from sqlalchemy import MetaData
from sqlalchemy import Column, MetaData, Table, types
from datetime import datetime
class Message(object):
pass
metadata = sqlalchemy.MetaData()
message = Table("message",
metadata,
Column('id', types.BigInteger(20), primary_key=True),
Column('parent_id', types.BigInteger(20)),
Column('incident_id', types.Integer(11)),
Column('user_id', types.Integer(11)),
Column('reporter_id', types.BigInteger(20)),
Column('service_messageid', types.Unicode(100)),
Column('message_from', types.Unicode(100)),
Column('message_to', types.Unicode(100)),
Column('message', types.Unicode),
Column('message_detail', types.Unicode),
Column('message_type', types.SmallInteger(4)),
Column('message_date', types.DateTime),
Column('message_level', types.SmallInteger(4)),
Column('type', types.Integer(4)),
mysql_engine='InnoDB',
mysql_charset='utf8')
class MultiplePrimaryKey(Model):
_id1 = Column(types.Integer(), primary_key=True)
_id2 = Column(types.Integer(), primary_key=True)
_id3 = Column(types.Integer(), primary_key=True)
class C(Model):
_id = Column(types.Integer(), primary_key=True)
class Doz(Model):
qux_id = Column(types.Integer(), ForeignKey('qux._id'), primary_key=True)
name = Column(types.String())
class AutoGenTableName(Model):
_id = Column(types.Integer(), primary_key=True)
name = Column(types.String())
from sqlalchemy import schema, types
metadata = schema.MetaData()
projects = schema.Table(
'projects',
metadata,
schema.Column('id', types.Integer, primary_key=True, unique=True),
schema.Column('num', types.Integer(), default=0),
schema.Column('loc', types.Text(), default=u''),
schema.Column('geo', types.Text(), default=u''),
)
updates = schema.Table(
'updates',
metadata,
schema.Column('id', types.Integer),
schema.Column('date', types.Integer(), default=0),
schema.Column('type', types.Text(), default=u''),
)
from sqlalchemy.engine import create_engine
db = create_engine('sqlite:///25th_ward.db', echo=False)
metadata.bind = db
metadata.create_all(checkfirst=True)
def init_schema(url):
"""
Setup the songs database connection and initialise the database schema.
:param url: The database to setup
The song database contains the following tables:
* authors
* authors_songs
* media_files
* media_files_songs
* song_books
* songs
* songs_songbooks
* songs_topics
* topics
**authors** Table
This table holds the names of all the authors. It has the following
columns:
* id
* first_name
* last_name
* display_name
**authors_songs Table**
This is a bridging table between the *authors* and *songs* tables, which
serves to create a many-to-many relationship between the two tables. It
has the following columns:
* author_id
* song_id
* author_type
**media_files Table**
* id
* file_name
* type
**song_books Table**
The *song_books* table holds a list of books that a congregation gets
their songs from, or old hymnals now no longer used. This table has the
following columns:
* id
* name
* publisher
**songs Table**
This table contains the songs, and each song has a list of attributes.
The *songs* table has the following columns:
* id
* title
* alternate_title
* lyrics
* verse_order
* copyright
* comments
* ccli_number
* theme_name
* search_title
* search_lyrics
**songs_songsbooks Table**
This is a mapping table between the *songs* and the *song_books* tables. It has the following columns:
* songbook_id
* song_id
* entry # The song number, like 120 or 550A
**songs_topics Table**
This is a bridging table between the *songs* and *topics* tables, which
serves to create a many-to-many relationship between the two tables. It
has the following columns:
* song_id
* topic_id
**topics Table**
The topics table holds a selection of topics that songs can cover. This
is useful when a worship leader wants to select songs with a certain
theme. This table has the following columns:
* id
* name
"""
session, metadata = init_db(url)
# Definition of the "authors" table
authors_table = Table(
'authors', metadata, Column('id', types.Integer(), primary_key=True),
Column('first_name', types.Unicode(128)),
Column('last_name', types.Unicode(128)),
Column('display_name', types.Unicode(255), index=True, nullable=False))
# Definition of the "media_files" table
media_files_table = Table(
'media_files', metadata, Column('id',
types.Integer(),
primary_key=True),
Column('song_id',
types.Integer(),
ForeignKey('songs.id'),
default=None),
Column('file_name', types.Unicode(255), nullable=False),
Column('type', types.Unicode(64), nullable=False, default='audio'),
Column('weight', types.Integer(), default=0))
# Definition of the "song_books" table
song_books_table = Table(
'song_books', metadata, Column('id', types.Integer(),
primary_key=True),
Column('name', types.Unicode(128), nullable=False),
Column('publisher', types.Unicode(128)))
# Definition of the "songs" table
songs_table = Table(
'songs', metadata, Column('id', types.Integer(), primary_key=True),
Column('title', types.Unicode(255), nullable=False),
Column('alternate_title', types.Unicode(255)),
Column('lyrics', types.UnicodeText, nullable=False),
Column('verse_order', types.Unicode(128)),
Column('copyright', types.Unicode(255)),
Column('comments', types.UnicodeText),
Column('ccli_number', types.Unicode(64)),
Column('theme_name', types.Unicode(128)),
Column('search_title', types.Unicode(255), index=True, nullable=False),
Column('search_lyrics', types.UnicodeText, nullable=False),
Column('create_date', types.DateTime(), default=func.now()),
Column('last_modified',
types.DateTime(),
default=func.now(),
onupdate=func.now()),
Column('temporary', types.Boolean(), default=False))
# Definition of the "topics" table
topics_table = Table(
'topics', metadata, Column('id', types.Integer(), primary_key=True),
Column('name', types.Unicode(128), index=True, nullable=False))
# Definition of the "authors_songs" table
authors_songs_table = Table(
'authors_songs', metadata,
Column('author_id',
types.Integer(),
ForeignKey('authors.id'),
primary_key=True),
Column('song_id',
types.Integer(),
ForeignKey('songs.id'),
primary_key=True),
Column('author_type',
types.Unicode(255),
primary_key=True,
nullable=False,
server_default=text('""')))
# Definition of the "songs_songbooks" table
songs_songbooks_table = Table(
'songs_songbooks', metadata,
Column('songbook_id',
types.Integer(),
ForeignKey('song_books.id'),
primary_key=True),
Column('song_id',
types.Integer(),
ForeignKey('songs.id'),
primary_key=True),
Column('entry', types.Unicode(255), primary_key=True, nullable=False))
# Definition of the "songs_topics" table
songs_topics_table = Table(
'songs_topics', metadata,
Column('song_id',
types.Integer(),
ForeignKey('songs.id'),
primary_key=True),
Column('topic_id',
types.Integer(),
ForeignKey('topics.id'),
primary_key=True))
mapper(Author,
authors_table,
properties={
'songs':
relation(Song, secondary=authors_songs_table, viewonly=True)
})
mapper(AuthorSong,
authors_songs_table,
properties={'author': relation(Author)})
mapper(SongBookEntry,
songs_songbooks_table,
properties={'songbook': relation(Book)})
mapper(Book, song_books_table)
mapper(MediaFile, media_files_table)
mapper(
Song,
songs_table,
properties={
# Use the authors_songs relation when you need access to the 'author_type' attribute
# or when creating new relations
'authors_songs':
relation(AuthorSong, cascade="all, delete-orphan"),
# Use lazy='joined' to always load authors when the song is fetched from the database (bug 1366198)
'authors':
relation(Author,
secondary=authors_songs_table,
viewonly=True,
lazy='joined'),
'media_files':
relation(MediaFile,
backref='songs',
order_by=media_files_table.c.weight),
'songbook_entries':
relation(SongBookEntry,
backref='song',
cascade='all, delete-orphan'),
'topics':
relation(Topic, backref='songs', secondary=songs_topics_table)
})
mapper(Topic, topics_table)
metadata.create_all(checkfirst=True)
return session
class SearchMany(Model):
_id = Column(types.Integer(), primary_key=True)
string = Column(types.String())
search_id = Column(types.Integer(), ForeignKey('search._id'))
class RouterTestModel(Base):
__tablename__ = "routertest"
id = Column(types.Integer(), default=3, primary_key=True)
text = Column(types.String(length=200))
class Owner(Base):
__tablename__ = "owners"
id = Column(types.Integer(), primary_key=True)
first_name = Column(types.Unicode(length=50))
last_name = Column(types.Unicode(length=50))
[1, 2, 3],
]
ONE_ROW_CONTENTS_DML = [
588,
datetime.datetime(2013, 10, 10, 11, 27, 16, tzinfo=timezone('UTC')),
'test', 40.76727216, False,
datetime.date(2013, 10, 10),
datetime.datetime(2013, 10, 10, 11, 27, 16),
datetime.time(11, 27, 16), 'test_bytes'
]
SAMPLE_COLUMNS = [
{
'name': 'integer',
'type': types.Integer(),
'nullable': True,
'default': None
},
{
'name': 'timestamp',
'type': types.TIMESTAMP(),
'nullable': True,
'default': None
},
{
'name': 'string',
'type': types.String(),
'nullable': True,
'default': None
},
class ModelWithJson(Base):
__tablename__ = "model_with_json"
id = Column(types.Integer(), Sequence("seq_id"), primary_key=True)
class Invoice(meta.BaseObject):
"""An invoice."""
