address review: record a walked directory only within the depth limit; stats skips symlinked sources and bounds the read

apps/gridcraft-cli/src/check.rs

@@ -278,18 +278,23 @@ fn collect_inputs(inputs: &[String]) -> Discovery {
 }
 
 /// Returns false only when the global file budget stops traversal.
-fn walk(dir: &Path, depth: usize, out: &mut Discovery, seen: &mut std::collections::HashSet<u64>) -> bool {
+fn walk(dir: &Path, depth: usize, out: &mut Discovery, seen: &mut std::collections::HashSet<(u64, u64)>) -> bool {
     // A directory is visited once by identity (dev+inode on unix): several links to the same
     // directory are aliases, not exponential branches of the scan. Other platforms keep the
     // depth cap only.
     #[cfg(unix)]
     {
         use std::os::unix::fs::MetadataExt as _;
-        if let Ok(m) = std::fs::metadata(dir)
-            && let Some(id) = m.dev().checked_mul(1 << 60).and_then(|d| d.checked_add(m.ino()))
-            && !seen.insert(id)
-        {
-            return true;
+        if let Ok(m) = std::fs::metadata(dir) {
+            let identity = (m.dev(), m.ino());
+            // Already scanned: an alias. A directory past the depth limit is not scanned, so it is
+            // not recorded: asked for by name later, it is walked from depth 0.
+            if seen.contains(&identity) {
+                return true;
+            }
+            if depth <= MAX_DEPTH {
+                seen.insert(identity);
+            }
         }
     }
     if depth > MAX_DEPTH {

apps/gridcraft-cli/tests/it/cli.rs

@@ -205,6 +205,24 @@ fn check_reports_a_depth_limited_scan_while_checking_readable_files() {
     let _ = std::fs::remove_dir_all(dir);
 }
 
+#[cfg(unix)]
+#[test]
+fn a_directory_past_the_depth_limit_is_still_scanned_when_named_as_an_input() {
+    let dir = tmpdir("check-depth-named");
+    // The first folder past the limit (33 levels down): the scan from the top reports it unreached.
+    let mut deep = dir.clone();
+    for _ in 0..33 {
+        deep.push("d");
+    }
+    std::fs::create_dir_all(&deep).unwrap();
+    ok(&["run", "--out", deep.join("deep.xlsx").to_str().unwrap(), "--quiet"]);
+    let out = run(&["check", dir.to_str().unwrap(), deep.to_str().unwrap(), "--json"]);
+    let report: serde_json::Value = serde_json::from_slice(&out.stdout).unwrap();
+    // The first scan reports the subtree it could not reach; the named directory is checked.
+    assert_eq!(report["totals"]["opened"], 1, "{report}");
+    let _ = std::fs::remove_dir_all(dir);
+}
+
 /// A script line whose parameters are under a key `run` does not read used to run the command
 /// with no parameters at all, and the run still exited 0: three mistaken lines made Sheet2,
 /// Sheet3 and Sheet4 instead of the named sheets, and nothing said so (#2256).

crates/engine/src/pivot/charts.rs

@@ -57,10 +57,6 @@ fn rows_ref(sheet: &str, origin: CellRef, col: u32, rows: &[u32]) -> Option<Stri
 /// Most category rows a PivotChart plots.
 const MAX_CHART_ROWS: usize = 10_000;
 
