From 0b654d8b02e14a4c9740fc8ff085b5337ca695df Mon Sep 17 00:00:00 2001
From: tonyrewin <anton.rewin@gmail.com>
Date: Tue, 31 May 2022 15:19:05 +0300
Subject: [PATCH] migration script fix

---
 .gitignore                                  |  4 +-
 README.md                                   |  1 +
 migration/tables/content_item_categories.py | 10 +--
 migration/tables/tags.py                    | 69 +++++++++++----------
 4 files changed, 46 insertions(+), 38 deletions(-)

diff --git a/.gitignore b/.gitignore
index a4f8206e..65200daa 100644
--- a/.gitignore
+++ b/.gitignore
@@ -137,4 +137,6 @@ Pipfile.lock
 
 migration/data
 migration/content/**/*.md
-.obsidian
\ No newline at end of file
+.obsidian
+
+*.zip
\ No newline at end of file
diff --git a/README.md b/README.md
index 8b30a670..52e5363c 100644
--- a/README.md
+++ b/README.md
@@ -34,3 +34,4 @@ Also see `Dockerfile`
 # How to do an authorized request
 
 Put the header 'Auth' with token from signInQuery or registerQuery.
+
diff --git a/migration/tables/content_item_categories.py b/migration/tables/content_item_categories.py
index 2380f4f2..1eae2191 100644
--- a/migration/tables/content_item_categories.py
+++ b/migration/tables/content_item_categories.py
@@ -16,16 +16,18 @@ def migrate(entry):
 		'slug': entry['slug'],
 		# 'createdBy': entry['createdBy'],
 		# 'createdAt': date_parse(entry['createdAt']),
-		'title': entry['title'], #.lower(),
+		'title': entry['title'].replace('&nbsp;', ' '), #.lower(),
 		'children': [],
 		'community' : Community.default_community.slug,
 		'body' : entry.get('description')
 	}
 	try:
 		with local_session() as session:
-			topic = session.query(Topic).filter(Topic.slug == entry['slug']).first()
-			if not topic:
-				topic = Topic.create(**topic_dict)
+			topic = session.query(Topic).filter(Topic.slug == topic_dict['slug']).first()
+			if not topic: 
+				topic = session.query(Topic).filter(Topic.title == topic_dict['title']).first()
+				if not topic:
+					topic = Topic.create(**topic_dict)
 	except Exception as e:
 		print(e)
 		raise e
diff --git a/migration/tables/tags.py b/migration/tables/tags.py
index 000766ab..0a6d743c 100644
--- a/migration/tables/tags.py
+++ b/migration/tables/tags.py
@@ -5,36 +5,39 @@ from orm import Topic, Community
 from dateutil.parser import parse as date_parse
 
 def migrate(entry):
-    '''
-    type Topic {
-        slug: String! # ID
-        createdBy: Int! # User
-        createdAt: DateTime!
-        title: String
-        parents: [String] # NOTE: topic can have parent topics
-        children: [String] # and children
-    }
-    '''
-    if type(entry['createdAt']) == type(''):
-        ts = date_parse(entry['createdAt'])
-    else:
-        ts = datetime.fromtimestamp(entry['createdAt']/1000)
-    topic_dict = {
-        'slug': entry['slug'],
-        # 'createdBy': entry['createdBy'],
-        # 'createdAt': ts,
-        'title': entry['title'], # .lower(),
-        'children': [],
-        'community' : Community.default_community.slug,
-        'body' : entry.get('description')
-    }
-    try:
-        with local_session() as session:
-            topic = session.query(Topic).filter(Topic.slug == entry['slug']).first()
-            if not topic: topic = Topic.create(**topic_dict)
-    except Exception as e:
-        print(e)
-        raise e
-    
-    topic_dict['tag_id'] = entry['_id']
-    return topic_dict
+	'''
+	type Topic {
+		slug: String! # ID
+		createdBy: Int! # User
+		createdAt: DateTime!
+		title: String
+		parents: [String] # NOTE: topic can have parent topics
+		children: [String] # and children
+	}
+	'''
+	if type(entry['createdAt']) == type(''):
+		ts = date_parse(entry['createdAt'])
+	else:
+		ts = datetime.fromtimestamp(entry['createdAt']/1000)
+	topic_dict = {
+		'slug': entry['slug'],
+		# 'createdBy': entry['createdBy'],
+		# 'createdAt': ts,
+		'title': entry['title'].replace('&nbsp;', ' '), # .lower(),
+		'children': [],
+		'community' : Community.default_community.slug,
+		'body' : entry.get('description')
+	}
+	try:
+		with local_session() as session:
+			topic = session.query(Topic).filter(Topic.slug == topic_dict['slug']).first()
+			if not topic: 
+				topic = session.query(Topic).filter(Topic.title == topic_dict['title']).first()
+				if not topic:
+					topic = Topic.create(**topic_dict)
+	except Exception as e:
+		print(e)
+		raise e
+	
+	topic_dict['tag_id'] = entry['_id']
+	return topic_dict