__tablename__ = "invoice"
id = schema.Column(types.Integer(),
schema.Sequence("invoice_id_seq", optional=True),
primary_key=True,
autoincrement=True)
_number = schema.Column("number", types.Integer())
customer_id = schema.Column(types.Integer(),
schema.ForeignKey(Customer.id,
onupdate="CASCADE",
ondelete="CASCADE"),
nullable=False)
customer = orm.relationship(Customer, backref="invoices")
sent = schema.Column(types.Date())
payment_term = schema.Column(types.Integer(), nullable=False, default=30)
paid = schema.Column(types.Date())
note = schema.Column(types.UnicodeText())
@orm.reconstructor
def _add_acls(self):
account_id = self.customer.account_id
self.__acl__ = [(security.Allow, account_id, ("comment", "view"))]
if not self.sent:
self.__acl__.append(
(security.Allow, account_id, ("delete", "edit")))
if len(self.entries):
self.__acl__.append((security.Allow, account_id, "send"))
if self.sent and not self.paid:
self.__acl__.append((security.Allow, account_id, "mark-paid"))
@property
def due(self):
if self.sent:
return self.sent + datetime.timedelta(days=self.payment_term)
return None
def total(self, type="gross"):
assert type in ["gross", "net", "vat"]
gross = sum([entry.total for entry in self.entries])
if type == "gross":
return gross
vat = sum([v[1] for v in self.VAT()])
if type == "vat":
return vat
return gross + vat
def VAT(self):
totals = {}
for entry in self.entries:
if not entry.vat:
continue
current = entry.total
totals[entry.vat] = totals.get(entry.vat, 0) + current
for (vat, total) in totals.items():
totals[vat] = (totals[vat] * vat) / 100
return sorted(totals.items())
@synonym_for("_number")
@property
def number(self):
if not self._number:
return None
return "%s.%04d" % (self.customer.invoice_code, self._number)
def state(self):
if not self.sent:
return "unsend"
elif self.paid:
return "paid"
today = datetime.date.today()
due = self.sent + datetime.timedelta(days=self.payment_term)
if due < today:
return "overdue"
else:
return "pending"
def overdue(self):
if self.paid or not self.sent:
return None
today = datetime.date.today()
due = self.sent + datetime.timedelta(days=self.payment_term)
if due >= today:
return None
return (today - due).days
def send(self):
assert self.sent is None
self.sent = datetime.datetime.now()
self._number = self.customer.account.newInvoiceNumber()
from sqlalchemy import orm
import datetime
from sqlalchemy import schema, types
metadata = schema.MetaData()
def now():
return datetime.datetime.now()
page_table = schema.Table(
'page',
metadata,
schema.Column('id',
types.Integer(),
schema.Sequence('page_seq_id', optional=True),
primary_key=True),
schema.Column('content', types.Text(), nullable=False),
schema.Column('posted', types.DateTime(), default=now),
schema.Column('title', types.Unicode(255), default=u'Unititled Page'),
schema.Column('heading', types.Unicode(255)),
)
comment_table = schema.Table(
'comment',
metadata,
schema.Column('id',
types.Integer(),
schema.Sequence('page_seq_id', optional=True),
primary_key=True),
class User(db.Model, UserMixin, DeploymentMixin): #pylint: disable=no-init,too-few-public-methods
'''
User
'''
__tablename__ = 'users'
id = Column(types.Integer, autoincrement=True, primary_key=True) #pylint: disable=invalid-name
picture_id = Column(
types.String,
default=lambda: base64.urlsafe_b64encode(os.urandom(20))[0:-2])
has_picture = Column(types.Boolean, default=False)
first_name = Column(types.String,
info={
'label': lazy_gettext('First Name'),
})
last_name = Column(types.String,
info={
'label': lazy_gettext('Last Name'),
})
email = Column(EmailType,
nullable=False,
info={
'label': lazy_gettext('Email'),
})
password = Column(types.String,
nullable=False,
info={
'label': lazy_gettext('Password'),
})
active = Column(types.Boolean, nullable=False)
last_login_at = Column(types.DateTime())
current_login_at = Column(types.DateTime())
confirmed_at = Column(types.DateTime())
last_login_ip = Column(types.Text)
current_login_ip = Column(types.Text)
login_count = Column(types.Integer)
position = Column(types.String, info={
'label': lazy_gettext('Position'),
})
organization = Column(types.String,
info={
'label': lazy_gettext('Organization'),
})
organization_type = Column(
types.String,
info={
'label': lazy_gettext('Type of Organization'),
'description':
lazy_gettext('The type of organization you work for'),
'choices': [(k, v) for k, v in ORG_TYPES.iteritems()]
})
country = Column(CountryType, info={
'label': lazy_gettext('Country'),
})
city = Column(types.String, info={'label': lazy_gettext('City')})
latlng = Column(types.String,
info={
'label': lazy_gettext('Location'),
'description': lazy_gettext('Enter your location')
})
projects = Column(
types.Text,
info={
'label':
lazy_gettext('Projects'),
'description':
lazy_gettext(
'Add name and url or short description of any current work projects.'
)
})
tutorial_step = Column(types.Integer())
created_at = Column(types.DateTime(), default=datetime.datetime.now)
updated_at = Column(types.DateTime(),
default=datetime.datetime.now,
onupdate=datetime.datetime.now)
def is_admin(self):
return self.email in current_app.config.get('ADMIN_UI_USERS', [])
@property
def full_name(self):
return u"%s %s" % (self.first_name, self.last_name)
@property
def full_location(self):
loc = []
if self.city:
loc.append(self.city)
if self.country and self.country.code != 'ZZ':
loc.append(self.country.name)
return ", ".join(loc)
@property
def display_in_search(self):
'''
Determine whether user has filled out bare minimum to display in search
results.
Specifically, we want to make sure that the first and last name
are both non-NULL and non-blank.
'''
return bool(self.first_name and self.last_name)
@property
def picture_path(self):
'''
Path where picture would be found (hosted on S3).
'''
return "{}/static/pictures/{}/{}".format(
current_app.config['NOI_DEPLOY'], self.id, self.picture_id)
@property
def picture_url(self):
'''
Full path to picture.
'''
return 'https://s3.amazonaws.com/{bucket}/{path}'.format(
bucket=current_app.config['S3_BUCKET_NAME'],
path=self.picture_path)
def remove_picture(self):
conn = S3Connection(current_app.config['S3_ACCESS_KEY_ID'],
current_app.config['S3_SECRET_ACCESS_KEY'])
bucket = conn.get_bucket(current_app.config['S3_BUCKET_NAME'])
if bucket.get_key(self.picture_path):
bucket.delete_key(self.picture_path)
self.has_picture = False
def upload_picture(self, fileobj, mimetype):
'''
Upload the given file object with the given mime type to S3 and
mark the user as having a picture.
'''
conn = S3Connection(current_app.config['S3_ACCESS_KEY_ID'],
current_app.config['S3_SECRET_ACCESS_KEY'])
bucket = conn.get_bucket(current_app.config['S3_BUCKET_NAME'])
bucket.make_public(recursive=False)
k = bucket.new_key(self.picture_path)
k.set_metadata('Content-Type', mimetype)
k.set_contents_from_file(fileobj)
k.make_public()
self.has_picture = True
@property
def helpful_users(self, limit=10):
'''
Returns a list of (user, score) tuples with matching positive skills,
ordered by the most helpful (highest score) descending.
'''
my_skills = aliased(UserSkill, name='my_skills', adapt_on_names=True)
their_skills = aliased(UserSkill,
name='their_skills',
adapt_on_names=True)
return User.query_in_deployment().\
add_column(func.sum(their_skills.level - my_skills.level)).\
filter(their_skills.user_id != my_skills.user_id).\
filter(User.id == their_skills.user_id).\
filter(their_skills.name == my_skills.name).\
filter(my_skills.user_id == self.id).\
filter(my_skills.level == LEVELS['LEVEL_I_WANT_TO_LEARN']['score']).\
group_by(User).\
order_by(desc(func.sum(their_skills.level - my_skills.level))).\
limit(limit)
@property
def nearest_neighbors(self, limit=10):
'''
Returns a list of (user, score) tuples with the closest matching
skills. If they haven't answered the equivalent skill question, we
consider that a very big difference (12).
Order is closest to least close, which is an ascending score.
'''
my_skills = aliased(UserSkill, name='my_skills', adapt_on_names=True)
their_skills = aliased(UserSkill,
name='their_skills',
adapt_on_names=True)
# difference we assume for user that has not answered question
unanswered_difference = (LEVELS['LEVEL_I_CAN_DO_IT']['score'] -
LEVELS['LEVEL_I_WANT_TO_LEARN']['score']) * 2
return User.query_in_deployment().\
add_column(((len(self.skills) - func.count(func.distinct(their_skills.id))) *
unanswered_difference) + \
func.sum(func.abs(their_skills.level - my_skills.level))).\
filter(their_skills.user_id != my_skills.user_id).\
filter(User.id == their_skills.user_id).\
filter(their_skills.name == my_skills.name).\
filter(my_skills.user_id == self.id).\
group_by(User).\
order_by(((len(self.skills) - func.count(func.distinct(their_skills.id)))
* unanswered_difference) + \
func.sum(func.abs(their_skills.level - my_skills.level))).\
limit(limit)
@property
def has_fully_registered(self):
'''
Returns whether the user has fully completed the registration/signup
flow.
'''
return db.session.query(UserJoinedEvent).\
filter_by(user_id=self.id).\
first() is not None
def set_fully_registered(self):
'''
Marks the user as having fully completed the registration/signup
flow, if they haven't already.
'''
if self.has_fully_registered:
return
db.session.add(UserJoinedEvent.from_user(self))
def match(self, level, limit=10):
'''
Returns a list of UserSkillMatch objects, in descending order of number
of skills matched for each user.
'''
skills_to_learn = [
s.name for s in self.skills
if s.level == LEVELS['LEVEL_I_WANT_TO_LEARN']['score']
]
if skills_to_learn:
matched_users = User.query_in_deployment().\
add_column(func.string_agg(UserSkill.name, ',')).\
add_column(func.count(UserSkill.id)).\
filter(UserSkill.name.in_(skills_to_learn)).\
filter(User.id == UserSkill.user_id).\
filter(UserSkill.level == level).\
filter(UserSkill.user_id != self.id).\
group_by(User).\
order_by(func.count().desc()).\
limit(limit)
else:
matched_users = []
for user, question_ids_by_comma, count in matched_users:
yield UserSkillMatch(user, question_ids_by_comma.split(','))
def match_against(self, user):
'''
Returns a list of three-tuples in the format:
(<questionnaire id>, <count of matching questions>, <skill dict>, )
In descending order of <count of matching questions>.
<skill dict> is keyed by the skill level of the other user, with each
value being a set of questions they can answer at that level.
'''
skills = UserSkill.query.\
filter(UserSkill.user_id == user.id).\
filter(UserSkill.name.in_(
[s.name for s in
self.skills if s.level == LEVELS['LEVEL_I_WANT_TO_LEARN']['score']
])).all()
resp = {}
for skill in skills:
question = QUESTIONS_BY_ID[skill.name]
questionnaire_id = question['questionnaire']['id']
if questionnaire_id not in resp:
resp[questionnaire_id] = dict()
if skill.level not in resp[questionnaire_id]:
resp[questionnaire_id][skill.level] = set()
resp[questionnaire_id][skill.level].add(skill.name)
resp = [(qname,
sum([len(questions)
for questions in skill_levels.values()]), skill_levels)
for qname, skill_levels in resp.items()]
return sorted(resp, lambda a, b: a[1] - b[1], reverse=True)
def match_against_with_progress(self, user):
'''
Like match_against(), but also includes information about
areas of expertise the target user has that we don't match
on.
'''
progress = user.questionnaire_progress
matches = self.match_against(user)
match_areas = {}
for questionnaire_id, _, _ in matches:
match_areas[questionnaire_id] = True
for questionnaire_id, progress in progress.items():
if (questionnaire_id not in match_areas
and progress['answered'] > 0):
matches.append((questionnaire_id, 0, {}))
return matches
def match_against_with_progress_in_area(self, user, areaid):
'''
Return a tuple of (matched_skill_dict, unmatched_skill_dict)
for the given area.
matched_skill_dict contains information about skills that
the target user has which we want to learn, while
unmatched_skill_dict contains all other skills the target
user has.
Each dict is keyed by the skill level of the other user,
with each value being a set of questions they can answer at that
level.
'''
questionnaire = QUESTIONNAIRES_BY_ID[areaid]
matches = self.match_against(user)
matched_skill_dict = {}
matched_skills = {}
for questionnaire_id, _, skill_dict in matches:
if questionnaire_id == areaid:
matched_skill_dict = skill_dict
break
for question_ids in matched_skill_dict.values():
for question_id in question_ids:
matched_skills[question_id] = True
unmatched_skill_dict = {}
skill_levels = user.skill_levels
for topic in questionnaire.get('topics', []):
for question in topic['questions']:
qid = question['id']
if (qid in skill_levels and qid not in matched_skills):
level = skill_levels[qid]
if level not in unmatched_skill_dict:
unmatched_skill_dict[level] = []
unmatched_skill_dict[level].append(qid)
return (matched_skill_dict, unmatched_skill_dict)
@property
def questionnaire_progress(self):
'''
Return a dictionary mapping top-level skill area IDs (e.g.,
'opendata', 'prizes') to information about how many questions
the user has answered in that skill area.
