gryf/ebook-converter
1
0
Fork 0
mirror of https://github.com/gryf/ebook-converter.git synced 2026-03-08 02:25:57 +01:00
Code Issues Projects Releases Wiki Activity

Clean up titlecase module

Browse Source
This commit is contained in:
gryf
2020-05-17 20:43:16 +02:00
parent 2d26e8916d
commit b4067b29c9
1 changed files with 13 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

43
ebook_converter/utils/titlecase.py
View File

@@ -6,11 +6,6 @@ License: http://www.opensource.org/licenses/mit-license.php
""" """
import re import re
from ebook_converter.utils.icu import capitalize, upper
__all__ = ['titlecase']
__version__ = '0.5'
SMALL = 'a|an|and|as|at|but|by|en|for|if|in|of|on|or|the|to|v\\.?|via|vs\\.?' SMALL = 'a|an|and|as|at|but|by|en|for|if|in|of|on|or|the|to|v\\.?|via|vs\\.?'
PUNCT = r"""!"#$%&'‘’()*+,\-‒–—―./:;?@[\\\]_`{|}~""" PUNCT = r"""!"#$%&'‘’()*+,\-‒–—―./:;?@[\\\]_`{|}~"""
@@ -19,23 +14,13 @@ SMALL_WORDS = re.compile(r'^(%s)$' % SMALL, re.I)
INLINE_PERIOD = re.compile(r'[a-z][.][a-z]', re.I) INLINE_PERIOD = re.compile(r'[a-z][.][a-z]', re.I)
UC_ELSEWHERE = re.compile(r'[%s]*?[a-zA-Z]+[A-Z]+?' % PUNCT) UC_ELSEWHERE = re.compile(r'[%s]*?[a-zA-Z]+[A-Z]+?' % PUNCT)
CAPFIRST = re.compile(str(r"^[%s]*?(\w)" % PUNCT), flags=re.UNICODE) CAPFIRST = re.compile(str(r"^[%s]*?(\w)" % PUNCT), flags=re.UNICODE)
SMALL_FIRST = re.compile(r'^([%s]*)(%s)\b' % (PUNCT, SMALL), re.I|re.U) SMALL_FIRST = re.compile(r'^([%s]*)(%s)\b' % (PUNCT, SMALL), re.I | re.U)
SMALL_LAST = re.compile(r'\b(%s)[%s]?$' % (SMALL, PUNCT), re.I|re.U) SMALL_LAST = re.compile(r'\b(%s)[%s]?$' % (SMALL, PUNCT), re.I | re.U)
SMALL_AFTER_NUM = re.compile(r'(\d+\s+)(a|an|the)\b', re.I|re.U) SMALL_AFTER_NUM = re.compile(r'(\d+\s+)(a|an|the)\b', re.I | re.U)
SUBPHRASE = re.compile(r'([:.;?!][ ])(%s)' % SMALL) SUBPHRASE = re.compile(r'([:.;?!][ ])(%s)' % SMALL)
APOS_SECOND = re.compile(r"^[dol]{1}['‘]{1}[a-z]+$", re.I) APOS_SECOND = re.compile(r"^[dol]{1}['‘]{1}[a-z]+$", re.I)
UC_INITIALS = re.compile(r"^(?:[A-Z]{1}\.{1}|[A-Z]{1}\.{1}[A-Z]{1})+$") UC_INITIALS = re.compile(r"^(?:[A-Z]{1}\.{1}|[A-Z]{1}\.{1}[A-Z]{1})+$")
_lang = None
def lang():
global _lang
if _lang is None:
from ebook_converter.utils.localization import get_lang
_lang = get_lang().lower()
return _lang
def titlecase(text): def titlecase(text):
""" """
@@ -49,7 +34,7 @@ def titlecase(text):
""" """
all_caps = upper(text) == text all_caps = text.upper() == text
pat = re.compile(r'(\s+)') pat = re.compile(r'(\s+)')
line = [] line = []
@@ -85,20 +70,18 @@ def titlecase(text):
result = "".join(line) result = "".join(line)
result = SMALL_FIRST.sub(lambda m: '%s%s' % ( result = SMALL_FIRST.sub(lambda m: '%s%s' % (m.group(1),
m.group(1), m.group(2).capitalize()),
capitalize(m.group(2)) result)
), result)
result = SMALL_AFTER_NUM.sub(lambda m: '%s%s' % (m.group(1), result = SMALL_AFTER_NUM.sub(lambda m: '%s%s' % (m.group(1),
capitalize(m.group(2)) m.group(2).capitalize()),
), result) result)
result = SMALL_LAST.sub(lambda m: capitalize(m.group(0)), result) result = SMALL_LAST.sub(lambda m: m.group(0).capitalize(), result)
result = SUBPHRASE.sub(lambda m: '%s%s' % ( result = SUBPHRASE.sub(lambda m: '%s%s' % (m.group(1),
m.group(1), m.group(2).capitalize()),
capitalize(m.group(2)) result)
), result)
return result return result