Adapt tests to latest changes in ckanext-harvest #263
Conversation
| # TODO move to ckanext-harvest | ||
| @pytest.fixture | ||
| def harvest_setup(): | ||
| harvest_model.setup() |
There was a problem hiding this comment.
Perhaps is best to keep the concerns for the clean_db and harvest_setup fixtures separated and keep harvest_setup but refactoring it to use migrate_db_for('harvest') rather than harvest_model.setup()?
There was a problem hiding this comment.
We could not solve the problem otherwise. Using migrate_db_for('harvest') in harvest_setup or redefining clean_db in the harvester-specific tests (test_harvester.py & test_json_harvester.py) does not seem to solve the issue.
All Tests that include clean_db still throw the following exception:
sqlalchemy.exc.ProgrammingError: (psycopg2.errors.UndefinedTable) relation "harvest_object_extra" does not exist
...
/srv/app/src/ckan/ckan/tests/helpers.py:83: in reset_db
model.repo.rebuild_db()
/srv/app/src/ckan/ckan/model/init.py:229: in rebuild_db
self.delete_all()
/srv/app/src/ckan/ckan/model/init.py:250: in delete_all
connection.execute('delete from "%s"' % table.name)
/usr/lib/python3.8/site-packages/sqlalchemy/engine/base.py:982: in execute
return self.execute_text(object, multiparams, params)
/usr/lib/python3.8/site-packages/sqlalchemy/engine/base.py:1149: in _execute_text
ret = self._execute_context(
/usr/lib/python3.8/site-packages/sqlalchemy/engine/base.py:1247: in _execute_context
self._handle_dbapi_exception(
/usr/lib/python3.8/site-packages/sqlalchemy/engine/base.py:1466: in _handle_dbapi_exception
util.raise_from_cause(sqlalchemy_exception, exc_info)
/usr/lib/python3.8/site-packages/sqlalchemy/util/compat.py:399: in raise_from_cause
reraise(type(exception), exception, tb=exc_tb, cause=cause)
/usr/lib/python3.8/site-packages/sqlalchemy/util/compat.py:153: in reraise
raise value.with_traceback(tb)
/usr/lib/python3.8/site-packages/sqlalchemy/engine/base.py:1243: in _execute_context
Any ideas on how to initialize/create the plugin db tables or to put the right scope for harvest_setup ?
Maybe this was also mentioned in this discussion
|
@amercader We would like to get the tests working again. Perhaps you'll find some time to have a quick look at this and let us know what do you think about our last investigations? Otherwise, we suggest merging the currently working solution to move forward. |
|
My apologies @seitenbau-govdata, reading the linked discussion this makes perfect sense. |
|
@amercader No problem. Don't worry. Thanks! 😃 |
The workflow tests are currently broken when run with the master version of
ckanext-harvest. With this commit inckanext-harvestthe way of initializing the db has been changed.This PR updates the tests for
ckanext-dcatto solve the issue.But I`m not quite sure if this is an ideal solution. Could someone take a closer look at this?