fix

[librarian.git] / src / librarian / hyphenator.py

diff --git a/src/librarian/hyphenator.py b/src/librarian/hyphenator.py
index aa5b4c3..c2fbb81 100644 (file)
--- a/src/librarian/hyphenator.py
+++ b/src/librarian/hyphenator.py
@@ -14,8 +14,6 @@ info@wilbertberendsen.nl
 License: LGPL.
 
 """
 License: LGPL.
 
 """

-from __future__ import print_function, unicode_literals
-

 import sys
 import re
 
 import sys
 import re
 


@@ -81,13 +79,13 @@ class Hyph_dict(object):
     """
     def __init__(self, filename):
         self.patterns = {}
     """
     def __init__(self, filename):
         self.patterns = {}

-        f = open(filename)
+        f = open(filename, 'rb')

         charset = f.readline().strip()
         charset = f.readline().strip()

-        if charset.startswith('charset '):
+        if charset.startswith(b'charset '):

             charset = charset[8:].strip()
 
         for pat in f:
             charset = charset[8:].strip()
 
         for pat in f:

-            pat = pat.decode(charset).strip()
+            pat = pat.decode(charset.decode('latin1')).strip()

             if not pat or pat[0] == '%': continue
             # replace ^^hh with the real character
             pat = parse_hex(hexrepl, pat)
             if not pat or pat[0] == '%': continue
             # replace ^^hh with the real character
             pat = parse_hex(hexrepl, pat)


@@ -211,7 +209,7 @@ class Hyphenator(object):
         the string 'let-ter-gre-pen'. The hyphen string to use can be
         given as the second parameter, that defaults to '-'.
         """
         the string 'let-ter-gre-pen'. The hyphen string to use can be
         given as the second parameter, that defaults to '-'.
         """

-        if isinstance(word, str):
+        if isinstance(word, bytes):

             word = word.decode('latin1')
         l = list(word)
         for p in reversed(self.positions(word)):
             word = word.decode('latin1')
         l = list(word)
         for p in reversed(self.positions(word)):