more utf-8 issues

This commit is contained in:
j 2014-10-31 15:41:46 +01:00
commit a306370f0d
3 changed files with 4 additions and 4 deletions

View file

@ -28,8 +28,8 @@ def info(path):
return data
def extract_text(path):
with open(path) as fd:
data = fd.read()
with open(path, 'rb') as fd:
data = fd.read().decode('utf-8', errors='replace')
return data
def extract_isbn(text):