from lxml import etree
import re
+from librarian.dcparser import Person
+
def _register_function(f):
""" Register extension function with lxml """
ns = etree.FunctionNamespace('http://wolnelektury.pl/functions')
ns[f.__name__] = f
-def reg_substitute_entities():
+def reg_substitute_entities():
ENTITY_SUBSTITUTIONS = [
(u'---', u'—'),
(u'--', u'–'),
(u'...', u'…'),
(u',,', u'„'),
+ #(u',,', u'“'),
(u'"', u'”'),
]
text = ''.join(text)
if not wrapping:
return text
-
+
words = re.split(r'\s', text)
-
+
line_length = 0
lines = [[]]
for word in words:
""" Converts "Name, Forename" to "Forename Name" """
if isinstance(text, list):
text = ''.join(text)
- return ' '.join([t.strip() for t in text.split(',', 1)[::-1]])
+ return Person.from_text(text).readable()
_register_function(person_name)
+def reg_texcommand():
+ def texcommand(context, text):
+ """Remove non-letters"""
+ if isinstance(text, list):
+ text = ''.join(text)
+ return re.sub(r'[^a-zA-Z]', '', text).strip()
+ _register_function(texcommand)
+
+