-/// Most series a PivotChart gets: the chart renderer reads at most 255 series, and each one owns
-/// reference text as long as its category rows, copied into every linked chart.
-const MAX_CHART_SERIES: usize = 255;
-
 /// A PivotChart's series for a report written at `origin`: categories are the innermost row
 /// items, one series per item column (column items × value fields); totals are left out.
 pub fn chart_series(sheet: &str, origin: CellRef, out: &Output) -> Vec<Series> {
@@ -68,7 +64,6 @@ pub fn chart_series(sheet: &str, origin: CellRef, out: &Output) -> Vec<Series> {
     let categories = out.label_col.and_then(|c| rows_ref(sheet, origin, c, &rows));
     out.chart_cols
         .iter()
-        .take(MAX_CHART_SERIES)
         .filter_map(|(c, name)| {
             Some(Series {
                 name: Some(format!("\"{}\"", name.replace('"', "\"\""))),
@@ -111,22 +106,3 @@ pub fn clear_output(wb: &mut Workbook, sheet: usize, idx: usize) {
         p.last_range = None;
     }
 }
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn a_pivot_chart_gets_at_most_the_series_the_renderer_reads() {
-        let out = Output {
-            grid: vec![],
-            row_kinds: vec![],
-            body_offset: 0,
-            width: 1000,
-            chart_rows: vec![1, 2, 3],
-            chart_cols: (1..1000).map(|c| (c, format!("S{c}"))).collect(),
-            label_col: Some(0),
-        };
-        assert_eq!(chart_series("Sheet1", CellRef::new(0, 0), &out).len(), MAX_CHART_SERIES);
-    }
-}

crates/engine/src/pivot/compute.rs

@@ -1,7 +1,5 @@
 use super::items::{Acc, FieldItems, build_tree, collect_siblings, is_collapsed, item_label};
-use super::output::{
-    AggMap, CellKind, ColKind, ColSpec, Ctx, Grid, Line, LineKind, Output, RowKind, ValSrc, caption, data_cols_exceed_sheet, flatten_cols,
-};
+use super::output::{AggMap, CellKind, ColKind, ColSpec, Ctx, Grid, Line, LineKind, Output, RowKind, ValSrc, data_cols_exceed_sheet, flatten_cols};
 use super::{
     CalcSet, CellError, CellRef, HashMap, HashSet, MAX_AXIS_FIELDS, MAX_COLS, MAX_VALUE_FIELDS, PivotDateGroup, PivotField, PivotFilter, PivotFunc,
     PivotLayout, PivotShowAs, PivotTable, PivotValue, Source, Value, Workbook, read_source,
@@ -94,10 +92,6 @@ struct AggregateData {
     col_leaves: HashSet<Vec<u32>>,
 }
 
-/// Most accumulators (groups times value columns) one report may hold: twice what the 5 million
-/// group limit allowed at one value field.
-const MAX_ACCUMULATORS: usize = 16_000_000;
-
 fn aggregate_record(
     input: &AggregateInput<'_>,
     data: &mut AggregateData,
@@ -156,9 +150,7 @@ fn aggregate_record(
     if !input.cols.is_empty() && data_cols_exceed_sheet(data.col_leaves.len(), input.pt.grand_totals_rows, input.vals.len()) {
         return Err("The PivotTable report will not fit on the sheet.".into());
     }
-    // Each entry holds one accumulator per value field and calculated-field source column, so the
-    // width counts too: a handful of groups over hundreds of fields is large as well.
-    if data.aggs.len() > 5_000_000 || data.aggs.len().saturating_mul(input.vals.len() + input.extra.len()) > MAX_ACCUMULATORS {
+    if data.aggs.len() > 5_000_000 {
         return Err("The PivotTable report is too large.".into());
     }
     Ok(())
@@ -265,8 +257,8 @@ fn write_headers(g: &mut Grid, input: HeaderInput<'_>) -> Result<(), String> {
                     let lv = spec.path.len().saturating_sub(1);
                     let label = spec.path.last().and_then(|it| col_items.get(lv).map(|fi| fi.label(*it))).unwrap_or_default();
                     let text = match vi {
-                        Some(vi) if k > 1 => caption(&label, &val_name(*vi)),
-                        _ => caption(&label, "Total"),
+                        Some(vi) if k > 1 => format!("{label} {}", val_name(*vi)),
+                        _ => format!("{label} Total"),
                     };
                     g.set(first_level + lv, c, Value::text(text), CellKind::ColLabel, 0, None);
                 }
@@ -324,7 +316,6 @@ struct ShowContext<'a> {
     value_index: usize,
     level0: &'a [u32],
     siblings: &'a HashMap<Vec<u32>, Vec<u32>>,
-    cache: &'a ShowCache,
 }
 
 fn number(value: &Option<Value>) -> Option<f64> {
@@ -341,67 +332,42 @@ fn ratio(a: Value, b: Option<Value>) -> Value {
     }
 }
 
