some fixes

migration/__init__.py

@@ -116,7 +116,7 @@ async def shouts_handle(storage, args):
 		topics_dataset_bodies.append(u" ".join([x.strip().lower() for x in texts]))
 		topics_dataset_tlist.append(shout['topics'])
   
-	np.savetxt('topics_dataset.csv', (topics_dataset_bodies, topics_dataset_tlist), delimiter=',', fmt='%s')
+	# np.savetxt('topics_dataset.csv', (topics_dataset_bodies, topics_dataset_tlist), delimiter=',', fmt='%s')
 
 	print('[migration] ' + str(counter) + ' content items were migrated')
 	print('[migration] ' + str(pub_counter) + ' have been published')
@@ -160,10 +160,10 @@ def bson_handle():
 	bson2json.json_tables()
 
 
-def export_one(slug, storage):
+def export_one(slug, storage, args = None):
 	topics_handle(storage)
 	users_handle(storage)
-	shouts_handle(storage)
+	shouts_handle(storage, args)
 	export_slug(slug, storage)
 
 

migration/tables/topics.py

@@ -1,3 +1,4 @@
+from logging import exception
 from migration.extract import extract_md, html2text
 from base.orm import local_session
 from orm import Topic, Community
@@ -14,15 +15,14 @@ def migrate(entry):
 	topic_dict['body'] = extract_md(html2text(body_orig), entry['_id'])
 	with local_session() as session:
 		slug = topic_dict['slug']
-		topic = session.query(Topic).filter(Topic.slug == slug).first()
+		t: Topic = session.query(Topic).filter(Topic.slug == slug).first() or Topic.create(**topic_d) or raise Exception('topic not created')  # type: ignore
-		if not topic: 
+		if t:
-			topic = Topic.create(**topic_dict)
+			if len(t.title) > len(topic_dict['title']): 
-		if len(topic.title) > len(topic_dict['title']):
+				Topic.update(t, {'title': topic_dict['title']})
-			topic.update({ 'title':  topic_dict['title'] })
+			if len(t.body) < len(topic_dict['body']): 
-		if len(topic.body) < len(topic_dict['body']):
+				Topic.update(t, { 'body':  topic_dict['body'] })
-			topic.update({ 'body':  topic_dict['body'] })
+			session.commit()
-		session.commit()
 	# print(topic.__dict__)
-	rt = topic.__dict__.copy()
+	rt = t.__dict__.copy()
 	del rt['_sa_instance_state']
 	return rt

schema.graphql

@@ -380,6 +380,8 @@ type Author {
   name: String!
   userpic: String
   caption: String # only for full shout 
+  bio: strings
+  links: [String]
 }
 
 # is publication