There are two possible solutions for this issue:
1.) check if we are operating on an empty database
2.) change our upgrade function in a way that if a upgrade script is not found 
for X -> Z it tries to run X -> Y -> Z

For 1.) there are two possible ways:
1.a.) check if file already exists before sqlite3.connect() to this file
1.b.) change _get_schema_version() to return -1 if SQLITE_MASTER is empty

I vote for 1.a.) as a fix for this bug, and work on 2.) for the future.

Upgrading database scheme is broken for newly created databases
You received this bug notification because you are a member of Zeitgeist
Framework Team, which is subscribed to Zeitgeist Framework.

Status in Zeitgeist Framework: In Progress

Bug description:
Todays newly created database version 2 broke the upgrade path, we have no 
upgrade script for 0 (which means 'no version entry found') to 2. This happens 
because no-existing databases do not have such table.

$ ZEITGEIST_DATABASE_PATH=/tmp/dfgdsfsddf ./zeitgeist-daemon                    
DEBUG:root:Checking for another running instance...
DEBUG:root:No running instances found.
INFO:zeitgeist.sql:Using database: /tmp/dfgdsfsddf
DEBUG:zeitgeist.sql:Schema 'core' not found: no such table: schema_version
INFO:zeitgeist.sql:Upgrading database 'core' from version 0 to 2. This may take 
a while
CRITICAL:zeitgeist.sql:Failed to upgrade database 'core' from version 0 to 2: 
No module named core_0_2

We have to check if we are operating on a new database.
(Side note: we need tests for upgrades....)

Mailing list:
Post to     :
Unsubscribe :
More help   :

Reply via email to