bias feature

Author: kmike

webstruct/features/__init__.py

@@ -15,6 +15,7 @@
                   'ZIPCODE', 'EMAIL', 'TEL', 'FAX', 'SUBJ', 'FUNC', 'HOURS'}
 
 EXAMPLE_TOKEN_FEATURES = [
+    bias,
     parent_tag,
     borders,
     block_length,

webstruct/features/token_features.py

@@ -3,6 +3,7 @@
 import re
 
 __all__ = [
+    'bias',
     'token_identity',
     'token_lower',
     'token_shape',
@@ -15,6 +16,11 @@
     'SuffixFeatures',
 ]
 
+
+def bias(html_token):
+    return {'bias': 1}
+
+
 def token_identity(html_token):
     return {'token': html_token.token}
 
@@ -112,4 +118,4 @@ def _shape(token):
     elif re.match('\w+$', token):
         return 'mixedcase'
     else:
-        return 'other'
+        return 'other'