-/// What Running Total and Rank need from other cells, computed once per sibling set instead of
-/// once per displayed value (which rescanned the whole set every time).
-type ShowKey = (Vec<u32>, Vec<u32>, usize);
-/// Cumulative values by top-level item, and the total of all of them.
-type RunningTotals = (HashMap<u32, f64>, f64);
-
-#[derive(Default)]
-struct ShowCache {
-    /// Ranked values' sibling numbers, sorted ascending without repeats, by (parent, column, value).
-    ranks: std::cell::RefCell<HashMap<ShowKey, Vec<f64>>>,
-    /// Running totals by top-level item, with the total of all of them, by (rest of path, column, value).
-    running: std::cell::RefCell<HashMap<ShowKey, RunningTotals>>,
-}
-
 fn running_total(raw: Value, lookup: &ValueLookup<'_>, context: &ShowContext<'_>) -> Value {
-    let Some((first, rest)) = context.row_path.split_first() else { return raw };
+    let Some((first, _)) = context.row_path.split_first() else { return raw };
     if !matches!(raw, Value::Number(_)) {
         return raw;
     }
-    let mut cache = context.cache.running.borrow_mut();
-    let (by_item, all) = cache.entry((rest.to_vec(), context.col_path.to_vec(), context.value_index)).or_insert_with(|| {
-        let mut total = 0.0;
-        let mut by_item = HashMap::new();
-        for item in context.level0 {
-            let mut path = Vec::with_capacity(rest.len() + 1);
-            path.push(*item);
-            path.extend_from_slice(rest);
-            // An item with no number here adds nothing and is not stored (the shown cell always has one).
-            let Some(value) = number(&lookup.get(&path, context.col_path, context.value_index)) else { continue };
-            total += value;
-            by_item.entry(*item).or_insert(total);
+    let mut total = 0.0;
+    for item in context.level0 {
+        let mut path = context.row_path.to_vec();
+        if let Some(first_item) = path.first_mut() {
+            *first_item = *item;
         }
-        (by_item, total)
-    });
-    Value::number(by_item.get(first).copied().unwrap_or(*all))
+        total += number(&lookup.get(&path, context.col_path, context.value_index)).unwrap_or(0.0);
+        if item == first {
+            break;
+        }
+    }
+    Value::number(total)
 }
 
 fn ranked_value(raw: Value, lookup: &ValueLookup<'_>, context: &ShowContext<'_>) -> Value {
     let Some((_, parent)) = context.row_path.split_last() else { return Value::Empty };
     let Some(me) = raw.as_f64() else { return raw };
-    let mut cache = context.cache.ranks.borrow_mut();
-    let sorted = cache.entry((parent.to_vec(), context.col_path.to_vec(), context.value_index)).or_insert_with(|| {
-        let mut all: Vec<f64> = context
-            .siblings
-            .get(parent)
-            .into_iter()
-            .flatten()
-            .filter_map(|sibling| {
-                let mut path = parent.to_vec();
-                path.push(*sibling);
-                number(&lookup.get(&path, context.col_path, context.value_index))
-            })
-            .collect();
-        all.sort_by(f64::total_cmp);
-        all.dedup();
-        all
-    });
+    let siblings = context.siblings.get(parent).cloned().unwrap_or_default();
     // Excel's ranks are dense: ties share a rank and the next value takes the next one
     // (5, 4, 3, 3, 4 rank 1, 2, 3, 3, 2), so count the distinct larger values.
-    let larger = sorted.len() - sorted.partition_point(|value| *value <= me);
-    Value::number(larger as f64 + 1.0)
+    let mut larger: Vec<f64> = siblings
+        .into_iter()
+        .filter_map(|sibling| {
+            let mut path = parent.to_vec();
+            path.push(sibling);
+            number(&lookup.get(&path, context.col_path, context.value_index)).filter(|value| *value > me)
+        })
+        .collect();
+    larger.sort_by(f64::total_cmp);
+    larger.dedup();
+    Value::number(larger.len() as f64 + 1.0)
 }
 
