Redmine #2619: more specific cordex-related tweaks

structure_assignment/assign_collocation_structures.py

@@ -50,7 +50,7 @@ def get_mwe_index_map(mapper_file_name):
 def get_mwe_structure_map(csv_file_name, mwe_components_map, mwe_index_map, structure_components_map):
     mwe_structure_map = defaultdict(set)
     csv_file = codecs.open(csv_file_name, 'r')
-    reader = csv.DictReader(csv_file, delimiter=',')
+    reader = csv.DictReader(csv_file, delimiter='\t')
     for row in reader:
         structure_id = int(row['Structure_ID'])
         collocation_id = int(row['Collocation_ID'])

structure_assignment/constants.py

@@ -8,7 +8,7 @@ FILE_MAP = {'strings-list': 'strings.txt',
             'tei-ids-collocation': 'tei_ids_collocations.xml',
             'tei-ids-all': 'tei_ids_all.xml',
             'collocations': 'collocation_matches.csv',
-            'collocation-mapper': 'mapper.txt',
+            'collocation-mapper': 'mapper.csv',
             'structures-old': 'structures_old.xml',
             'structures-new': 'structures_new.xml',
             'dictionary': 'dictionary.xml',

structure_assignment/pipeline.py

@@ -204,8 +204,9 @@ class Pipeline:
         structure_file_name = self.file_map['structures-old']
         input_file_name = self.file_map['tei-initial']
         output_file_name = self.file_map['collocations']
+        mapper_file_name = self.file_map['collocation-mapper']
 
-        extractor = cordex.Pipeline(structure_file_name, fixed_restriction_order=True, statistics=False, collocation_sentence_map_dest=self.tmp_directory)
+        extractor = cordex.Pipeline(structure_file_name, fixed_restriction_order=True, statistics=False, collocation_sentence_map_dest=mapper_file_name)
         extraction = extractor(input_file_name)
         extraction.write(output_file_name)