#3166: a web page opens after a leading comment and keeps its non-breaking spaces

Addresses CodeRabbit review on #3223:
- "Skip leading HTML comments before the prefix check": a page starting
  with an exporter's <!-- comment --> was read as CSV text. Complete
  leading comments are now skipped before looking for the page's start.
- "Preserve nonbreaking spaces in HTML text imports": split_whitespace
  turned New&nbsp;&nbsp;York into "New York". The decoded cell text is
  used as is (ordinary whitespace is already collapsed), so the
  non-breaking spaces stay, as Excel keeps them.

Test: a_web_page_table_opens_after_a_leading_comment_and_keeps_nonbreaking_spaces.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

crates/engine/src/io.rs

@@ -357,7 +357,15 @@ fn part_too_large_message(name: &str, part: &str, size: u64) -> String {
 fn read_html_table(name: &str, kind: Option<FileKind>, bytes: &[u8], csv_as_text: bool) -> Option<Result<(Workbook, Vec<String>)>> {
     // Decoded as a CSV is (BOM, UTF-8, else Windows-1252/1256), so a legacy page keeps its letters.
     let text = gridcraft_xlsx::decode_text(bytes);
-    let head: String = text.trim_start_matches('\u{feff}').trim_start().chars().take(16).collect::<String>().to_ascii_lowercase();
+    // Comments an exporter writes before the page (`<!-- ... -->`) don't make it something else.
+    let mut rest = text.trim_start_matches('\u{feff}').trim_start();
+    while let Some(after) = rest.strip_prefix("<!--") {
+        match after.find("-->").and_then(|end| after.get(end + 3..)) {
+            Some(more) => rest = more.trim_start(),
+            None => break,
+        }
+    }
+    let head: String = rest.chars().take(16).collect::<String>().to_ascii_lowercase();
     if !["<!doctype html", "<html", "<table", "<head", "<body", "<meta"].iter().any(|t| head.starts_with(t)) {
         return None;
     }
@@ -375,7 +383,8 @@ fn read_html_table(name: &str, kind: Option<FileKind>, bytes: &[u8], csv_as_text
                 // cell keeps the text it shows.
                 let field = match cell.num.filter(|_| !csv_as_text) {
                     Some(n) => gridcraft_core::number_to_text(n),
-                    None => cell.text.split_whitespace().collect::<Vec<_>>().join(" "),
+                    // Whitespace is already collapsed; non-breaking spaces stay, as in Excel.
+                    None => cell.text.clone(),
                 };
                 if field.contains(['\t', '\n', '\r', '"']) { format!("\"{}\"", field.replace('"', "\"\"")) } else { field }
             })

crates/engine/src/io/tests.rs

@@ -656,3 +656,15 @@ fn a_web_page_table_keeps_its_characters_text_and_header_formats() {
     let (wb, _) = open_bytes_with("menu.xls", &page, None, true).unwrap();
     assert_eq!(wb.sheet(0).unwrap().value(CellRef::parse("B2").unwrap()), gridcraft_core::Value::text("4,50"));
 }
+
+#[test]
+fn a_web_page_table_opens_after_a_leading_comment_and_keeps_nonbreaking_spaces() {
+    // An exporter's comment before <html> still makes a web page; `&nbsp;` stays a non-breaking
+    // space, as Excel keeps it, while ordinary whitespace collapses.
+    let page = "<!-- exported by a web app -->\n<html><body><table><tr><td>New&nbsp;&nbsp;York</td><td>a \n b</td></tr></table></body></html>";
+    let (wb, warnings) = open_bytes("cities.xls", page.as_bytes()).unwrap();
+    assert!(warnings.first().is_some_and(|w| w.contains("is a web page (HTML)")), "{warnings:?}");
+    let sh = wb.sheet(0).unwrap();
+    assert_eq!(sh.value(CellRef::parse("A1").unwrap()), gridcraft_core::Value::text("New\u{a0}\u{a0}York"));
+    assert_eq!(sh.value(CellRef::parse("B1").unwrap()), gridcraft_core::Value::text("a b"));
+}