From 826221c0f520b5fcbd5d931032a3cf5035f12df7 Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Fri, 10 Dec 2010 12:27:06 -0700
Subject: [PATCH] Various fixes to Title Casing. Fixes #7846 (Title Case
 doesn't quite work)

---
 src/calibre/utils/icu.py       | 15 +++++++++------
 src/calibre/utils/titlecase.py | 15 ++++++++++-----
 2 files changed, 19 insertions(+), 11 deletions(-)

diff --git a/src/calibre/utils/icu.py b/src/calibre/utils/icu.py
index 4b0f6d4821..22611813f4 100644
--- a/src/calibre/utils/icu.py
+++ b/src/calibre/utils/icu.py
@@ -77,6 +77,9 @@ def py_strcmp(a, b):
 def icu_case_sensitive_strcmp(collator, a, b):
     return collator.strcmp(a, b)
 
+def icu_capitalize(s):
+    s = lower(s)
+    return s.replace(s[0], upper(s[0]), 1)
 
 load_icu()
 load_collator()
@@ -104,10 +107,6 @@ lower = (lambda s: s.lower()) if _icu_not_ok else \
 title_case = (lambda s: s.title()) if _icu_not_ok else \
     partial(_icu.title, get_locale())
 
-def icu_capitalize(s):
-    s = lower(s)
-    return s.replace(s[0], upper(s[0]))
-
 capitalize = (lambda s: s.capitalize()) if _icu_not_ok else \
     (lambda s: icu_capitalize(s))
 
@@ -226,12 +225,16 @@ pêché'''
     test_strcmp(german + french)
 
     print '\nTesting case transforms in current locale'
-    for x in ('a', 'Alice\'s code'):
+    from calibre.utils.titlecase import titlecase
+    for x in ('a', 'Alice\'s code', 'macdonald\'s machine', '02 the wars'):
         print 'Upper:     ', x, '->', 'py:', x.upper().encode('utf-8'), 'icu:', upper(x).encode('utf-8')
         print 'Lower:     ', x, '->', 'py:', x.lower().encode('utf-8'), 'icu:', lower(x).encode('utf-8')
-        print 'Title:     ', x, '->', 'py:', x.title().encode('utf-8'), 'icu:', title_case(x).encode('utf-8')
+        print 'Title:     ', x, '->', 'py:', x.title().encode('utf-8'), 'icu:', title_case(x).encode('utf-8'), 'titlecase:', titlecase(x).encode('utf-8')
         print 'Capitalize:', x, '->', 'py:', x.capitalize().encode('utf-8'), 'icu:', capitalize(x).encode('utf-8')
         print
 
 # }}}
 
+if __name__ == '__main__':
+    test()
+
diff --git a/src/calibre/utils/titlecase.py b/src/calibre/utils/titlecase.py
index bbc4c26688..bf2f9a78d4 100755
--- a/src/calibre/utils/titlecase.py
+++ b/src/calibre/utils/titlecase.py
@@ -23,11 +23,12 @@ UC_ELSEWHERE = re.compile(r'[%s]*?[a-zA-Z]+[A-Z]+?' % PUNCT)
 CAPFIRST = re.compile(r"^[%s]*?([A-Za-z])" % PUNCT)
 SMALL_FIRST = re.compile(r'^([%s]*)(%s)\b' % (PUNCT, SMALL), re.I)
 SMALL_LAST = re.compile(r'\b(%s)[%s]?$' % (SMALL, PUNCT), re.I)
+SMALL_AFTER_NUM = re.compile(r'(\d+\s+)(a|an|the)\b', re.I)
 SUBPHRASE = re.compile(r'([:.;?!][ ])(%s)' % SMALL)
 APOS_SECOND = re.compile(r"^[dol]{1}['‘]{1}[a-z]+$", re.I)
 ALL_CAPS = re.compile(r'^[A-Z\s%s]+$' % PUNCT)
 UC_INITIALS = re.compile(r"^(?:[A-Z]{1}\.{1}|[A-Z]{1}\.{1}[A-Z]{1})+$")
-MAC_MC = re.compile(r"^([Mm]a?c)(\w+)")
+MAC_MC = re.compile(r"^([Mm]a?c)(.+)")
 
 def titlecase(text):
 
@@ -44,7 +45,7 @@ def titlecase(text):
 
     all_caps = ALL_CAPS.match(text)
 
-    words = re.split('\s', text)
+    words = re.split('\s+', text)
     line = []
     for word in words:
         if all_caps:
@@ -55,8 +56,8 @@ def titlecase(text):
                 word = icu_lower(word)
 
         if APOS_SECOND.match(word):
-            word = word.replace(word[0], icu_upper(word[0]))
-            word = word.replace(word[2], icu_upper(word[2]))
+            word = word.replace(word[0], icu_upper(word[0]), 1)
+            word = word[:2] + icu_upper(word[2]) + word[3:]
             line.append(word)
             continue
         if INLINE_PERIOD.search(word) or UC_ELSEWHERE.match(word):
@@ -67,7 +68,7 @@ def titlecase(text):
             continue
 
         match = MAC_MC.match(word)
-        if match:
+        if match and not match.group(2).startswith('hin'):
             line.append("%s%s" % (capitalize(match.group(1)),
                                   capitalize(match.group(2))))
             continue
@@ -85,6 +86,10 @@ def titlecase(text):
         capitalize(m.group(2))
     ), result)
 
+    result = SMALL_AFTER_NUM.sub(lambda m: '%s%s' % (m.group(1),
+        capitalize(m.group(2))
+    ), result)
+
     result = SMALL_LAST.sub(lambda m: capitalize(m.group(0)), result)
 
     result = SUBPHRASE.sub(lambda m: '%s%s' % (