Merge PR #1115: Cap CSV import fields per record at the worksheet column limit (#1037)

crates/xlsx/src/csv.rs

@@ -87,8 +87,9 @@ pub fn sniff_delimiter(text: &str, quote: char) -> u8 {
     best.0
 }
 
-/// Splits CSV text into records of fields.
-fn parse_records(text: &str, delim: char, quote: char, mut on_record: impl FnMut(Vec<String>) -> bool) {
+/// Splits CSV text into records of fields. A record keeps at most [`MAX_COLS`] fields; the rest are
+/// dropped as they are read, so a row of endless delimiters cannot allocate a string per field.
+pub(crate) fn parse_records(text: &str, delim: char, quote: char, mut on_record: impl FnMut(Vec<String>) -> bool) {
     let mut field = String::new();
     let mut record: Vec<String> = vec![];
     let mut in_q = false;
@@ -112,13 +113,19 @@ fn parse_records(text: &str, delim: char, quote: char, mut on_record: impl FnMut
             in_q = true;
             field_started_quoted = true;
         } else if c == delim {
-            record.push(std::mem::take(&mut field));
+            let f = std::mem::take(&mut field);
+            if record.len() < MAX_COLS as usize {
+                record.push(f);
+            }
             field_started_quoted = false;
         } else if c == '\r' || c == '\n' {
             if c == '\r' && chars.peek() == Some(&'\n') {
                 chars.next();
             }
-            record.push(std::mem::take(&mut field));
+            let f = std::mem::take(&mut field);
+            if record.len() < MAX_COLS as usize {
+                record.push(f);
+            }
             field_started_quoted = false;
             if !on_record(std::mem::take(&mut record)) {
                 return;
@@ -128,7 +135,9 @@ fn parse_records(text: &str, delim: char, quote: char, mut on_record: impl FnMut
         }
     }
     if !field.is_empty() || !record.is_empty() || field_started_quoted {
-        record.push(field);
+        if record.len() < MAX_COLS as usize {
+            record.push(field);
+        }
         on_record(record);
     }
 }

crates/xlsx/src/tests/csv_tests.rs

@@ -1,7 +1,7 @@
-use gridcraft_core::{CellRef, DateSystem, Value};
+use gridcraft_core::{CellRef, DateSystem, MAX_COLS, Value};
 use gridcraft_model::{Sheet, Workbook};
 
-use crate::csv::{decode_text, sniff_delimiter};
+use crate::csv::{decode_text, parse_records, sniff_delimiter};
 use crate::{CsvOptions, Format, IoError, read_csv, sniff, write_csv, write_csv_limited};
 
 fn v(wb: &Workbook, r: u32, c: u32) -> Value {
@@ -209,3 +209,24 @@ fn the_cap_counts_the_bytes_the_cells_hold_and_accepted_exports_keep_every_posit
     let tabs = write_csv_limited(&s, b'\t', 14, text_of(&s)).unwrap();
     assert_eq!(tabs.len(), 14);
 }
+
+#[test]
+fn parser_stops_collecting_fields_at_max_cols() {
+    let extra = 50;
+    // MAX_COLS + extra fields on the first row (the last one filled), then a short second row.
+    let mut text = ",".repeat((MAX_COLS + extra) as usize - 1);
+    text.push_str("end\nx,y\n");
+    let mut lens = vec![];
+    parse_records(&text, ',', '"', |rec| {
+        lens.push(rec.len());
+        true
+    });
+    assert_eq!(lens, vec![MAX_COLS as usize, 2]);
+    // An unterminated last row is capped the same way.
+    let mut lens = vec![];
+    parse_records(&",".repeat(MAX_COLS as usize + extra as usize), ',', '"', |rec| {
+        lens.push(rec.len());
+        true
+    });
+    assert_eq!(lens, vec![MAX_COLS as usize]);
+}