/src/cmd/print/text.rs

@@ -3,57 +3,42 @@ use super::{CellRef, Color, Font, HAlign, MAX_COLS, Page, Point, RangeRef, Sheet
 
 const PAD_PX: f32 = 2.0;
 
-/// Most lines [`wrap_lines`] returns (it may return one more, the rest of a paragraph).
-const MAX_WRAP_LINES: usize = 2000;
-
 /// Word-wraps `text` to `width` points.
-///
-/// Widths are kept as running sums of glyph widths (the sum `text_width` makes), so each
-/// character is measured once: re-measuring the line for every word, or the rest of an
-/// over-long word for every piece cut from it, made a long paragraph or word quadratic.
 pub(super) fn wrap_lines(text: &str, font: Font, size: f32, width: f32) -> Vec<String> {
-    let units_of = |s: &str| -> u64 { s.chars().map(|c| u64::from(font.char_width(c))).sum() };
-    let fits = |units: u64| units as f32 * size / 1000.0 <= width;
-    let space = units_of(" ");
     let mut out = Vec::new();
     for para in text.split('\n') {
         let mut line = String::new();
-        let mut line_units: u64 = 0;
         for word in para.split(' ') {
-            let word_units = units_of(word);
-            if line.is_empty() {
-                line.push_str(word);
-                line_units = word_units;
-            } else if fits(line_units + space + word_units) {
-                line.push(' ');
-                line.push_str(word);
-                line_units += space + word_units;
+            let cand = if line.is_empty() { word.to_string() } else { format!("{line} {word}") };
+            if line.is_empty() || gridcraft_pdf::text_width(font, size, &cand) <= width {
+                line = cand;
             } else {
                 out.push(std::mem::take(&mut line));
-                line.push_str(word);
-                line_units = word_units;
+                line = word.to_string();
             }
-            // Break a single over-long word by characters: each piece is the longest that fits
-            // (at least one character).
-            if !fits(line_units) && line.chars().nth(1).is_some() {
-                let mut piece = String::new();
-                let mut units: u64 = 0;
-                for c in line.chars() {
-                    let w = u64::from(font.char_width(c));
-                    if !piece.is_empty() && !fits(units + w) {
-                        out.push(std::mem::take(&mut piece));
-                        units = 0;
-                        if out.len() > MAX_WRAP_LINES {
-                            return out;
-                        }
+            // Break a single over-long word by characters. The cut that fits is found by binary
+            // search (each try measures one candidate prefix), so an over-long word costs
+            // O(n log n) measurements instead of re-measuring the whole cut per character for
+            // every emitted line (`a` × 100k pasted text would be cubic worst case).
+            while gridcraft_pdf::text_width(font, size, &line) > width && line.chars().count() > 1 {
+                let chars: Vec<char> = line.chars().collect();
+                let (mut lo, mut hi) = (1usize, chars.len());
+                while lo < hi {
+                    let mid = lo + (hi - lo).div_ceil(2);
+                    let cut: String = chars[..mid].iter().collect();
+                    if gridcraft_pdf::text_width(font, size, &cut) <= width {
+                        lo = mid;
+                    } else {
+                        hi = mid - 1;
                     }
-                    piece.push(c);
-                    units += w;
                 }
-                line = piece;
-                line_units = units;
+                if lo == chars.len() {
+                    break;
+                }
+                out.push(chars[..lo].iter().collect());
+                line = chars[lo..].iter().collect();
             }
-            if out.len() > MAX_WRAP_LINES {
+            if out.len() > 2000 {
                 return out;
             }
         }