Local part of JID and domain part of JID needs to be compared in case insensitive way. We need to verify code and update it if required.
Andrzej Wójcik (Tigase) commented 8 years ago
I updated code to work with lowercased JIDs on database side to properly compare JID.
However this change requires update of MongoDB database, so that existing jids are also lowercased and hashes of JIDs are updated with hashes of lowercased jids. For that I prepared updateSchema method in every repository which needs to be converted and created MongoSchemaUpdater class with main method which allows to run schema update from command line.
%wojtek Anything else I should do? If not then please close this task.
wojciech.kapcia@tigase.net commented 8 years ago
After the discussion it was decided we should use:
sha1 columns for the purpose of querying data;
maintain JID case in JID column.
This way it will be in line with specification and Tigase JID/BareJID implementation (maintaining JID case while performing case insensitive comparison).
Andrzej Wójcik (Tigase) commented 8 years ago
I improved support for JID comparison in Mongo DB implementation by usage of JID hashes for lookup and storing original JID value as well.
Andrzej Wójcik (Tigase) commented 8 years ago
MongoDB schema is updated but in case of Mongo we use SHA-256 hashing algorithm.
For schemas for other databases and hosted in different projects I created separate tasks.
Local part of JID and domain part of JID needs to be compared in case insensitive way. We need to verify code and update it if required.