 fn show_value(definition: &PivotValue, raw: Value, lookup: &ValueLookup<'_>, context: &ShowContext<'_>) -> Value {
@@ -671,7 +637,6 @@ pub(super) fn compute_from(wb: &Workbook, pt: &PivotTable, src: &Source, mut pro
         })
         .collect();
     let lookup = ValueLookup { wb, source: src, calc: &calc, vals: &vals, extra: &extra, aggs: &aggs, value_count: k };
-    let show_cache = ShowCache::default();
     let mut chart_rows: Vec<u32> = Vec::new();
     let chart_cols: Vec<(u32, String)> = data_cols
         .iter()
@@ -729,7 +694,7 @@ pub(super) fn compute_from(wb: &Workbook, pt: &PivotTable, src: &Source, mut pro
             let Some((_, vdef)) = vals.get(*vi) else { continue };
             let cp = spec.path.as_slice();
             let Some(raw) = lookup.get(rp, cp, *vi) else { continue };
-            let context = ShowContext { row_path: rp, col_path: cp, value_index: *vi, level0: &level0, siblings: &siblings, cache: &show_cache };
+            let context = ShowContext { row_path: rp, col_path: cp, value_index: *vi, level0: &level0, siblings: &siblings };
             let shown = show_value(vdef, raw, &lookup, &context);
             if let Some(pr) = probe.as_deref_mut() {
                 if pr.at == Some((r, l + j)) {

crates/engine/src/pivot/output.rs

@@ -121,18 +121,6 @@ pub(super) enum ColKind {
     Grand,
 }
 
-/// Longest item caption Excel keeps; a subtotal caption repeats its parent's label for every value
-/// field at every level, so a long source label is cut rather than multiplied.
-const MAX_CAPTION_LABEL_CHARS: usize = 255;
-
-/// `label` (cut to [`MAX_CAPTION_LABEL_CHARS`]) and `suffix`, as in "East Total".
-pub(super) fn caption(label: &str, suffix: &str) -> String {
-    let mut out: String = label.chars().take(MAX_CAPTION_LABEL_CHARS).collect();
-    out.push(' ');
-    out.push_str(suffix);
-    out
-}
-
 pub(super) struct ColSpec {
     pub(super) kind: ColKind,
     pub(super) path: Vec<u32>,
@@ -174,7 +162,7 @@ impl Ctx<'_> {
                 leaf: false,
                 path: prefix.to_vec(),
                 values: true,
-                labels: vec![(col, Value::text(caption(label, name)), indent, None)],
+                labels: vec![(col, Value::text(format!("{label} {name}")), indent, None)],
                 vi: Some(vi),
             });
         }
@@ -261,7 +249,7 @@ impl Ctx<'_> {
                         leaf: false,
                         path: prefix.clone(),
                         values: true,
-                        labels: vec![(col, Value::text(caption(&label, "Total")), indent, None)],
+                        labels: vec![(col, Value::text(format!("{label} Total")), indent, None)],
                         vi,
                     });
                 }
@@ -328,7 +316,7 @@ impl Ctx<'_> {
                         leaf: false,
                         path: prefix.clone(),
                         values: true,
-                        labels: vec![(col, Value::text(caption(&label, "Total")), 0, None)],
+                        labels: vec![(col, Value::text(format!("{label} Total")), 0, None)],
                         vi,
                     });
                 }
@@ -450,17 +438,3 @@ pub(super) fn cell_style(lk: &Look, rk: RowKind, cell: &OutCell, last_header: bo
     }
     s
 }
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn captions_cut_a_long_parent_label() {
-        assert_eq!(caption("East", "Total"), "East Total");
-        let long = "é".repeat(100_000);
-        let c = caption(&long, "Sum of Sales");
-        assert_eq!(c.chars().count(), MAX_CAPTION_LABEL_CHARS + " Sum of Sales".len());
-        assert!(c.ends_with(" Sum of Sales"));
-    }
-}