'''
skill_levels = self.skill_levels
progress = {}
for questionnaire in QUESTIONNAIRES:
topic_progress = {'answered': 0, 'total': 0}
progress[questionnaire['id']] = topic_progress
for topic in questionnaire.get('topics', []):
for question in topic['questions']:
topic_progress['total'] += 1
if question['id'] in skill_levels:
topic_progress['answered'] += 1
return progress
@property
def skill_levels(self):
'''
Dictionary of this user's entered skills, keyed by the id of the skill.
'''
return dict([(skill.name, skill.level) for skill in self.skills])
@property
def connections(self):
'''
Count the number of distinct email addresses this person has sent or
received messages from in the deployment.
'''
sent = db.session.query(func.count(func.distinct(Email.to_user_id))).\
filter(Email.to_user_id != self.id).\
filter(Email.from_user_id == self.id).first()[0]
received = db.session.query(func.count(func.distinct(Email.from_user_id))).\
filter(Email.from_user_id != self.id).\
filter(Email.to_user_id == self.id).first()[0]
return sent + received
def set_skill(self, skill_name, skill_level):
'''
Set the level of a single skill by name.
'''
if skill_name not in QUESTIONS_BY_ID:
return
try:
if int(skill_level) not in VALID_SKILL_LEVELS:
return
except ValueError:
return
for skill in self.skills:
if skill_name == skill.name:
skill.level = skill_level
db.session.add(skill)
return
db.session.add(
UserSkill(user_id=self.id, name=skill_name, level=skill_level))
def email_connect(self, users):
'''
Indicate that this user has opened an email window with this list of
users as recipients.
'''
event = ConnectionEvent()
for user in users:
event.emails.append(Email(from_user_id=self.id,
to_user_id=user.id))
db.session.add(event)
return event
roles = orm.relationship('Role',
secondary='role_users',
backref=orm.backref('users', lazy='dynamic'))
expertise_domains = orm.relationship('UserExpertiseDomain',
cascade='all,delete-orphan',
backref='user')
languages = orm.relationship('UserLanguage',
cascade='all,delete-orphan',
backref='user')
skills = orm.relationship('UserSkill',
cascade='all,delete-orphan',
backref='user')
@classmethod
def get_most_complete_profiles(cls, limit=10):
'''
Obtain a list of most complete profiles, as (User, score) tuples.
'''
return User.query_in_deployment().\
add_column(func.count(UserSkill.id)).\
filter(User.id == UserSkill.user_id).\
group_by(User).\
order_by(func.count(UserSkill.id).desc()).\
limit(limit)
@classmethod
def get_most_connected_profiles(cls, limit=10):
'''
Obtain a list of most connected profiles, as descending (User, score)
tuples.
'''
count_of_unique_emails = func.count(
func.distinct(
cast(Email.to_user_id, String) + '-' +
cast(Email.from_user_id, String)))
return User.query_in_deployment().\
add_column(count_of_unique_emails).\
filter((User.id == Email.from_user_id) | (User.id == Email.to_user_id)).\
group_by(User).\
order_by(count_of_unique_emails.desc()).\
limit(limit)
@hybrid_property
def expertise_domain_names(self):
'''
Convenient list of expertise domains by name.
'''
return [ed.name for ed in self.expertise_domains]
@expertise_domain_names.setter
def _expertise_domains_setter(self, values):
'''
Update expertise domains in bulk. Values are array of names.
'''
# Only add new expertise
for val in values:
if val not in self.expertise_domain_names:
db.session.add(UserExpertiseDomain(name=val, user_id=self.id))
# delete expertise no longer found
expertise_to_remove = []
for exp in self.expertise_domains:
if exp.name not in values:
expertise_to_remove.append(exp)
for exp in expertise_to_remove:
self.expertise_domains.remove(exp)
def get_area_scores(self):
skill_levels = self.skill_levels
result = {}
NO_ANSWER = -999
for questionnaire in QUESTIONNAIRES:
if not questionnaire['questions']:
continue
max_score = NO_ANSWER
answers_with_score = {}
for question in questionnaire['questions']:
if question['id'] in skill_levels:
score = skill_levels[question['id']]
if score > max_score:
max_score = score
if score not in answers_with_score:
answers_with_score[score] = 0
answers_with_score[score] += 1
reported_max_score = None
if max_score != NO_ANSWER:
reported_max_score = max_score
result[questionnaire['id']] = {
'skills': scores_to_skills(answers_with_score),
'max_score': reported_max_score
}
return result
@hybrid_property
def locales(self):
'''
Convenient list of locales for this user.
'''
return [l.locale for l in self.languages]
@locales.setter
def _languages_setter(self, values):
'''
Update locales for this user in bulk. Values are an array of language
codes.
'''
locale_codes = [l.language for l in self.locales]
# only add new languages
for val in values:
if val not in locale_codes:
db.session.add(UserLanguage(locale=val, user_id=self.id))
# delete languages no longer found
languages_to_remove = []
for lan in self.languages:
if lan.locale.language not in values:
languages_to_remove.append(lan)
for lan in languages_to_remove:
self.languages.remove(lan)
__table_args__ = (UniqueConstraint('deployment', 'email'), )
def test_should_integer_convert_int():
assert get_field(types.Integer()).type == graphene.Int
class Post(BaseTable):
__tablename__ = 'posts'
id = Column(types.Integer(10), primary_key=True)
title = Column(types.Unicode(255))
content = Column(types.UnicodeText)
summary = Column(types.UnicodeText, nullable=True)
slug = Column(types.Unicode(255), unique=True)
category_id = Column(types.Integer(10), ForeignKey('categories.id'))
category = relation('Category')
user_id = Column(types.Integer(10), ForeignKey('users.id'))
user = relation('User')
posted = Column(types.DateTime, default=datetime.datetime.now)
comments = relation('Comment', order_by=['id'])
comments_count = Column(types.Integer(10), nullable=True, default=0)
def __init__(self, title, category_id, content, user_id, **kwargs):
self.title = title
self.category_id = category_id
self.content = content
self.user_id = user_id
self.slug = slug(kwargs.get('slug', ''))
if not self.slug:
# Automatically generate a slug (URL) from the title if there was
# none supplied.
self.slug = slug(self.title)
self.summary = kwargs.get('summary', '')
def __repr__(self):
return '<Post: "%s" by %s>' % (self.title, self.user.name)
@classmethod
def all(self):
q = session.query(Post).order_by(Post.posted.desc())
return q.all()
@classmethod
def by_category(self, id):
q = session.query(Post).filter(Post.category_id == id)
return q
@classmethod
def by_id(self, id):
q = session.query(Post).filter(Post.id == id)
return q
@classmethod
def by_slug(self, slug):
q = session.query(Post).filter(Post.slug == slug)
return q
@classmethod
def by_slug_category(self, slug, category):
q = session.query(Post).filter(
and_(Post.slug == slug, Category.slug == category))
return q
@classmethod
def by_user(self, id):
q = session.query(Post).filter(Post.user_id == id)
return q
def test_should_primary_integer_convert_id():
assert get_field(types.Integer(), primary_key=True).type == graphene.NonNull(graphene.ID)
from six import text_type
import vdm.sqlalchemy
import meta
import core
import domain_object
__all__ = [
'system_info_revision_table', 'system_info_table', 'SystemInfo',
'SystemInfoRevision', 'get_system_info', 'set_system_info'
]
system_info_table = Table(
'system_info',
meta.metadata,
Column('id', types.Integer(), primary_key=True, nullable=False),
Column('key', types.Unicode(100), unique=True, nullable=False),
Column('value', types.UnicodeText),
Column('state', types.UnicodeText, default=core.State.ACTIVE),
)
system_info_revision_table = core.make_revisioned_table(system_info_table)
class SystemInfo(vdm.sqlalchemy.RevisionedObjectMixin,
core.StatefulObjectMixin, domain_object.DomainObject):
def __init__(self, key, value):
super(SystemInfo, self).__init__()
self.key = key
class TestModel(Model):
__tablename__ = 'test'
_id = Column(types.Integer(), primary_key=True)
query_class = None
class BaseEngineSpec: # pylint: disable=too-many-public-methods
"""Abstract class for database engine specific configurations"""
engine = "base" # str as defined in sqlalchemy.engine.engine
engine_aliases: Optional[Tuple[str]] = None
engine_name: Optional[
str] = None # used for user messages, overridden in child classes
_date_trunc_functions: Dict[str, str] = {}
_time_grain_expressions: Dict[Optional[str], str] = {}
column_type_mappings: Tuple[Tuple[Pattern[str],
Union[TypeEngine, Callable[[Match[str]],
TypeEngine]],
GenericDataType, ],
..., ] = (
(
re.compile(r"^smallint",
re.IGNORECASE),
types.SmallInteger(),
GenericDataType.NUMERIC,
),
(
re.compile(r"^integer", re.IGNORECASE),
types.Integer(),
GenericDataType.NUMERIC,
),
(
re.compile(r"^bigint", re.IGNORECASE),
types.BigInteger(),
GenericDataType.NUMERIC,
),
(
re.compile(r"^decimal", re.IGNORECASE),
types.Numeric(),
GenericDataType.NUMERIC,
),
(
re.compile(r"^numeric", re.IGNORECASE),
types.Numeric(),
GenericDataType.NUMERIC,
),
(
re.compile(r"^real", re.IGNORECASE),
types.REAL,
GenericDataType.NUMERIC,
),
(
re.compile(r"^smallserial",
re.IGNORECASE),
types.SmallInteger(),
GenericDataType.NUMERIC,
),
(
re.compile(r"^serial", re.IGNORECASE),
types.Integer(),
GenericDataType.NUMERIC,
),
(
re.compile(r"^bigserial",
re.IGNORECASE),
types.BigInteger(),
GenericDataType.NUMERIC,
),
(
re.compile(r"^string", re.IGNORECASE),
types.String(),
utils.GenericDataType.STRING,
),
(
re.compile(r"^N((VAR)?CHAR|TEXT)",
re.IGNORECASE),
UnicodeText(),
utils.GenericDataType.STRING,
),
(
re.compile(
r"^((VAR)?CHAR|TEXT|STRING)",
re.IGNORECASE),
String(),
utils.GenericDataType.STRING,
),
(
re.compile(r"^date", re.IGNORECASE),
types.Date(),
GenericDataType.TEMPORAL,
),
(
re.compile(r"^timestamp",
re.IGNORECASE),
types.TIMESTAMP(),
GenericDataType.TEMPORAL,
),
(
re.compile(r"^interval",
re.IGNORECASE),
types.Interval(),
GenericDataType.TEMPORAL,
),
(
re.compile(r"^time", re.IGNORECASE),
types.Time(),
GenericDataType.TEMPORAL,
),
(
re.compile(r"^boolean", re.IGNORECASE),
types.Boolean(),
GenericDataType.BOOLEAN,
),
)
time_groupby_inline = False
limit_method = LimitMethod.FORCE_LIMIT
time_secondary_columns = False
allows_joins = True
allows_subqueries = True
allows_alias_in_select = True
allows_alias_in_orderby = True
allows_sql_comments = True
force_column_alias_quotes = False
arraysize = 0
max_column_name_length = 0
try_remove_schema_from_table_name = True # pylint: disable=invalid-name
run_multiple_statements_as_one = False
@classmethod
def get_dbapi_exception_mapping(
cls) -> Dict[Type[Exception], Type[Exception]]:
"""
Each engine can implement and converge its own specific exceptions into
Superset DBAPI exceptions
Note: On python 3.9 this method can be changed to a classmethod property
without the need of implementing a metaclass type
:return: A map of driver specific exception to superset custom exceptions
"""
return {}
@classmethod
def get_dbapi_mapped_exception(cls, exception: Exception) -> Exception:
"""
Get a superset custom DBAPI exception from the driver specific exception.
