Fix leading space in surnames after capitalize() with empty middle name

capitalize() split each attribute with str.split(' '), which returns ['']
(not []) for an empty string. cap_piece() returns '' for an empty part, so an
empty middle name produced middle_list = [''], which leaked into surnames_list
(middle_list + last_list) and yielded a leading space in the surnames property:

    >>> hn = HumanName('john doe'); hn.capitalize(); hn.surnames
    ' Doe'   # leading space (should be 'Doe')

The same spurious '' element also appeared in title_list/first_list/last_list
for empty attributes. Using str.split() instead returns [] for empty strings
and is equivalent for the already-whitespace-collapsed pieces cap_piece()
returns. The suffix split (', ') is intentionally left unchanged.

Added a regression test in HumanNameCapitalizationTestCase.

Author: patchwright

nameparser/parser.py

@@ -980,10 +980,10 @@ def capitalize(self, force: bool | None = None) -> None:
 
         if not force and not (name == name.upper() or name == name.lower()):
             return
-        self.title_list = self.cap_piece(self.title, 'title').split(' ')
-        self.first_list = self.cap_piece(self.first, 'first').split(' ')
-        self.middle_list = self.cap_piece(self.middle, 'middle').split(' ')
-        self.last_list = self.cap_piece(self.last, 'last').split(' ')
+        self.title_list = self.cap_piece(self.title, 'title').split()
+        self.first_list = self.cap_piece(self.first, 'first').split()
+        self.middle_list = self.cap_piece(self.middle, 'middle').split()
+        self.last_list = self.cap_piece(self.last, 'last').split()
         self.suffix_list = self.cap_piece(self.suffix, 'suffix').split(', ')
 
     def handle_capitalization(self) -> None:

tests.py

@@ -2124,6 +2124,28 @@ def test_capitization_middle_initial_is_also_a_conjunction(self) -> None:
         hn.capitalize()
         self.m(str(hn), 'Scott E. Werner', hn)
 
+    def test_capitalize_empty_name_part_has_no_leading_space_in_surnames(self) -> None:
+        # capitalize() splits each attribute with str.split(' '), which returns
+        # [''] (rather than []) for an empty string. That spurious element
+        # leaked into surnames_list (middle_list + last_list) and produced a
+        # leading space in the surnames property, e.g. ' Doe' instead of 'Doe'.
+        hn = HumanName('john doe')
+        hn.capitalize()
+        self.m(hn.surnames, 'Doe', hn)
+        self.assertEqual(hn.middle_list, [])
+        self.assertEqual(hn.surnames_list, ['Doe'])
+
+        # force=True on a mixed-case name hits the same code path
+        hn = HumanName('Jane Doe')
+        hn.capitalize(force=True)
+        self.m(hn.surnames, 'Doe', hn)
+        self.assertEqual(hn.middle_list, [])
+
+        # other empty attribute lists are also free of the spurious '' element
+        self.assertEqual(hn.title_list, [])
+        self.assertEqual(hn.first_list, ['Jane'])
+        self.assertEqual(hn.last_list, ['Doe'])
+
     # Leaving already-capitalized names alone
     def test_no_change_to_mixed_chase(self) -> None:
         hn = HumanName('Shirley Maclaine')