Merge pull request #19 from akb89/develop

Develop
akb89 · Jun 6, 2019 · bdd28d2 · bdd28d2
2 parents 9817ce1 + d126b9a
commit bdd28d2
Show file tree

Hide file tree

Showing 6 changed files with 12 additions and 9 deletions.
diff --git a/pyfn/main.py b/pyfn/main.py
@@ -69,9 +69,9 @@ def _convert(args):
                 'need to specify the --sent parameter pointing at the '
                 '.sentences file absolute filepath')
         annosets = semaforu.unmarshall_annosets(args.source_path, args.sent)
-    ## Starting marshalling
-    os.makedirs(args.target_path, exist_ok=True)
+    # Starting marshalling
     if args.target_format == 'bios':
+        os.makedirs(args.target_path, exist_ok=True)
         biosm.marshall_annosets_dict(annosets_dict, args.target_path,
                                      args.filter, args.output_sentences,
                                      args.excluded_frames,
@@ -90,6 +90,7 @@ def _convert(args):
                                   args.excluded_sentences,
                                   args.excluded_annosets)
     if args.target_format == 'semafor':
+        os.makedirs(args.target_path, exist_ok=True)
         semaform.marshall_annosets_dict(annosets_dict, args.target_path,
                                         args.filter, args.output_sentences,
                                         args.excluded_frames,
@@ -126,7 +127,9 @@ def main():
                                 help='absolute filepath to source dir')
     parser_convert.add_argument('--target', required=True,
                                 dest='target_path',
-                                help='absolute filepath to target dir')
+                                help='absolute path to target dir or file.'
+                                     'is dirpath if --to is semafor or bios. '
+                                     'is filepath if --to is semeval')
     parser_convert.add_argument('--from', required=True,
                                 dest='source_format',
                                 choices=['semafor', 'bios', 'semeval',

diff --git a/pyfn/marshalling/unmarshallers/bios.py b/pyfn/marshalling/unmarshallers/bios.py
@@ -115,7 +115,7 @@ def _create_annoset(lines, sentence, annoset_id):
 
 def _create_sent_dict(bios_filepath):
     sent_dict = {}
-    with open(bios_filepath, 'r') as bios_stream:
+    with open(bios_filepath, 'r', encoding='utf-8') as bios_stream:
         tokens = []
         sent_index = -1
         for line in bios_stream:
@@ -141,7 +141,7 @@ def unmarshall_annosets(bios_filepath, sent_filepath=None):
         sent_dict = marsh_utils.get_sent_dict(sent_filepath)
     else:
         sent_dict = _create_sent_dict(bios_filepath)
-    with open(bios_filepath, 'r') as bios_stream:
+    with open(bios_filepath, 'r', encoding='utf-8') as bios_stream:
         index = -1
         annoset_id = 0
         lines = []

diff --git a/pyfn/marshalling/unmarshallers/semafor.py b/pyfn/marshalling/unmarshallers/semafor.py
@@ -71,7 +71,7 @@ def unmarshall_annosets(semafor_filepath, sent_filepath):
     """Unmarshall a semafor .frame.elements file to pyfn.AnnotationSets."""
     annosets = []
     sent_dict = marsh_utils.get_sent_dict(sent_filepath)
-    with open(semafor_filepath, 'r') as semafor_stream:
+    with open(semafor_filepath, 'r', encoding='utf-8') as semafor_stream:
         for line in semafor_stream:
             line = line.strip()
             line_split = line.split('\t')

diff --git a/pyfn/utils/config.py b/pyfn/utils/config.py
@@ -16,6 +16,6 @@
 def load(config_file):
     """Load an ImmutableConfig from a YAML configuration file."""
     logger.info('Loading config from file {}'.format(config_file))
-    with open(config_file, 'r') as config_stream:
+    with open(config_file, 'r', encoding='utf-8') as config_stream:
         config = yaml.safe_load(config_stream)
         return ImmutableConfig(config)
diff --git a/pyfn/utils/marshalling.py b/pyfn/utils/marshalling.py
@@ -33,7 +33,7 @@ def get_sent_dict(sent_filepath):
     """
     sent_dict = {}
     sent_iter = 0
-    with open(sent_filepath, 'r') as sent_stream:
+    with open(sent_filepath, 'r', encoding='utf-8') as sent_stream:
         for line in sent_stream:
             line = line.rstrip()
             sent_dict[sent_iter] = line

diff --git a/setup.py b/setup.py
@@ -16,7 +16,7 @@
     author_email='[email protected]',
     long_description=long_description,
     long_description_content_type='text/markdown',
-    version='1.3.4',
+    version='1.3.6',
     url='https://gitlab.com/akb89/pyfn',
     download_url='https://pypi.org/project/pyfn/#files',
     license='MIT',