Override if the engine needs to perform extra changes to the exception, for
example change the exception message or implement custom more complex logic
:param exception: The driver specific exception
:return: Superset custom DBAPI exception
"""
new_exception = cls.get_dbapi_exception_mapping().get(type(exception))
if not new_exception:
return exception
return new_exception(str(exception))
@classmethod
def get_allow_cost_estimate(cls, extra: Dict[str, Any]) -> bool:
return False
@classmethod
def get_engine(
cls,
database: "Database",
schema: Optional[str] = None,
source: Optional[str] = None,
) -> Engine:
user_name = utils.get_username()
return database.get_sqla_engine(schema=schema,
nullpool=True,
user_name=user_name,
source=source)
@classmethod
def get_timestamp_expr(
cls,
col: ColumnClause,
pdf: Optional[str],
time_grain: Optional[str],
type_: Optional[str] = None,
) -> TimestampExpression:
"""
Construct a TimestampExpression to be used in a SQLAlchemy query.
:param col: Target column for the TimestampExpression
:param pdf: date format (seconds or milliseconds)
:param time_grain: time grain, e.g. P1Y for 1 year
:param type_: the source column type
:return: TimestampExpression object
"""
if time_grain:
time_expr = cls.get_time_grain_expressions().get(time_grain)
if not time_expr:
raise NotImplementedError(
f"No grain spec for {time_grain} for database {cls.engine}"
)
if type_ and "{func}" in time_expr:
date_trunc_function = cls._date_trunc_functions.get(type_)
if date_trunc_function:
time_expr = time_expr.replace("{func}",
date_trunc_function)
if type_ and "{type}" in time_expr:
date_trunc_function = cls._date_trunc_functions.get(type_)
if date_trunc_function:
time_expr = time_expr.replace("{type}", type_)
else:
time_expr = "{col}"
# if epoch, translate to DATE using db specific conf
if pdf == "epoch_s":
time_expr = time_expr.replace("{col}", cls.epoch_to_dttm())
elif pdf == "epoch_ms":
time_expr = time_expr.replace("{col}", cls.epoch_ms_to_dttm())
return TimestampExpression(time_expr, col, type_=DateTime)
@classmethod
def get_time_grains(cls) -> Tuple[TimeGrain, ...]:
"""
Generate a tuple of supported time grains.
:return: All time grains supported by the engine
"""
ret_list = []
time_grains = builtin_time_grains.copy()
time_grains.update(config["TIME_GRAIN_ADDONS"])
for duration, func in cls.get_time_grain_expressions().items():
if duration in time_grains:
name = time_grains[duration]
ret_list.append(TimeGrain(name, _(name), func, duration))
return tuple(ret_list)
@classmethod
def get_time_grain_expressions(cls) -> Dict[Optional[str], str]:
"""
Return a dict of all supported time grains including any potential added grains
but excluding any potentially disabled grains in the config file.
:return: All time grain expressions supported by the engine
"""
# TODO: use @memoize decorator or similar to avoid recomputation on every call
time_grain_expressions = cls._time_grain_expressions.copy()
grain_addon_expressions = config["TIME_GRAIN_ADDON_EXPRESSIONS"]
time_grain_expressions.update(
grain_addon_expressions.get(cls.engine, {}))
denylist: List[str] = config["TIME_GRAIN_DENYLIST"]
for key in denylist:
time_grain_expressions.pop(key)
return time_grain_expressions
@classmethod
def make_select_compatible(
cls, groupby_exprs: Dict[str, ColumnElement],
select_exprs: List[ColumnElement]) -> List[ColumnElement]:
"""
Some databases will just return the group-by field into the select, but don't
allow the group-by field to be put into the select list.
:param groupby_exprs: mapping between column name and column object
:param select_exprs: all columns in the select clause
:return: columns to be included in the final select clause
"""
return select_exprs
@classmethod
def fetch_data(cls,
cursor: Any,
limit: Optional[int] = None) -> List[Tuple[Any, ...]]:
"""
:param cursor: Cursor instance
:param limit: Maximum number of rows to be returned by the cursor
:return: Result of query
"""
if cls.arraysize:
cursor.arraysize = cls.arraysize
try:
if cls.limit_method == LimitMethod.FETCH_MANY and limit:
return cursor.fetchmany(limit)
return cursor.fetchall()
except Exception as ex:
raise cls.get_dbapi_mapped_exception(ex)
@classmethod
def expand_data(
cls, columns: List[Dict[Any, Any]], data: List[Dict[Any, Any]]
) -> Tuple[List[Dict[Any, Any]], List[Dict[Any, Any]], List[Dict[Any,
Any]]]:
"""
Some engines support expanding nested fields. See implementation in Presto
spec for details.
:param columns: columns selected in the query
:param data: original data set
:return: list of all columns(selected columns and their nested fields),
expanded data set, listed of nested fields
"""
return columns, data, []
@classmethod
def alter_new_orm_column(cls, orm_col: "TableColumn") -> None:
"""Allow altering default column attributes when first detected/added
For instance special column like `__time` for Druid can be
set to is_dttm=True. Note that this only gets called when new
columns are detected/created"""
# TODO: Fix circular import caused by importing TableColumn
@classmethod
def epoch_to_dttm(cls) -> str:
"""
SQL expression that converts epoch (seconds) to datetime that can be used in a
query. The reference column should be denoted as `{col}` in the return
expression, e.g. "FROM_UNIXTIME({col})"
:return: SQL Expression
"""
raise NotImplementedError()
@classmethod
def epoch_ms_to_dttm(cls) -> str:
"""
SQL expression that converts epoch (milliseconds) to datetime that can be used
in a query.
:return: SQL Expression
"""
return cls.epoch_to_dttm().replace("{col}", "({col}/1000)")
@classmethod
def get_datatype(cls, type_code: Any) -> Optional[str]:
"""
Change column type code from cursor description to string representation.
:param type_code: Type code from cursor description
:return: String representation of type code
"""
if isinstance(type_code, str) and type_code != "":
return type_code.upper()
return None
@classmethod
def normalize_indexes(
cls, indexes: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
"""
Normalizes indexes for more consistency across db engines
noop by default
:param indexes: Raw indexes as returned by SQLAlchemy
:return: cleaner, more aligned index definition
"""
return indexes
@classmethod
def extra_table_metadata(cls, database: "Database", table_name: str,
schema_name: str) -> Dict[str, Any]:
"""
Returns engine-specific table metadata
:param database: Database instance
:param table_name: Table name
:param schema_name: Schema name
:return: Engine-specific table metadata
"""
# TODO: Fix circular import caused by importing Database
return {}
@classmethod
def apply_limit_to_sql(cls, sql: str, limit: int,
database: "Database") -> str:
"""
Alters the SQL statement to apply a LIMIT clause
:param sql: SQL query
:param limit: Maximum number of rows to be returned by the query
:param database: Database instance
:return: SQL query with limit clause
"""
# TODO: Fix circular import caused by importing Database
if cls.limit_method == LimitMethod.WRAP_SQL:
sql = sql.strip("\t\n ;")
qry = (select("*").select_from(
TextAsFrom(text(sql), ["*"]).alias("inner_qry")).limit(limit))
return database.compile_sqla_query(qry)
if cls.limit_method == LimitMethod.FORCE_LIMIT:
parsed_query = sql_parse.ParsedQuery(sql)
sql = parsed_query.set_or_update_query_limit(limit)
return sql
@classmethod
def get_limit_from_sql(cls, sql: str) -> Optional[int]:
"""
Extract limit from SQL query
:param sql: SQL query
:return: Value of limit clause in query
"""
parsed_query = sql_parse.ParsedQuery(sql)
return parsed_query.limit
@classmethod
def set_or_update_query_limit(cls, sql: str, limit: int) -> str:
"""
Create a query based on original query but with new limit clause
:param sql: SQL query
:param limit: New limit to insert/replace into query
:return: Query with new limit
"""
parsed_query = sql_parse.ParsedQuery(sql)
return parsed_query.set_or_update_query_limit(limit)
@staticmethod
def csv_to_df(**kwargs: Any) -> pd.DataFrame:
"""Read csv into Pandas DataFrame
:param kwargs: params to be passed to DataFrame.read_csv
:return: Pandas DataFrame containing data from csv
"""
kwargs["encoding"] = "utf-8"
kwargs["iterator"] = True
chunks = pd.read_csv(**kwargs)
df = pd.concat(chunk for chunk in chunks)
return df
@classmethod
def df_to_sql(cls, df: pd.DataFrame, **kwargs: Any) -> None:
"""Upload data from a Pandas DataFrame to a database. For
regular engines this calls the DataFrame.to_sql() method. Can be
overridden for engines that don't work well with to_sql(), e.g.
BigQuery.
:param df: Dataframe with data to be uploaded
:param kwargs: kwargs to be passed to to_sql() method
"""
df.to_sql(**kwargs)
@classmethod
def create_table_from_csv( # pylint: disable=too-many-arguments
cls,
filename: str,
table: Table,
database: "Database",
csv_to_df_kwargs: Dict[str, Any],
df_to_sql_kwargs: Dict[str, Any],
) -> None:
"""
Create table from contents of a csv. Note: this method does not create
metadata for the table.
"""
df = cls.csv_to_df(filepath_or_buffer=filename, **csv_to_df_kwargs)
engine = cls.get_engine(database)
if table.schema:
# only add schema when it is preset and non empty
df_to_sql_kwargs["schema"] = table.schema
if engine.dialect.supports_multivalues_insert:
df_to_sql_kwargs["method"] = "multi"
cls.df_to_sql(df=df, con=engine, **df_to_sql_kwargs)
@classmethod
def convert_dttm(cls, target_type: str, dttm: datetime) -> Optional[str]:
"""
Convert Python datetime object to a SQL expression
:param target_type: The target type of expression
:param dttm: The datetime object
:return: The SQL expression
"""
return None
@classmethod
def create_table_from_excel( # pylint: disable=too-many-arguments
cls,
filename: str,
table: Table,
database: "Database",
excel_to_df_kwargs: Dict[str, Any],
df_to_sql_kwargs: Dict[str, Any],
) -> None:
"""
Create table from contents of a excel. Note: this method does not create
metadata for the table.
