Resolve "process_poleval plain text"

FastAPI

poldeepner2/utils/data_utils.py

@@ -510,3 +510,45 @@ def map_json_to_iob(json_ann, iob):
                     out_iob += line
     failed_to_add = len(token_dict) - successfully_added
     return out_iob, successfully_added, failed_to_add, derives
+
+
+def is_continued(annotation, next_annotations):
+    if next_annotations == ['O']:
+        return False
+    searched_ann = 'I-{0}'.format(annotation)
+    return searched_ann in next_annotations
+
+
+def iob2_to_iob(iob2_text):
+    iob2_list = []
+    iob1_list = []
+    for line in iob2_text.split('\n'):
+        split = line.split(' ')
+        iob2_list.append((split[0], split[1].split('#')))
+    for i, line in enumerate(iob2_list):
+        current_ann = []
+        if len(iob2_list) == 1:
+            for ann in line[1]:
+                split = ann.split('-')
+                if split[0] == 'B':
+                    current_ann.append('I-{0}'.format(split[1]))
+                else:
+                    current_ann.append(ann)
+            iob1_list.append((line[0], '#'.join(current_ann)))
+        elif i == len(iob2_list)-1:
+            for ann in line[1]:
+                split = ann.split('-')
+                if split[0] == 'B':
+                    current_ann.append('I-{0}'.format(split[1]))
+                else:
+                    current_ann.append(ann)
+            iob1_list.append((line[0], '#'.join(current_ann)))
+        else:
+            for ann in line[1]:
+                split = ann.split('-')
+                if split[0] == 'B' and not is_continued(split[1], iob2_list[i+1][1]):
+                    current_ann.append('I-{0}'.format(split[1]))
+                else:
+                    current_ann.append(ann)
+            iob1_list.append((line[0], '#'.join(current_ann)))
+    return '\n'.join(map(lambda x: '{} {}'.format(x[0], x[1]) , iob1_list))
\ No newline at end of file