##// END OF EJS Templates
caches: use different namespaces for conditional_cache wrapper....
caches: use different namespaces for conditional_cache wrapper. Before we used one namespace for ALL meaning caches were kept in single file, or single DB row. While this makes sense for single functions, a per-user, per-group cache should be divided to single files.

File last commit:

r1061:3f46ec25 default
r2814:45a75c50 default
Show More
062_version_4_5_0.py
31 lines | 939 B | text/x-python | PythonLexer
import logging
from sqlalchemy import Column, MetaData, Unicode
from rhodecode.lib.dbmigrate.versions import _reset_base
log = logging.getLogger(__name__)
def upgrade(migrate_engine):
"""
Upgrade operations go here.
Don't create your own engine; bind migrate_engine to your metadata
"""
_reset_base(migrate_engine)
from rhodecode.lib.dbmigrate.schema import db_4_5_0_0 as db
# Add shadow merge ref column to pull request table.
pr_table = db.PullRequest.__table__
pr_col = Column('shadow_merge_ref', Unicode(255), nullable=True)
pr_col.create(table=pr_table)
# Add shadow merge ref column to pull request version table.
pr_version_table = db.PullRequestVersion.__table__
pr_version_col = Column('shadow_merge_ref', Unicode(255), nullable=True)
pr_version_col.create(table=pr_version_table)
def downgrade(migrate_engine):
meta = MetaData()
meta.bind = migrate_engine