"""
df = pd.read_excel(io=filename, **excel_to_df_kwargs)
engine = cls.get_engine(database)
if table.schema:
# only add schema when it is preset and non empty
df_to_sql_kwargs["schema"] = table.schema
if engine.dialect.supports_multivalues_insert:
df_to_sql_kwargs["method"] = "multi"
cls.df_to_sql(df=df, con=engine, **df_to_sql_kwargs)
@classmethod
def get_all_datasource_names(
cls, database: "Database",
datasource_type: str) -> List[utils.DatasourceName]:
"""Returns a list of all tables or views in database.
:param database: Database instance
:param datasource_type: Datasource_type can be 'table' or 'view'
:return: List of all datasources in database or schema
"""
# TODO: Fix circular import caused by importing Database
schemas = database.get_all_schema_names(
cache=database.schema_cache_enabled,
cache_timeout=database.schema_cache_timeout,
force=True,
)
all_datasources: List[utils.DatasourceName] = []
for schema in schemas:
if datasource_type == "table":
all_datasources += database.get_all_table_names_in_schema(
schema=schema,
force=True,
cache=database.table_cache_enabled,
cache_timeout=database.table_cache_timeout,
)
elif datasource_type == "view":
all_datasources += database.get_all_view_names_in_schema(
schema=schema,
force=True,
cache=database.table_cache_enabled,
cache_timeout=database.table_cache_timeout,
)
else:
raise Exception(
f"Unsupported datasource_type: {datasource_type}")
return all_datasources
@classmethod
def handle_cursor(cls, cursor: Any, query: Query,
session: Session) -> None:
"""Handle a live cursor between the execute and fetchall calls
The flow works without this method doing anything, but it allows
for handling the cursor and updating progress information in the
query object"""
# TODO: Fix circular import error caused by importing sql_lab.Query
@classmethod
def extract_error_message(cls, ex: Exception) -> str:
return f"{cls.engine} error: {cls._extract_error_message(ex)}"
@classmethod
def _extract_error_message(cls, ex: Exception) -> str:
"""Extract error message for queries"""
return utils.error_msg_from_exception(ex)
@classmethod
def extract_errors(cls, ex: Exception) -> List[Dict[str, Any]]:
return [
dataclasses.asdict(
SupersetError(
error_type=SupersetErrorType.GENERIC_DB_ENGINE_ERROR,
message=cls._extract_error_message(ex),
level=ErrorLevel.ERROR,
extra={"engine_name": cls.engine_name},
))
]
@classmethod
def adjust_database_uri(cls, uri: URL,
selected_schema: Optional[str]) -> None:
"""
Mutate the database component of the SQLAlchemy URI.
The URI here represents the URI as entered when saving the database,
``selected_schema`` is the schema currently active presumably in
the SQL Lab dropdown. Based on that, for some database engine,
we can return a new altered URI that connects straight to the
active schema, meaning the users won't have to prefix the object
names by the schema name.
Some databases engines have 2 level of namespacing: database and
schema (postgres, oracle, mssql, ...)
For those it's probably better to not alter the database
component of the URI with the schema name, it won't work.
Some database drivers like presto accept '{catalog}/{schema}' in
the database component of the URL, that can be handled here.
"""
@classmethod
def patch(cls) -> None:
"""
TODO: Improve docstring and refactor implementation in Hive
"""
@classmethod
def get_schema_names(cls, inspector: Inspector) -> List[str]:
"""
Get all schemas from database
:param inspector: SqlAlchemy inspector
:return: All schemas in the database
"""
return sorted(inspector.get_schema_names())
@classmethod
def get_table_names(cls, database: "Database", inspector: Inspector,
schema: Optional[str]) -> List[str]:
"""
Get all tables from schema
:param inspector: SqlAlchemy inspector
:param schema: Schema to inspect. If omitted, uses default schema for database
:return: All tables in schema
"""
tables = inspector.get_table_names(schema)
if schema and cls.try_remove_schema_from_table_name:
tables = [re.sub(f"^{schema}\\.", "", table) for table in tables]
return sorted(tables)
@classmethod
def get_view_names(cls, database: "Database", inspector: Inspector,
schema: Optional[str]) -> List[str]:
"""
Get all views from schema
:param inspector: SqlAlchemy inspector
:param schema: Schema name. If omitted, uses default schema for database
:return: All views in schema
"""
views = inspector.get_view_names(schema)
if schema and cls.try_remove_schema_from_table_name:
views = [re.sub(f"^{schema}\\.", "", view) for view in views]
return sorted(views)
@classmethod
def get_table_comment(cls, inspector: Inspector, table_name: str,
schema: Optional[str]) -> Optional[str]:
"""
Get comment of table from a given schema and table
:param inspector: SqlAlchemy Inspector instance
:param table_name: Table name
:param schema: Schema name. If omitted, uses default schema for database
:return: comment of table
"""
comment = None
try:
comment = inspector.get_table_comment(table_name, schema)
comment = comment.get("text") if isinstance(comment,
dict) else None
except NotImplementedError:
# It's expected that some dialects don't implement the comment method
pass
except Exception as ex: # pylint: disable=broad-except
logger.error("Unexpected error while fetching table comment")
logger.exception(ex)
return comment
@classmethod
def get_columns(cls, inspector: Inspector, table_name: str,
schema: Optional[str]) -> List[Dict[str, Any]]:
"""
Get all columns from a given schema and table
:param inspector: SqlAlchemy Inspector instance
:param table_name: Table name
:param schema: Schema name. If omitted, uses default schema for database
:return: All columns in table
"""
return inspector.get_columns(table_name, schema)
@classmethod
def where_latest_partition( # pylint: disable=too-many-arguments
cls,
table_name: str,
schema: Optional[str],
database: "Database",
query: Select,
columns: Optional[List[Dict[str, str]]] = None,
) -> Optional[Select]:
"""
Add a where clause to a query to reference only the most recent partition
:param table_name: Table name
:param schema: Schema name
:param database: Database instance
:param query: SqlAlchemy query
:param columns: List of TableColumns
:return: SqlAlchemy query with additional where clause referencing latest
partition
"""
# TODO: Fix circular import caused by importing Database, TableColumn
return None
@classmethod
def _get_fields(cls, cols: List[Dict[str, Any]]) -> List[Any]:
return [column(c["name"]) for c in cols]
@classmethod
def select_star( # pylint: disable=too-many-arguments,too-many-locals
cls,
database: "Database",
table_name: str,
engine: Engine,
schema: Optional[str] = None,
limit: int = 100,
show_cols: bool = False,
indent: bool = True,
latest_partition: bool = True,
cols: Optional[List[Dict[str, Any]]] = None,
) -> str:
"""
Generate a "SELECT * from [schema.]table_name" query with appropriate limit.
WARNING: expects only unquoted table and schema names.
:param database: Database instance
:param table_name: Table name, unquoted
:param engine: SqlALchemy Engine instance
:param schema: Schema, unquoted
:param limit: limit to impose on query
:param show_cols: Show columns in query; otherwise use "*"
:param indent: Add indentation to query
:param latest_partition: Only query latest partition
:param cols: Columns to include in query
:return: SQL query
"""
fields: Union[str, List[Any]] = "*"
cols = cols or []
if (show_cols or latest_partition) and not cols:
cols = database.get_columns(table_name, schema)
if show_cols:
fields = cls._get_fields(cols)
quote = engine.dialect.identifier_preparer.quote
if schema:
full_table_name = quote(schema) + "." + quote(table_name)
else:
full_table_name = quote(table_name)
qry = select(fields).select_from(text(full_table_name))
if limit:
qry = qry.limit(limit)
if latest_partition:
partition_query = cls.where_latest_partition(table_name,
schema,
database,
qry,
columns=cols)
if partition_query is not None:
qry = partition_query
sql = database.compile_sqla_query(qry)
if indent:
sql = sqlparse.format(sql, reindent=True)
return sql
@classmethod
def estimate_statement_cost(
cls,
statement: str,
cursor: Any,
) -> Dict[str, Any]:
"""
Generate a SQL query that estimates the cost of a given statement.
:param statement: A single SQL statement
:param cursor: Cursor instance
:return: Dictionary with different costs
"""
raise Exception("Database does not support cost estimation")
@classmethod
def query_cost_formatter(
cls, raw_cost: List[Dict[str, Any]]) -> List[Dict[str, str]]:
"""
Format cost estimate.
:param raw_cost: Raw estimate from `estimate_query_cost`
:return: Human readable cost estimate
"""
raise Exception("Database does not support cost estimation")
@classmethod
def process_statement(cls, statement: str, database: "Database",
user_name: str) -> str:
"""
Process a SQL statement by stripping and mutating it.
:param statement: A single SQL statement
:param database: Database instance
:param username: Effective username
:return: Dictionary with different costs
"""
parsed_query = ParsedQuery(statement)
sql = parsed_query.stripped()
sql_query_mutator = config["SQL_QUERY_MUTATOR"]
if sql_query_mutator:
sql = sql_query_mutator(sql, user_name, security_manager, database)
return sql
@classmethod
def estimate_query_cost(
cls,
database: "Database",
schema: str,
sql: str,
source: Optional[str] = None) -> List[Dict[str, Any]]:
"""
Estimate the cost of a multiple statement SQL query.
:param database: Database instance
:param schema: Database schema
:param sql: SQL query with possibly multiple statements
:param source: Source of the query (eg, "sql_lab")
"""
extra = database.get_extra() or {}
if not cls.get_allow_cost_estimate(extra):
raise Exception("Database does not support cost estimation")
user_name = g.user.username if g.user else None
parsed_query = sql_parse.ParsedQuery(sql)
statements = parsed_query.get_statements()
engine = cls.get_engine(database, schema=schema, source=source)
costs = []
with closing(engine.raw_connection()) as conn:
cursor = conn.cursor()
for statement in statements:
processed_statement = cls.process_statement(
statement, database, user_name)
costs.append(
cls.estimate_statement_cost(processed_statement, cursor))
return costs
@classmethod
def modify_url_for_impersonation(cls, url: URL, impersonate_user: bool,
username: Optional[str]) -> None:
"""
Modify the SQL Alchemy URL object with the user to impersonate if applicable.
:param url: SQLAlchemy URL object
:param impersonate_user: Flag indicating if impersonation is enabled
:param username: Effective username
"""
if impersonate_user and username is not None:
url.username = username
@classmethod
def update_impersonation_config(
cls,
connect_args: Dict[str, Any],
uri: str,
username: Optional[str],
) -> None:
"""
Update a configuration dictionary
that can set the correct properties for impersonating users
:param connect_args: config to be updated
:param uri: URI
:param impersonate_user: Flag indicating if impersonation is enabled
:param username: Effective username
:return: None
"""
@classmethod
def execute(cls, cursor: Any, query: str, **kwargs: Any) -> None:
"""
Execute a SQL query
:param cursor: Cursor instance
:param query: Query to execute
:param kwargs: kwargs to be passed to cursor.execute()
:return:
"""
if not cls.allows_sql_comments:
query = sql_parse.strip_comments_from_sql(query)
if cls.arraysize:
cursor.arraysize = cls.arraysize
try:
cursor.execute(query)
except Exception as ex:
raise cls.get_dbapi_mapped_exception(ex)
@classmethod
def make_label_compatible(cls, label: str) -> Union[str, quoted_name]:
"""
Conditionally mutate and/or quote a sqlalchemy expression label. If
force_column_alias_quotes is set to True, return the label as a
sqlalchemy.sql.elements.quoted_name object to ensure that the select query
and query results have same case. Otherwise return the mutated label as a
regular string. If maxmimum supported column name length is exceeded,
generate a truncated label by calling truncate_label().
:param label: expected expression label/alias
:return: conditionally mutated label supported by the db engine
"""
label_mutated = cls._mutate_label(label)
if (cls.max_column_name_length
and len(label_mutated) > cls.max_column_name_length):
label_mutated = cls._truncate_label(label)
if cls.force_column_alias_quotes:
label_mutated = quoted_name(label_mutated, True)
return label_mutated
@classmethod
def get_sqla_column_type(
cls,
column_type: Optional[str],
column_type_mappings: Tuple[Tuple[Pattern[str],
Union[TypeEngine,
Callable[[Match[str]],
TypeEngine]],
GenericDataType, ],
..., ] = column_type_mappings,
) -> Union[Tuple[TypeEngine, GenericDataType], None]:
"""
Return a sqlalchemy native column type that corresponds to the column type
defined in the data source (return None to use default type inferred by
SQLAlchemy). Override `column_type_mappings` for specific needs
(see MSSQL for example of NCHAR/NVARCHAR handling).
:param column_type: Column type returned by inspector
:return: SqlAlchemy column type
"""
if not column_type:
return None
for regex, sqla_type, generic_type in column_type_mappings:
match = regex.match(column_type)
if match:
if callable(sqla_type):
return sqla_type(match), generic_type
return sqla_type, generic_type
return None
@staticmethod
def _mutate_label(label: str) -> str:
"""
Most engines support mixed case aliases that can include numbers
and special characters, like commas, parentheses etc. For engines that
have restrictions on what types of aliases are supported, this method
can be overridden to ensure that labels conform to the engine's
limitations. Mutated labels should be deterministic (input label A always
yields output label X) and unique (input labels A and B don't yield the same
output label X).
:param label: Preferred expression label
:return: Conditionally mutated label
"""
return label
@classmethod
def _truncate_label(cls, label: str) -> str:
"""
In the case that a label exceeds the max length supported by the engine,
this method is used to construct a deterministic and unique label based on
the original label. By default this returns an md5 hash of the original label,
conditionally truncated if the length of the hash exceeds the max column length
of the engine.
:param label: Expected expression label
:return: Truncated label
"""
label = hashlib.md5(label.encode("utf-8")).hexdigest()
# truncate hash if it exceeds max length
if cls.max_column_name_length and len(
label) > cls.max_column_name_length:
label = label[:cls.max_column_name_length]
return label
@classmethod
def column_datatype_to_string(cls, sqla_column_type: TypeEngine,
dialect: Dialect) -> str:
"""
Convert sqlalchemy column type to string representation.
By default removes collation and character encoding info to avoid unnecessarily
long datatypes.
:param sqla_column_type: SqlAlchemy column type
:param dialect: Sqlalchemy dialect
:return: Compiled column type
"""
sqla_column_type = sqla_column_type.copy()
if hasattr(sqla_column_type, "collation"):
sqla_column_type.collation = None
if hasattr(sqla_column_type, "charset"):
sqla_column_type.charset = None
return sqla_column_type.compile(dialect=dialect).upper()
@classmethod
def get_function_names(cls, database: "Database") -> List[str]:
"""
Get a list of function names that are able to be called on the database.
Used for SQL Lab autocomplete.
:param database: The database to get functions for
:return: A list of function names useable in the database
"""
return []
@staticmethod
def pyodbc_rows_to_tuples(data: List[Any]) -> List[Tuple[Any, ...]]:
"""
Convert pyodbc.Row objects from `fetch_data` to tuples.
:param data: List of tuples or pyodbc.Row objects
:return: List of tuples
"""
if data and type(data[0]).__name__ == "Row":
data = [tuple(row) for row in data]
return data
@staticmethod
def mutate_db_for_connection_test(database: "Database") -> None:
"""
Some databases require passing additional parameters for validating database
connections. This method makes it possible to mutate the database instance prior
to testing if a connection is ok.
:param database: instance to be mutated
"""
return None
@staticmethod
def get_extra_params(database: "Database") -> Dict[str, Any]:
"""
Some databases require adding elements to connection parameters,
like passing certificates to `extra`. This can be done here.
:param database: database instance from which to extract extras
:raises CertificateException: If certificate is not valid/unparseable
"""
extra: Dict[str, Any] = {}
if database.extra:
try:
extra = json.loads(database.extra)
except json.JSONDecodeError as ex:
logger.error(ex)
raise ex
return extra
@classmethod
def is_readonly_query(cls, parsed_query: ParsedQuery) -> bool:
"""Pessimistic readonly, 100% sure statement won't mutate anything"""
return (parsed_query.is_select() or parsed_query.is_explain()
or parsed_query.is_show())
@classmethod
def get_column_spec(
cls,
native_type: Optional[str],
source: utils.ColumnTypeSource = utils.ColumnTypeSource.GET_TABLE,
column_type_mappings: Tuple[Tuple[Pattern[str],
Union[TypeEngine,
Callable[[Match[str]],
TypeEngine]],
GenericDataType, ],
..., ] = column_type_mappings,
) -> Union[ColumnSpec, None]:
"""
Converts native database type to sqlalchemy column type.
:param native_type: Native database typee
:param source: Type coming from the database table or cursor description
:return: ColumnSpec object
"""
column_type = None
if (cls.get_sqla_column_type(native_type,
column_type_mappings=column_type_mappings)
is not None):
column_type, generic_type = cls.get_sqla_column_type( # type: ignore
native_type,
column_type_mappings=column_type_mappings)
is_dttm = generic_type == GenericDataType.TEMPORAL
if column_type:
return ColumnSpec(sqla_type=column_type,
generic_type=generic_type,
is_dttm=is_dttm)
return None
def mixed_in_column(cls):
return Column(types.Integer(),
primary_key=cls.__name__.endswith('Primary'))
class Order(Model):
__tablename__ = 'orders'
_id = Column(types.Integer(), primary_key=True)
status = Column(OrderStatus.db_type(), default=OrderStatus.pending)
side = Column(OrderSide.db_type(), default=OrderSide.buy)
def foreign_id(cls):
return Column(types.Integer(),
ForeignKey(MixedInTablePrimary.mixed_in_column),
primary_key=cls.__name__.endswith('Primary'))
class BaseAutoGen(object):
_id = Column(types.Integer(), primary_key=True)
def local_primary_key(cls):
return Column(types.Integer(), primary_key=True)
class AC(Model):
a_id = Column(types.Integer(), ForeignKey('a._id'), primary_key=True)
c_id = Column(types.Integer(), ForeignKey('c._id'), primary_key=True)
key = Column(types.String())
c = orm.relationship('C', lazy=False)
class GeoLite2CityBlocksIpv4(Base, ModelDictMixin):
"""GeoLite2 City Blocks database model
References:
* `GeoIP2 City and Country CSV Databases <https://dev.maxmind.com/geoip/geoip2/geoip2-city-country-csv-databases/>`_
Args:
id (str):
* The primary key
network (str):
* This is the IPv4 network in CIDR format such as “2.21.92.0/29”
geoname_id (int):
* A unique identifier for the network's location as specified by
GeoNames. This ID can be used to look up the location
information in the Location file.
registered_country_geoname_id (int):
* The registered country is the country in which the ISP has
registered the network. This column contains a unique
identifier for the network's registered country as specified
by GeoNames. This ID can be used to look up the location
information in the Location file.
represented_country_geoname_id (int):
* The represented country is the country which is represented by
users of the IP address. For instance, the country represented
by an overseas military base. This column contains a unique
identifier for the network's registered country as specified
by GeoNames. This ID can be used to look up the location
information in the Location file.
is_anonymous_proxy (boolean):
* Deprecated. Please see our GeoIP2 Anonymous IP database to
determine whether the IP address is used by an anonymizing
service.
is_satellite_provider (boolean):
* Deprecated. Please see our GeoIP2 Anonymous IP database.
postal_code (str):
* A postal code close to the user's location. We return the first
3 characters for Canadian postal codes. We return the
first 2-4 characters (outward code) for postal codes in the
United Kingdom.
latitude (float):
* The approximate latitude of the postal code, city, subdivision
or country associated with the IP address.
longitude (float):
* The approximate longitude of the postal code, city, subdivision
or country associated with the IP address.
accuracy_radius (int):
* The approximate accuracy radius, in kilometers, around the
latitude and longitude for the geographical entity
(country, subdivision, city or postal code) associated with
the IP address. We have a 67% confidence that the location of
the end-user falls within the area defined by the accuracy
radius and the latitude and longitude coordinates.
""" # noqa
__tablename__ = "geolite2_city_blocks_ipv4"
id = Column(types.String(16),
index=True,
primary_key=True,
default=generate_uuid)
network = Column(
types.String(18),
nullable=False,
)
geoname_id = Column(types.Integer())
registered_country_geoname_id = Column(types.Integer())
represented_country_geoname_id = Column(types.Integer())
_is_anonymous_proxy = Column(types.Boolean())
_is_satellite_provider = Column(types.Boolean())
postal_code = Column(types.String(9))
_latitude = Column(types.Float())
_longitude = Column(types.Float())
accuracy_radius = Column(types.Integer())
@hybrid_property
def latitude(self):
return self._latitude
@latitude.setter
def latitude(self, value):
value = value or float()
self._latitude = float(value)
@hybrid_property
def longitude(self):
return self._longitude
@longitude.setter
def longitude(self, value):
value = value or float()
self._longitude = float(value)
@hybrid_property
def is_anonymous_proxy(self):
return self._is_anonymous_proxy
@is_anonymous_proxy.setter
def is_anonymous_proxy(self, value):
self._is_anonymous_proxy = bool(int(value))
@hybrid_property
def is_satellite_provider(self):
return self._is_satellite_provider
@is_satellite_provider.setter
def is_satellite_provider(self, value):
self._is_satellite_provider = bool(int(value))
def __repr__(self):
return 'GeoLite2CityBlocksIpv4(id={!r})'.format(self.id)
class SearchOne(Model):
_id = Column(types.Integer(), primary_key=True)
string = Column(types.String())
class Entry(EntryJsonSerializer, BaseMixin, VersionMinxin):
__table_args__ = {'mysql_charset': 'utf8', 'mysql_engine': 'InnoDB'}
TagClass = None
category_id = Column(Integer, nullable=False, index=True, doc=u'分类ID')
author_id = Column(Integer, nullable=False, index=True, doc=u'作者ID')
published_id = Column(Integer,
nullable=False,
index=True,
default=0,
doc=u'发布人ID')
entry_type = Column(ChoiceType(ENTRY_TYPE_CHOICES, types.Integer()),
nullable=False,
default=EntryType.article,
index=True,
doc=u'类型')
title = Column(String(200), nullable=False, doc=u'标题')
# content = Column(Text, doc=u'内容')
body = db.Column(db.Text)
body_html = db.Column(db.Text)
summary = Column(String(200), doc=u'摘要')
tag_id_list = Column(DenormalizedText, doc=u'标签ID列表')
tag_list = Column(DenormalizedText, doc=u'标签文本列表')
logo = Column(String(200), doc=u'图标地址')
thumbnail_id_list = db.Column(DenormalizedText,
nullable=True,
default=set(),
doc=u'缩略图ID列表')
thumbs_width = Column(Integer, nullable=False, default=256, doc=u'缩略图宽度')
thumbs_height = Column(Integer, nullable=False, default=256, doc=u'缩略图高度')
source_title = Column(String(200), doc=u'来源标题')
source_url = Column(String(200), doc=u'来源地址')
source_type = Column(ChoiceType(ENTRY_SOURCE_TYPE_CHOICES,
types.Integer()),
nullable=False,
default=EntrySourceType.original,
doc=u'来源类型：0-原创、1-翻译、2-转帖、3-专稿')
slug = Column(String(200), unique=True, index=True, doc=u'固定地址')
recommend = Column(ChoiceType(ENTRY_RECOMMEND_CHOICES, types.Integer()),
nullable=False,
default=EntryRecommend.general,
doc=u'推荐级别：0-默认（未推荐）、1-不错、2-良好、3-精华')
comment_status = Column(ChoiceType(ENTRY_COMMENT_STATUS_CHOICES,
types.Integer()),
nullable=False,
default=EntryCommentStatus.open,
doc=u'评论状态：0-允许、1-关闭')
view_status = Column(ChoiceType(ENTRY_VIEW_STATUS_CHOICES,
types.Integer()),
nullable=False,
default=EntryViewStatus.open,
doc=u'显示状态：0-公开、1-私有、2-仅作者可见、3-隐藏')
entry_status = Column(ChoiceType(ENTRY_STATUS_CHOICES, types.Integer()),
nullable=False,
default=EntryStatus.draft,
doc=u'内容状态：0-草稿、1-待审、2-已删除、3-发布、4-已接受')
be_modified = Column(Integer, nullable=False, default=0, doc=u'被别人修改了')
num_favorites = Column(Integer, nullable=False, default=0, doc=u'收藏数')
num_retweet = Column(Integer, nullable=False, default=0, doc=u'转发数')
num_views = Column(Integer, nullable=False, default=0, doc=u'阅读数')
num_comments = Column(Integer, nullable=False, default=0, doc=u'评论/回复数')
num_uninterested = Column(Integer, nullable=False, default=0, doc=u'不感兴趣数')
# 置顶周期、推首页周期、
#文章排序：（置顶周期） 标题样式：
on_top = Column(Integer, nullable=False, default=0, doc=u'是否置顶')
on_top_period = Column(Integer,
nullable=True,
default=14,
doc=u'置顶周期: 缺省2周')
on_portal = Column(Integer, nullable=False, default=0, doc=u'是否首页显示')
on_portal_period = Column(Integer,
nullable=True,
default=14,
doc=u'首页显示周期: 缺省2周')
created_ip = Column(Integer, nullable=False, doc=u'创建时的IP')
upward_time = Column(DateTime, nullable=False, doc=u'贴子互动/变更时间')
published_time = Column(DateTime, doc=u'发布时间')
ranking = Column(Integer,
nullable=False,
default=0,
index=True,
doc=u'热门度、权重')
# 最后回复ID
last_comment_id = Column(Integer, doc=u'最后回复ID')
# 文档文件ID列表
document_id_list = Column(DenormalizedText, default='[]', doc=u'文档文件ID列表')
# 普通文件ID列表
general_id_list = Column(DenormalizedText, default='[]', doc=u'普通文件ID列表')
# 图片文件ID列表
picture_id_list = Column(DenormalizedText, default='[]', doc=u'图片文件ID列表')
@property
def last_comment(self):
return Comment.get_by_id(self.last_comment_id)
@property
def hot_support_comment(self):
return Comment.get_hot_support_comment(self.id)
@property
def author(self):
return UserService.get_brief_by_id(self.author_id)
@property
def author_name(self):
return UserService.get_brief_by_id(self.author_id).nickname
@property
def category(self):
print 'self.category_id: ', self.category_id
return Category.get_by_id(self.category_id)
@property
def category_name(self):
return Category.get_by_id(self.category_id).category_name
@property
def tags(self):
tag_list = []
for _id in self.tag_id_list:
# todo
tag_list.append(self.TagClass.get_by_id(_id))
return tag_list
@property
def permissions(self):
if g.user:
return {
'can_edit': self.author.id == g.user.id
or g.user.is_supervisor,
'can_del': self.author.id == g.user.id or g.user.is_supervisor
}
else:
return {'can_edit': 0, 'can_del': 0}
def __unicode__(self):
return self.id
def __repr__(self):
return self.id
@classmethod
def get_by_id(cls, id_):
#global simple_cache
#key = 'entry_%d' % id
#print key
#if key in simple_cache:
# query = loads(simple_cache[key], scoped_session=db.session)
#else:
# todo cache
return cls.query.get(int(id_))
#serialized = dumps(query)
#simple_cache[key] = serialized
# return query
@staticmethod
def get_entries_total():
_total = Entry.query.filter(
and_(Entry.entry_status == EntryStatus.published,
Entry.view_status == EntryViewStatus.open)).count()
_pages = int(ceil(_total / float(PER_PAGE)))
return _total, _pages
@staticmethod
def get_latest_entries_by_type(entry_type, page, size):
return Entry.query.filter(
and_(Entry.entry_type == entry_type,
Entry.entry_status == EntryStatus.published,
Entry.view_status == EntryViewStatus.open)).order_by(
'%s %s' % ('id', Order.DESC)).limit(size).offset(
size * (page - 1))
@staticmethod
def get_all_entries():
return Entry.query.filter(
and_(Entry.entry_status == EntryStatus.published,
Entry.view_status == EntryViewStatus.open)).order_by(
'%s %s' % ('id', Order.DESC)).all()
@staticmethod
def get_latest_entries(page, size):
return Entry.query.filter(
and_(Entry.entry_status == EntryStatus.published,
Entry.view_status == EntryViewStatus.open)).order_by(
'%s %s' % ('id', Order.DESC)).limit(size).offset(
size * (page - 1))
@staticmethod
def get_by_category(category_id, page, size):
return Entry.query.filter(
and_(Entry.on_portal == True,
Entry.entry_status == EntryStatus.published,
Entry.view_status == EntryViewStatus.open,
Entry.category_id == category_id)).order_by(
'on_top DESC, %s %s' %
('upward_time', Order.DESC)).limit(size).offset(
size * (page - 1))
@staticmethod
def get_by_entry_type(entry_type, page, size):
return Entry.query.filter(
and_(Entry.on_portal == True,
Entry.entry_status == EntryStatus.published,
Entry.view_status == EntryViewStatus.open,
Entry.entry_type == entry_type)).order_by(
'on_top DESC, %s %s' %
('upward_time', Order.DESC)).limit(size).offset(
size * (page - 1))
@staticmethod
def get_latest_portal_entries(page, size):
return Entry.query.filter(
and_(Entry.on_portal == True,
Entry.entry_status == EntryStatus.published,
Entry.view_status == EntryViewStatus.open)).order_by(
'on_top DESC, %s %s' %
('upward_time', Order.DESC)).limit(size).offset(
size * (page - 1))
@staticmethod
def get_hot_entries(page, size):
return Entry.query.filter(
and_(Entry.entry_status == EntryStatus.published,
Entry.view_status == EntryViewStatus.open)).order_by(
'%s %s' %
('num_comments', Order.DESC)).limit(size).offset(
size * (page - 1))
@staticmethod
def get_page_by_tag(tag, query, sort, order, page):
q = Entry.query.join(TagEntryRelation,
TagEntryRelation.entry_id == Entry.id).filter(
TagEntryRelation.tag_id == tag.id)
if q is None:
abort(404)
_total = q.count()
_list = q.order_by('-entry.id').limit(PER_PAGE).offset(
PER_PAGE * (page - 1)).all()
#_list = db.session.query(Entry).from_statement("SELECT * FROM entry WHERE category_id =:category_id").params(category_id=category.id).all()
_pages = int(ceil(_total / float(PER_PAGE)))
return _total, _pages, _list
@classmethod
def get_page_by_category(cls, category, tag, query, sort, order, page):
# /栏目/search?q=mysql&catalog=7
# /栏目/list?show=reply
# /search?q=[python]+mysql+using+sqlalchemy
# 查询规则：
#
# 根据 category.current_level 决定从Entry的哪一个category_id查询，一个3级
# category1_id、category2_id、category3_id、
#q = None
q = cls.query.filter_by(category_id=category)
#if tag:
# if category.current_level == 1:
# q = Entry.query.join(TagEntryRelation, TagEntryRelation.entry_id == Entry.id).filter(
# and_(Entry.category1_id == category.id, TagEntryRelation.tag_id == tag.id))
#
# if category.current_level == 2:
# q = Entry.query.join(TagEntryRelation, TagEntryRelation.entry_id == Entry.id).filter(
# and_(Entry.category2_id == category.id, TagEntryRelation.tag_id == tag.id))
#
# if category.current_level == 3:
# q = Entry.query.join(TagEntryRelation, TagEntryRelation.entry_id == Entry.id).filter(
# and_(Entry.category3_id == category.id, TagEntryRelation.tag_id == tag.id))
#else:
#if category.current_level == 1:
# q = Entry.query.filter_by(category1_id=category.id)
#
#if category.current_level == 2:
# q = Entry.query.filter_by(category2_id=category.id)
#
#if category.current_level == 3:
# q = Entry.query.filter_by(category3_id=category.id)
if q is None:
abort(404)
_total = q.count()
#print '123:%s %s' % (sort, order)
_list = q.order_by('%s %s' % (sort, order)).limit(PER_PAGE).offset(
PER_PAGE * (page - 1)).all()
#_list = db.session.query(Entry).from_statement("SELECT * FROM entry WHERE category_id =:category_id").params(category_id=category.id).all()
_pages = int(ceil(_total / float(PER_PAGE)))
entry_list = []
