Fix salary tool edge cases (#75)

* fix: handle salary tool edge cases

* fix: preserve Danish salary compounds

salary_lookup.py

@@ -204,12 +204,15 @@ def format_entry(entry, metadata):
             count = data.get("count")
             index = data.get("index")
             if count is not None or index is not None:
-                count_str = str(count) if count else "-"
-                if index is not None:
+                count_str = str(count) if count is not None else "-"
+                if isinstance(index, (int, float)):
                     diff = index - baseline
                     sign = "+" if diff >= 0 else ""
                     index_str = f"{index:.1f}"
                     diff_str = f"{sign}{diff:.1f}%"
+                elif index is not None:
+                    index_str = str(index)
+                    diff_str = ""
                 else:
                     index_str = "N/A*"
                     diff_str = ""

tests/test_convert_salary_excel.py

@@ -37,6 +37,13 @@ class DetectColumnTypeTests(unittest.TestCase):
             with self.subTest(header=header):
                 self.assertEqual(detect_column_type(header), "count")
 
+    def test_count_inside_word_does_not_make_count_header(self):
+        self.assertIsNone(detect_column_type("Accounting Total"))
+        self.assertEqual(detect_column_type("Accounting Index"), "index")
+
+    def test_danish_compound_headers_still_match(self):
+        self.assertEqual(detect_column_type("Lønindeks"), "index")
+
     def test_parse_sheet_preserves_category_name_with_letter_n(self):
         ws = FakeWorksheet([
             ("Company", "Engineering Count", "Engineering Index"),
@@ -47,6 +54,16 @@ class DetectColumnTypeTests(unittest.TestCase):
 
         self.assertEqual(companies[0]["categories"]["engineering"], {"count": 12, "index": 105.5})
 
+    def test_parse_sheet_groups_accounting_count_index_pair(self):
+        ws = FakeWorksheet([
+            ("Company", "Accounting Count", "Accounting Index"),
+            ("Example Corp", 12, 105.5),
+        ])
+
+        companies = parse_sheet(ws)
+
+        self.assertEqual(companies[0]["categories"]["accounting"], {"count": 12, "index": 105.5})
+
 
 if __name__ == "__main__":
     unittest.main()

tests/test_salary_lookup.py

@@ -0,0 +1,41 @@
+import unittest
+
+from salary_lookup import format_entry
+
+
+class FormatEntryTests(unittest.TestCase):
+    def test_zero_count_is_displayed_as_zero(self):
+        entry = {
+            "company": "Example Corp",
+            "city": "",
+            "categories": {
+                "public_data": {
+                    "count": 0,
+                    "index": 100.0,
+                },
+            },
+        }
+
+        rendered = format_entry(entry, {"index_baseline": 100, "index_label": "Index"})
+
+        self.assertRegex(rendered, r"Public Data\s+0\s+100\.0")
+
+    def test_text_index_does_not_crash(self):
+        entry = {
+            "company": "Example Corp",
+            "city": "",
+            "categories": {
+                "sample": {
+                    "count": 3,
+                    "index": "private",
+                },
+            },
+        }
+
+        rendered = format_entry(entry, {"index_baseline": 100, "index_label": "Index"})
+
+        self.assertIn("private", rendered)
+
+
+if __name__ == "__main__":
+    unittest.main()

tools/convert_salary_excel.py

@@ -42,6 +42,7 @@ COMPANY_PATTERNS = {"firma", "company", "virksomhed", "employer", "arbejdsgiver"
 CITY_PATTERNS = {"by", "city", "kommune", "location", "lokation", "sted"}
 COUNT_PATTERNS = {"antal", "count", "number", "n", "employees", "medarbejdere"}
 INDEX_PATTERNS = {"indeks", "index", "idx", "salary", "løn", "median", "average", "gennemsnit"}
+DANISH_COMPOUND_PATTERNS = {"antal", "indeks", "løn", "gennemsnit", "medarbejdere"}
 
 
 def header_matches(header, patterns):
@@ -50,10 +51,9 @@ def header_matches(header, patterns):
     tokens = set(re.findall(r"[a-zæøåöäü0-9]+", h))
 
     for p in patterns:
-        if len(p) == 1:
-            if p in tokens:
-                return True
-        elif p in h:
+        if p in tokens:
+            return True
+        if p in DANISH_COMPOUND_PATTERNS and p in h:
             return True
     return False
 
@@ -62,10 +62,7 @@ def strip_type_patterns(header, patterns):
     """Remove count/index words from a header to derive a category name."""
     name = header.lower()
     for p in patterns:
-        if len(p) == 1:
-            name = re.sub(rf"\b{re.escape(p)}\b", "", name)
-        else:
-            name = name.replace(p, "")
+        name = re.sub(rf"(?<![a-zæøåöäü0-9]){re.escape(p)}(?![a-zæøåöäü0-9])", "", name)
     return name.strip(" _-")