fix(salary): skip whitespace-only workbook company rows (#576)

* fix(salary): skip whitespace-only workbook company rows

* docs(changelog): record fix for #576

CHANGELOG.md

@@ -232,6 +232,8 @@ per-file diff commands.
   before changing rank metadata; current-day deadlines, explicit future
   extensions, invalid dates, dry runs, and scoring errors keep their behavior.
 
+- **Salary workbook conversion skips whitespace-only companies** (#576) - blank-looking company cells no longer produce empty company entries.
+
 - **`/reset profile` keeps the `framework_version` frontmatter on the two files it rewrites**
   (`.claude/commands/reset.md` Step 3, `tests/test_reset_command.py`) - the full-file skeletons
   for `01-candidate-profile.md` and `02-behavioral-profile.md` started at the `#` title, while

tests/test_salary_empty_company_cli.py

@@ -0,0 +1,52 @@
+"""Workbook blank rows must not make the generated salary data invalid."""
+import json
+import shutil
+import subprocess
+import sys
+import unittest
+from pathlib import Path
+from tempfile import TemporaryDirectory
+
+try:
+    import openpyxl
+except ImportError:
+    openpyxl = None
+
+REPO = Path(__file__).resolve().parent.parent
+
+
+@unittest.skipUnless(openpyxl, "requires optional openpyxl (installed in CI)")
+class EmptyCompanyWorkbookCLI(unittest.TestCase):
+    def convert_and_validate(self, blank):
+        with TemporaryDirectory() as directory:
+            root = Path(directory)
+            workbook = root / "salary.xlsx"
+            output = root / "salary_data.json"
+            book = openpyxl.Workbook()
+            try:
+                book.active.append(["Company", "Count", "Index"])
+                book.active.append(["Acme Corp", 12, 105])
+                book.active.append([blank, 20, 110])
+                book.save(workbook)
+            finally:
+                book.close()
+            conversion = subprocess.run(
+                [sys.executable, str(REPO / "tools/convert_salary_excel.py"), str(workbook), "--output", str(output)],
+                capture_output=True, text=True, encoding="utf-8",
+            )
+            self.assertEqual(conversion.returncode, 0, conversion.stderr)
+            shutil.copyfile(REPO / "salary_lookup.py", root / "salary_lookup.py")
+            validation = subprocess.run(
+                [sys.executable, str(root / "salary_lookup.py"), "--validate"],
+                capture_output=True, text=True, encoding="utf-8",
+            )
+            self.assertEqual(validation.returncode, 0, validation.stdout + validation.stderr)
+            data = json.loads(output.read_text(encoding="utf-8"))
+            self.assertEqual(data["companies"], [{"company": "Acme Corp", "city": "", "categories": {
+                "all_employees": {"count": 12, "index": 105.0}}}])
+
+    def test_space_only_company_rows_are_skipped(self):
+        self.convert_and_validate(" \u00a0 ")
+
+    def test_truly_empty_company_rows_are_still_skipped(self):
+        self.convert_and_validate(None)

tools/convert_salary_excel.py

@@ -280,6 +280,8 @@ def parse_sheet(ws, sheet_label=None):
             continue
 
         company_name = str(row[company_col]).strip()
+        if not company_name:
+            continue
         if city_col is not None and city_col < len(row) and row[city_col]:
             city_name = str(row[city_col]).strip()
         else: