Changeset - ad374c171656
[Not reviewed]
default
0 1 0
Mads Kiilerich - 11 years ago 2015-01-06 00:54:36
madski@unity3d.com
comments: introduce lazy join of ChangesetStatus on ChangesetComment

99% of all uses of ChangesetComment will also need the optional corresponding
ChangesetStatus (if any) on the status_change relationship. Fetching it on
demand gives a lot of roundtrips and might be slow ... but adding explicit bulk
queries everywhere do not seem feasible. Adding lazy=joined in the data model
seems like the best solution.

Loading a pull request with 10 comments on a slow repo goes from 3.5 s to
2.7 s.
1 file changed with 4 insertions and 1 deletions:
0 comments (0 inline, 0 general)
kallithea/model/db.py
Show inline comments
 
@@ -2056,193 +2056,196 @@ class CacheInvalidation(Base, BaseModel)
 
        """
 
        # prefix, repo_name, suffix
 
        return self._cache_key_partition()[2]
 

	
 
    @classmethod
 
    def clear_cache(cls):
 
        """
 
        Delete all cache keys from database.
 
        Should only be run when all instances are down and all entries thus stale.
 
        """
 
        cls.query().delete()
 
        Session().commit()
 

	
 
    @classmethod
 
    def _get_cache_key(cls, key):
 
        """
 
        Wrapper for generating a unique cache key for this instance and "key".
 
        key must / will start with a repo_name which will be stored in .cache_args .
 
        """
 
        import kallithea
 
        prefix = kallithea.CONFIG.get('instance_id', '')
 
        return "%s%s" % (prefix, key)
 

	
 
    @classmethod
 
    def set_invalidate(cls, repo_name, delete=False):
 
        """
 
        Mark all caches of a repo as invalid in the database.
 
        """
 
        inv_objs = Session().query(cls).filter(cls.cache_args == repo_name).all()
 
        log.debug('for repo %s got %s invalidation objects'
 
                  % (safe_str(repo_name), inv_objs))
 

	
 
        for inv_obj in inv_objs:
 
            log.debug('marking %s key for invalidation based on repo_name=%s'
 
                      % (inv_obj, safe_str(repo_name)))
 
            if delete:
 
                Session().delete(inv_obj)
 
            else:
 
                inv_obj.cache_active = False
 
                Session().add(inv_obj)
 
        Session().commit()
 

	
 
    @classmethod
 
    def test_and_set_valid(cls, repo_name, kind, valid_cache_keys=None):
 
        """
 
        Mark this cache key as active and currently cached.
 
        Return True if the existing cache registration still was valid.
 
        Return False to indicate that it had been invalidated and caches should be refreshed.
 
        """
 

	
 
        key = (repo_name + '_' + kind) if kind else repo_name
 
        cache_key = cls._get_cache_key(key)
 

	
 
        if valid_cache_keys and cache_key in valid_cache_keys:
 
            return True
 

	
 
        inv_obj = cls.query().filter(cls.cache_key == cache_key).scalar()
 
        if not inv_obj:
 
            inv_obj = CacheInvalidation(cache_key, repo_name)
 
        if inv_obj.cache_active:
 
            return True
 
        inv_obj.cache_active = True
 
        Session().add(inv_obj)
 
        Session().commit()
 
        return False
 

	
 
    @classmethod
 
    def get_valid_cache_keys(cls):
 
        """
 
        Return opaque object with information of which caches still are valid
 
        and can be used without checking for invalidation.
 
        """
 
        return set(inv_obj.cache_key for inv_obj in cls.query().filter(cls.cache_active).all())
 

	
 

	
 
class ChangesetComment(Base, BaseModel):
 
    __tablename__ = 'changeset_comments'
 
    __table_args__ = (
 
        Index('cc_revision_idx', 'revision'),
 
        {'extend_existing': True, 'mysql_engine': 'InnoDB',
 
         'mysql_charset': 'utf8', 'sqlite_autoincrement': True},
 
    )
 
    comment_id = Column('comment_id', Integer(), nullable=False, primary_key=True)
 
    repo_id = Column('repo_id', Integer(), ForeignKey('repositories.repo_id'), nullable=False)
 
    revision = Column('revision', String(40), nullable=True)
 
    pull_request_id = Column("pull_request_id", Integer(), ForeignKey('pull_requests.pull_request_id'), nullable=True)
 
    line_no = Column('line_no', Unicode(10), nullable=True)
 
    hl_lines = Column('hl_lines', Unicode(512), nullable=True)
 
    f_path = Column('f_path', Unicode(1000), nullable=True)
 
    user_id = Column('user_id', Integer(), ForeignKey('users.user_id'), nullable=False)
 
    text = Column('text', UnicodeText(25000), nullable=False)
 
    created_on = Column('created_on', DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
    modified_at = Column('modified_at', DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 

	
 
    author = relationship('User', lazy='joined')
 
    repo = relationship('Repository')
 
    status_change = relationship('ChangesetStatus', cascade="all, delete-orphan")
 
