fix capitalization of prefixes when they are not part of last name #70

Author: derek73

docs/release_log.rst

@@ -3,6 +3,7 @@ Release Log
 * 1.0.0 - August 30, 2018
     - Fix support for nicknames in single quotes (#74)
     - Change prefix handling to support prefixes on first names (#60)
+    - Fix prefix capitalization when not part of last name (#70)
     - No other big changes, just bumping to v1 to indicate approprite project maturity
 * 0.5.8 - August 19, 2018
     - Add "Junior" to suffixes (#76)

nameparser/parser.py

@@ -779,8 +779,8 @@ def join_on_conjunctions(self, pieces, additional_parts_count=0):
 
     ### Capitalization Support
 
-    def cap_word(self, word):
-        if self.is_prefix(word) or self.is_conjunction(word):
+    def cap_word(self, word, attribute):
+        if (self.is_prefix(word) and attribute=='last') or self.is_conjunction(word):
             return word.lower()
         exceptions = self.C.capitalization_exceptions
         if lc(word) in exceptions:
@@ -793,10 +793,10 @@ def cap_after_mac(m):
         else:
             return word.capitalize()
 
-    def cap_piece(self, piece):
+    def cap_piece(self, piece, attribute):
         if not piece:
             return ""
-        replacement = lambda m: self.cap_word(m.group(0))
+        replacement = lambda m: self.cap_word(m.group(0), attribute)
         return self.C.regexes.word.sub(replacement, piece)
 
     def capitalize(self, force=False):
@@ -829,8 +829,8 @@ def capitalize(self, force=False):
         name = u(self)
         if not force and not (name == name.upper() or name == name.lower()):
             return
-        self.title_list  = self.cap_piece(self.title ).split(' ')
-        self.first_list  = self.cap_piece(self.first ).split(' ')
-        self.middle_list = self.cap_piece(self.middle).split(' ')
-        self.last_list   = self.cap_piece(self.last  ).split(' ')
-        self.suffix_list = self.cap_piece(self.suffix).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(', ')

tests.py

@@ -1941,6 +1941,11 @@ def test_portuguese_prefixes(self):
         hn.capitalize()
         self.m(str(hn), 'Joao da Silva do Amaral de Souza', hn)
 
+    def test_capitalize_prefix_clash_on_first_name(self):
+        hn = HumanName("van nguyen")
+        hn.capitalize()
+        self.m(str(hn), 'Van Nguyen', hn)
+
 
 class HumanNameOutputFormatTests(HumanNameTestBase):