fix(persistence): CSV export quoting, number errors, formula suffix, parse locations

- export_csv writes through csv::Writer (RFC 4180) and honors the view's
  number format (improvise-1cz)
- malformed numbers yield CellValue::Error naming the text instead of
  silent 0.0 (improvise-4yc)
- formula [Category] suffix moved into the pest grammar; walker extracts
  it structurally instead of rfind(" [") string-scanning (improvise-zgb)
- parse errors now report line, column, and the offending line's text
  (improvise-6kj)

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

crates/improvise-io/src/persistence/improv.pest

@@ -58,10 +58,17 @@ formulas_section = {
     formula_line*
 }
 
+// `- raw expression [Category]` — the trailing `[Category]` suffix is
+// optional (defaults to _Measure) and its name follows the usual quoting
+// rules. `formula_raw` stops where a suffix-then-newline (or bare newline)
+// begins, so pipe-quoted identifiers containing " [" stay in the raw text.
 formula_line = {
-    "- " ~ rest_of_line ~ NEWLINE ~ blank_lines
+    "- " ~ formula_raw ~ formula_target? ~ NEWLINE ~ blank_lines
 }
 
+formula_target = _{ " "* ~ "[" ~ name ~ "]" }
+formula_raw    = @{ (!(formula_target? ~ NEWLINE) ~ ANY)+ }
+
 // ── Data ─────────────────────────────────────────────────────────────────────
 
 data_section = {

crates/improvise-io/src/persistence/mod.rs

@@ -224,7 +224,7 @@ pub fn format_md(workbook: &Workbook) -> String {
             if f.target_category == "_Measure" {
                 w!(out, "- {}", f.raw);
             } else {
-                w!(out, "- {} [{}]", f.raw, f.target_category);
+                w!(out, "- {} [{}]", f.raw, quote_name(&f.target_category));
             }
         }
     }
@@ -302,7 +302,17 @@ pub fn parse_md(text: &str) -> Result<Workbook> {
     use pest::iterators::{Pair, Pairs};
 
     let file = ImprovParser::parse(Rule::file, text)
-        .map_err(|e| anyhow::anyhow!("Parse error: {e}"))?
+        .map_err(|e| {
+            use pest::error::LineColLocation;
+            let (line, col) = match e.line_col {
+                LineColLocation::Pos(p) | LineColLocation::Span(p, _) => p,
+            };
+            let line_text = text.lines().nth(line - 1).unwrap_or("").trim_end();
+            anyhow::anyhow!(
+                "Parse error at line {line}, column {col}: {line_text}\n  ({})",
+                e.variant.message()
+            )
+        })?
         .next()
         .ok_or_else(|| anyhow::anyhow!("Empty parse result"))?;
 
@@ -410,19 +420,18 @@ pub fn parse_md(text: &str) -> Result<Workbook> {
             Rule::formulas_section => {
                 for fl in pair.into_inner() {
                     if fl.as_rule() == Rule::formula_line {
-                        let raw = first_str(fl)?;
-                        if let Some(i) = raw.rfind(" [")
-                            && raw.ends_with(']')
-                        {
-                            formulas.push((
-                                raw[..i].to_string(),
-                                raw[i + 2..raw.len() - 1].to_string(),
-                            ));
-                            continue;
-                        }
+                        let mut parts = fl.into_inner();
+                        let raw = next(&mut parts, "formula_line")?
+                            .as_str()
+                            .trim()
+                            .to_string();
                         // No [Category] suffix — default to _Measure
-                        if !raw.is_empty() && raw.contains('=') {
-                            formulas.push((raw, "_Measure".to_string()));
+                        let category = match parts.next() {
+                            Some(name_pair) => extract_name(name_pair)?,
+                            None => "_Measure".to_string(),
+                        };
+                        if raw.contains('=') {
+                            formulas.push((raw, category));
                         }
                     }
                 }
@@ -904,6 +913,59 @@ mod tests {
         assert_eq!(f.target_category, "Type");
     }
 
+    /// Bug improvise-zgb (b): the formulas walker located the `[Category]`
+    /// suffix via `raw.rfind(" [")`, so a hand-written line with no space
+    /// before the bracket (`- Total = Food + Gas[Type]`) was not recognized
+    /// and the formula was silently retargeted to _Measure. The suffix is now
+    /// part of the grammar (`formula_line`) and walked structurally.
+    #[test]
+    fn parse_md_formula_category_suffix_without_space() {
+        let text = "v2025-04-09\n# T\n\
+                    ## Category: Type\n\
+                    - Food, Gas, Total\n\
+                    ## Formulas\n\
+                    - Total = Food + Gas[Type]\n";
+        let m = parse_md(text).unwrap();
+        let f = &m.model.formulas()[0];
+        assert_eq!(f.raw, "Total = Food + Gas");
+        assert_eq!(f.target_category, "Type");
+    }
+
+    /// Bug improvise-zgb (a): `rfind(" [")` mis-split a suffix whose category
+    /// name itself contains " [" — the writer emitted it unquoted
+    /// (`[Margin [%]]`) and the parser then split inside the name, producing
+    /// garbage raw text and category "%]". Category names in the suffix are
+    /// now pipe-quoted on write and parsed via the grammar's `name` rule.
+    #[test]
+    fn parse_md_round_trips_formula_category_with_special_chars() {
+        let mut m = two_cat_model();
+        m.add_category("Margin [%]").unwrap();
+        m.model
+            .category_mut("Margin [%]")
+            .unwrap()
+            .add_item("Total");
+        m.model
+            .add_formula(parse_formula("Total = Food + Gas", "Margin [%]").unwrap());
+        let m2 = parse_md(&format_md(&m)).unwrap();
+        let f = &m2.model.formulas()[0];
+        assert_eq!(f.raw, "Total = Food + Gas");
+        assert_eq!(f.target_category, "Margin [%]");
+    }
+
+    /// Companion to improvise-zgb: a formula *ending* in a pipe-quoted
+    /// identifier that contains " [" must not have the quoted name mistaken
+    /// for a category suffix — it stays a _Measure formula with intact raw.
+    #[test]
+    fn parse_md_formula_ending_in_bracketed_pipe_name_keeps_raw() {
+        let text = "v2025-04-09\n# T\n\
+                    ## Formulas\n\
+                    - Total = Cost + |Revenue [USD]|\n";
+        let m = parse_md(text).unwrap();
+        let f = &m.model.formulas()[0];
+        assert_eq!(f.raw, "Total = Cost + |Revenue [USD]|");
+        assert_eq!(f.target_category, "_Measure");
+    }
+
     #[test]
     fn parse_md_round_trips_hidden_item() {
         let _ = two_cat_model();
@@ -2071,6 +2133,21 @@ mod parser_edge_cases {
         );
     }
 
+    /// Bug improvise-6kj: parse failures were wrapped as `Parse error: {e}`,
+    /// leaning on pest's raw rendering instead of stating the location in
+    /// plain words. The message must name the line number and quote the
+    /// offending line so a user can find the problem in a hand-edited file.
+    #[test]
+    fn parse_error_names_line_number_and_contents() {
+        let text = "v2025-04-09\n# T\n## Data\nType=Food = 3.14.15\n";
+        let err = parse_md(text).unwrap_err().to_string();
+        assert!(err.contains("line 4"), "missing line number: {err}");
+        assert!(
+            err.contains("Type=Food = 3.14.15"),
+            "missing line contents: {err}"
+        );
+    }
+
     #[test]
     fn parse_data_without_value() {
         // Malformed data line: no " = " separator — pest rejects it