    # status_change is frequently used directly in templates - make it a lazy
 
    # join to avoid fetching each related ChangesetStatus on demand.
 
    # There will only be one ChangesetStatus referencing each comment so the join will not explode.
 
    status_change = relationship('ChangesetStatus', cascade="all, delete-orphan", lazy='joined')
 
    pull_request = relationship('PullRequest')
 

	
 
    @classmethod
 
    def get_users(cls, revision=None, pull_request_id=None):
 
        """
 
        Returns user associated with this ChangesetComment. ie those
 
        who actually commented
 

	
 
        :param cls:
 
        :param revision:
 
        """
 
        q = Session().query(User)\
 
                .join(ChangesetComment.author)
 
        if revision:
 
            q = q.filter(cls.revision == revision)
 
        elif pull_request_id:
 
            q = q.filter(cls.pull_request_id == pull_request_id)
 
        return q.all()
 

	
 

	
 
class ChangesetStatus(Base, BaseModel):
 
    __tablename__ = 'changeset_statuses'
 
    __table_args__ = (
 
        Index('cs_revision_idx', 'revision'),
 
        Index('cs_version_idx', 'version'),
 
        UniqueConstraint('repo_id', 'revision', 'version'),
 
        {'extend_existing': True, 'mysql_engine': 'InnoDB',
 
         'mysql_charset': 'utf8', 'sqlite_autoincrement': True}
 
    )
 
    STATUS_NOT_REVIEWED = DEFAULT = 'not_reviewed'
 
    STATUS_APPROVED = 'approved'
 
    STATUS_REJECTED = 'rejected'
 
    STATUS_UNDER_REVIEW = 'under_review'
 

	
 
    STATUSES = [
 
        (STATUS_NOT_REVIEWED, _("Not Reviewed")),  # (no icon) and default
 
        (STATUS_APPROVED, _("Approved")),
 
        (STATUS_REJECTED, _("Rejected")),
 
        (STATUS_UNDER_REVIEW, _("Under Review")),
 
    ]
 

	
 
    changeset_status_id = Column('changeset_status_id', Integer(), nullable=False, primary_key=True)
 
    repo_id = Column('repo_id', Integer(), ForeignKey('repositories.repo_id'), nullable=False)
 
    user_id = Column("user_id", Integer(), ForeignKey('users.user_id'), nullable=False, unique=None)
 
    revision = Column('revision', String(40), nullable=False)
 
    status = Column('status', String(128), nullable=False, default=DEFAULT)
 
    changeset_comment_id = Column('changeset_comment_id', Integer(), ForeignKey('changeset_comments.comment_id'))
 
    modified_at = Column('modified_at', DateTime(), nullable=False, default=datetime.datetime.now)
 
    version = Column('version', Integer(), nullable=False, default=0)
 
    pull_request_id = Column("pull_request_id", Integer(), ForeignKey('pull_requests.pull_request_id'), nullable=True)
 

	
 
    author = relationship('User', lazy='joined')
 
    repo = relationship('Repository')
 
    comment = relationship('ChangesetComment', lazy='joined')
 
    pull_request = relationship('PullRequest')
 

	
 
    def __unicode__(self):
 
        return u"<%s('%s:%s')>" % (
 
            self.__class__.__name__,
 
            self.status, self.author
 
        )
 

	
 
    @classmethod
 
    def get_status_lbl(cls, value):
 
        return dict(cls.STATUSES).get(value)
 

	
 
    @property
 
    def status_lbl(self):
 
        return ChangesetStatus.get_status_lbl(self.status)
 

	
 

	
 
class PullRequest(Base, BaseModel):
 
    __tablename__ = 'pull_requests'
 
    __table_args__ = (
 
        {'extend_existing': True, 'mysql_engine': 'InnoDB',
 
         'mysql_charset': 'utf8', 'sqlite_autoincrement': True},
 
    )
 

	
 
    # values for .status
 
    STATUS_NEW = u'new'
 
    STATUS_CLOSED = u'closed'
 

	
 
    pull_request_id = Column('pull_request_id', Integer(), nullable=False, primary_key=True)
 
    title = Column('title', Unicode(256), nullable=True)
 
    description = Column('description', UnicodeText(10240), nullable=True)
 
    status = Column('status', Unicode(256), nullable=False, default=STATUS_NEW) # only for closedness, not approve/reject/etc
 
    created_on = Column('created_on', DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
    updated_on = Column('updated_on', DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
    user_id = Column("user_id", Integer(), ForeignKey('users.user_id'), nullable=False, unique=None)
 
    _revisions = Column('revisions', UnicodeText(20500))  # 500 revisions max
 
    org_repo_id = Column('org_repo_id', Integer(), ForeignKey('repositories.repo_id'), nullable=False)
 
    org_ref = Column('org_ref', Unicode(256), nullable=False)
 
    other_repo_id = Column('other_repo_id', Integer(), ForeignKey('repositories.repo_id'), nullable=False)
 
    other_ref = Column('other_ref', Unicode(256), nullable=False)
 

	
 
    @hybrid_property
0 comments (0 inline, 0 general)