xtask/src/stats.rs

@@ -13,7 +13,10 @@ struct Counts {
     proptests: usize,
 }
 
-fn walk(dir: &Path, c: &mut Counts, seen: &mut std::collections::HashSet<u64>, budget: &mut usize) {
+/// Larger files (or devices behind a symlink) are not source files and are not read.
+const MAX_SOURCE_BYTES: u64 = 16 << 20;
+
+fn walk(dir: &Path, c: &mut Counts, seen: &mut std::collections::HashSet<(u64, u64)>, budget: &mut usize) {
     if *budget == 0 {
         // Hard stop for pathological trees: two directory symlinks pointing at each other
         // would otherwise be walked exponentially.
@@ -22,9 +25,9 @@ fn walk(dir: &Path, c: &mut Counts, seen: &mut std::collections::HashSet<u64>, b
     // Directory identities (dev+inode on unix) make a symlink to a visited directory a
     // no-op, not another exponential branch of the scan.
     #[cfg(unix)]
-    let id = std::fs::metadata(dir).ok().and_then(|m| m.dev().checked_mul(1 << 60).and_then(|d| d.checked_add(m.ino())));
+    let id = std::fs::metadata(dir).ok().map(|m| (m.dev(), m.ino()));
     #[cfg(not(unix))]
-    let id: Option<u64> = None;
+    let id: Option<(u64, u64)> = None;
     if let Some(id) = id
         && !seen.insert(id)
     {
@@ -45,7 +48,9 @@ fn walk(dir: &Path, c: &mut Counts, seen: &mut std::collections::HashSet<u64>, b
                 walk(&p, c, seen, budget);
             }
         } else if name.ends_with(".rs")
-            && let Ok(s) = std::fs::read_to_string(&p)
+            // The entry's own metadata does not follow a symlink: only regular files are sources.
+            && e.metadata().is_ok_and(|m| m.is_file() && m.len() <= MAX_SOURCE_BYTES)
+            && let Some(s) = read_source(&p)
         {
             c.files += 1;
             c.lines += s.lines().count();
@@ -56,6 +61,15 @@ fn walk(dir: &Path, c: &mut Counts, seen: &mut std::collections::HashSet<u64>, b
     }
 }
 
+/// A source file's text, read up to [`MAX_SOURCE_BYTES`]: one that grew or was replaced since its
+/// size was checked is skipped, not read whole.
+fn read_source(p: &Path) -> Option<String> {
+    use std::io::Read as _;
+    let mut s = String::new();
+    std::fs::File::open(p).ok()?.take(MAX_SOURCE_BYTES + 1).read_to_string(&mut s).ok()?;
+    (s.len() as u64 <= MAX_SOURCE_BYTES).then_some(s)
+}
+
 /// `#[test]` attributes and `proptest!` blocks (each block may hold many
 /// property tests, and macro-generated tests are only counted once — use
 /// `--exact` for harness-accurate numbers).
@@ -129,11 +143,25 @@ pub fn run(_root: &Path, exact: bool) -> Result<(), String> {
 
 #[cfg(test)]
 mod tests {
-    use super::count_tests;
+    use super::{MAX_SOURCE_BYTES, count_tests, read_source};
 
     #[test]
     fn counts_attributes_and_proptest_blocks() {
         let src = "#[test]\nfn a() {}\n// #[test] commented\n    #[test]\nfn b() {}\nproptest! {\n}\n";
         assert_eq!(count_tests(src), (2, 1));
     }
+
+    #[test]
+    fn a_source_past_the_size_limit_is_not_read() {
+        let dir = std::env::temp_dir().join(format!("xtask-stats-{}", std::process::id()));
+        std::fs::create_dir_all(&dir).unwrap();
+        let (small, big) = (dir.join("a.rs"), dir.join("b.rs"));
+        std::fs::write(&small, "fn a() {}\n").unwrap();
+        let file = std::fs::File::create(&big).unwrap();
+        file.set_len(MAX_SOURCE_BYTES + 1).unwrap(); // sparse: nothing is written
+        drop(file);
+        assert_eq!(read_source(&small).as_deref(), Some("fn a() {}\n"));
+        assert_eq!(read_source(&big), None);
+        let _ = std::fs::remove_dir_all(dir);
+    }
 }