for e in _list:
entry_list.append(e.to_dict())
return _total, _pages, entry_list
@classmethod
def get_page_by_category_id(cls,
category,
sort,
order,
page,
page_size=20):
# /栏目/search?q=mysql&catalog=7
# /栏目/list?show=reply
# /search?q=[python]+mysql+using+sqlalchemy
# 查询规则：
#
q = cls.query.filter_by(category_id=category)
if q is None:
abort(404)
_total = q.count()
_list = q.order_by('%s %s' % (sort, order)).limit(page_size).offset(
page_size * (page - 1)).all()
_pages = int(ceil(_total / float(PER_PAGE)))
entry_list = []
for e in _list:
entry_list.append(e.to_dict())
return _total, _pages, entry_list
@classmethod
def get_page_by_category_and_entryType(cls, category, tag, query, sort,
order, page, entryType):
# /栏目/search?q=mysql&catalog=7
# /栏目/list?show=reply
# /search?q=[python]+mysql+using+sqlalchemy
# 查询规则：
#
# 根据 category.current_level 决定从Entry的哪一个category_id查询，一个3级
# category1_id、category2_id、category3_id、
q = None
if tag:
if category.current_level == 1:
q = Entry.query.join(
TagEntryRelation,
TagEntryRelation.entry_id == Entry.id).filter(
and_(Entry.category1_id == category.id,
TagEntryRelation.tag_id == tag.id))
if category.current_level == 2:
q = Entry.query.join(
TagEntryRelation,
TagEntryRelation.entry_id == Entry.id).filter(
and_(Entry.category2_id == category.id,
TagEntryRelation.tag_id == tag.id))
if category.current_level == 3:
q = Entry.query.join(
TagEntryRelation,
TagEntryRelation.entry_id == Entry.id).filter(
and_(Entry.category3_id == category.id,
TagEntryRelation.tag_id == tag.id))
else:
if category.current_level == 1:
q = Entry.query.filter_by(category1_id=category.id,
entry_type=entryType)
if category.current_level == 2:
q = Entry.query.filter_by(category2_id=category.id,
entry_type=entryType)
if category.current_level == 3:
q = Entry.query.filter_by(category3_id=category.id,
entry_type=entryType)
if q is None:
abort(404)
_total = q.count()
#print '123:%s %s' % (sort, order)
_list = q.order_by('on_top DESC, %s %s' %
(sort, order)).limit(PER_PAGE).offset(PER_PAGE *
(page - 1))
#_list = db.session.query(Entry).from_statement("SELECT * FROM entry WHERE category_id =:category_id").params(category_id=category.id).all()
_pages = int(ceil(_total / float(PER_PAGE)))
return _total, _pages, _list
@staticmethod
def get_page_by_entryType(sort, order, page, entryType):
# 根据类型获取文章列表
q = None
q = Entry.query.filter_by(entry_type=entryType)
if q is None:
abort(404)
print "entrytype============================================================================="
print q.count()
print "entrytype============================================================================="
_total = q.count()
_list = q.order_by('on_top DESC, %s %s' %
(sort, order)).limit(PER_PAGE).offset(PER_PAGE *
(page - 1))
_pages = int(ceil(_total / float(PER_PAGE)))
return _total, _pages, _list
@staticmethod
def get_entry_list_by_tag_id(tag_id,
page=1,
page_size=PER_PAGE,
sort=None,
order=None):
q = Entry.query.join(TagEntryRelation,
TagEntryRelation.entry_id == Entry.id).filter(
TagEntryRelation.tag_id == tag_id)
_total = q.count()
_list = q.order_by('-entry.id').limit(PER_PAGE).offset(PER_PAGE *
(page - 1))
#_list = db.session.query(Entry).from_statement("SELECT * FROM entry WHERE category_id =:category_id").params(category_id=category.id).all()
_pages = int(ceil(_total / float(PER_PAGE)))
return _total, _pages, _list
@staticmethod
def get_page_by_user(user_id, tag, entry_type, query, sort, order, page):
q = None
if entry_type:
q = Entry.query.filter(
and_(Entry.author_id == user_id,
Entry.entry_type == entry_type))
else:
q = Entry.query.filter(and_(Entry.author_id == user_id))
if q is None:
abort(404)
_total = q.count()
_list = q.order_by('-entry.id').limit(PER_PAGE).offset(PER_PAGE *
(page - 1))
_pages = int(ceil(_total / float(PER_PAGE)))
return _total, _pages, _list
@staticmethod
def add_or_update(category, entry, is_draft, is_auto_save_img=False):
# 保存数据
is_new = False
if entry.id is None:
# 是新建的数据
is_new = True
# todo wqq:如果是管理员权限才能允许表单递交author_id
#if not g.user.is_supervisor:
# entry.author_id = g.user.id
entry.author_id = 1
entry.created_ip = get_remote_ip()
# entry.created_time = sys_now()
# entry.updated_time = entry.created_time
entry.upward_time = sys_now()
entry.category_id = category.id
entry.entry_type = category.entry_type
#if is_auto_save_img:
# # 是否自动保存图片
# entry.content = auto_save_img(entry.content, app.config['SITE_DOMAIN'], PHOTOS_RELATIVE_PATH)
# 只有管理员可以设置是否推首页、置顶
#if not g.user.is_supervisor:
# entry.on_top = 0
# todo 先默认是推到首页的，以后考虑可以配置
entry.on_portal = 1
if is_draft:
entry.entry_status = EntryStatus.draft
else:
# 低权限的文章要审核
#if category.need_review and (not g.user.is_supervisor):
#print '低权限的文章要审核:', app.config['SAFE_POST_START_TIME'], app.config['SAFE_POST_END_TIME'], check_in_rank(app.config['SAFE_POST_START_TIME'], app.config['SAFE_POST_END_TIME'])
#if (not g.user.is_editor) and (
# not check_in_rank(app.config['SAFE_POST_START_TIME'], app.config['SAFE_POST_END_TIME'])):
# # 需要审核
# entry.entry_status = EntryStatus.pending
#else:
# entry.entry_status = EntryStatus.published
# entry.published_id = g.user.id
#
# if not entry.published_time:
# entry.published_time = sys_now()
#
#if gfw.check(entry.content) or gfw.check(entry.title):
# # 需要审核
# entry.entry_status = EntryStatus.pending
entry.entry_status = EntryStatus.pending
# 处理摘要
if not entry.summary:
# 自动产生摘要
entry.summary = extract_summary(entry.content)
# 开始处理tags
old_tags_list = []
if not is_new:
# old_tags_list = json_decode(entry.tags)
entry.updated_time = sys_now()
new_tags_list = request.form.getlist('m_tags')
#entry.tags = json_encode(new_tags_list)
if is_new:
db.session.add(entry)
db.session.flush()
if entry.id is None:
print '出错了'
db.session.rollback()
return False
# # 要删除的标签
# del_tags_list = [val for val in old_tags_list if val not in new_tags_list]
#
# # 要添加的标签
# add_tags_list = [val for val in new_tags_list if val not in old_tags_list]
#
# if del_tags_list:
# # 删除 tag_entry_relation （标签的文章关联）
# TagEntryRelationService.del_relation(entry.id, del_tags_list)
#
# # 更新 tag_category_relation （标签的分类计数）
# TagCategoryRelationService.del_relation(category, del_tags_list)
#
# if add_tags_list:
# # 添加 tag_entry_relation （标签的文章关联）
# TagEntryRelationService.add_relation(entry.id, add_tags_list)
#
# # 标签内容数 + 1
# TagService.inc_entry_count(add_tags_list)
#
# # 更新 tag_category_relation （标签的分类关联、计数）
# TagCategoryRelationService.add_relation(category, add_tags_list)
if not entry.slug:
# 文章URL自动生成规则
# 文章标题: {title_name}
# 文章ID: {title_id}
# 栏目名: {category_name}
# 栏目固定地址: {category_slug}
# 时间戳: {timestamp}
# 时间: {time}
# 日期: {date}
entry.slug = category.entry_url_rule.replace('{title_name}', '%s' % entry.title) \
.replace('{title_id}', '%d' % entry.id) \
.replace('{category_name}', '%s' % category.category_name) \
.replace('{category_slug}', category.slug) \
.replace('{timestamp}', '%s' % str(int(time()))) \
.replace('{time}', format_time_with_path(int(time()))) \
.replace('{date}', format_date_with_path(int(time()))).replace('//', '/')
if entry.slug[0] == '/':
entry.slug = entry.slug[1:]
db.session.commit()
#print '保存成功了'
return True
@staticmethod
def inc_num_comments(entry_id, num):
entry = Entry.get_by_id(entry_id)
entry.num_comments = Entry.__table__.c.num_comments + num
entry.ranking = entry.num_views + (10 * entry.num_useful) + (10 * entry.num_favorites) + (
10 * entry.num_retweet) \
+ (5 * entry.num_oppositions) + (7 * entry.num_comments) + (5 * entry.num_useless) + (
7 * entry.num_supports)
db.session.commit()
app.config['content_updated_time'] = sys_now()
@staticmethod
def calc_ranking(entry, ext_num=0):
entry.ranking = entry.num_views + (10 * entry.num_useful) + (10 * entry.num_favorites) \
+ (10 * entry.num_retweet) + (5 * entry.num_oppositions) + (7 * entry.num_comments) \
+ (-5 * entry.num_useless) + (7 * entry.num_thanks) + (-7 * entry.num_uninterested) + ext_num
@staticmethod
def inc_num_views(entry_id, num):
entry = Entry.get_by_id(entry_id)
entry.num_views = Entry.__table__.c.num_views + num
Entry.calc_ranking(entry)
db.session.commit()
@staticmethod
def favorites(user_id, entry_id, title, tags, description):
# 收藏
# 取出,还是客户端传人？
interact = EntryUserInteractService.get_by_entry_and_user_id(
entry_id, user_id)
if interact and interact.has_favorites:
return True
if not interact:
interact = EntryUserInteract()
interact.user_id = user_id
interact.entry_id = entry_id
interact.has_favorites = 1
EntryUserInteractService.add_or_update(interact)
entry = Entry.get_by_id(entry_id)
if not entry:
return False
favorites_id = FavoritesService.add_favorites(user_id, entry_id,
entry.author_id, title,
tags, description,
entry.entry_type)
if not favorites_id:
return False
entry.num_favorites = Entry.__table__.c.num_favorites + 1
Entry.calc_ranking(entry, 8)
db.session.commit()
return True
@staticmethod
def useful(entry_id, user_id):
# 有用数
interact = EntryUserInteractService.get_by_entry_and_user_id(
entry_id, user_id)
if interact and interact.has_useful:
return True
if not interact:
interact = EntryUserInteract()
interact.user_id = user_id
interact.entry_id = entry_id
interact.has_useful = 1
EntryUserInteractService.add_or_update(interact)
entry = Entry.get_by_id(entry_id)
entry.num_useful = Entry.__table__.c.num_useful + 1
Entry.calc_ranking(entry)
db.session.commit()
return True
@staticmethod
def useless(entry_id, user_id):
# 无用数
interact = EntryUserInteractService.get_by_entry_and_user_id(
entry_id, user_id)
if interact and interact.has_useless:
return True
if not interact:
interact = EntryUserInteract()
interact.user_id = user_id
interact.entry_id = entry_id
interact.has_useless = 1
EntryUserInteractService.add_or_update(interact)
entry = Entry.get_by_id(entry_id)
entry.num_useless = Entry.__table__.c.num_useless + 1
Entry.calc_ranking(entry)
db.session.commit()
return True
@staticmethod
def supports(entry_id, user_id):
# 支持
interact = EntryUserInteractService.get_by_entry_and_user_id(
entry_id, user_id)
if interact and interact.has_supports:
return True
if not interact:
interact = EntryUserInteract()
interact.user_id = user_id
interact.entry_id = entry_id
interact.has_supports = 1
EntryUserInteractService.add_or_update(interact)
entry = Entry.get_by_id(entry_id)
entry.num_supports = Entry.__table__.c.num_supports + 1
Entry.calc_ranking(entry)
db.session.commit()
return True
@staticmethod
def oppositions(entry_id, user_id):
# 反对
interact = EntryUserInteractService.get_by_entry_and_user_id(
entry_id, user_id)
if interact and interact.has_oppositions:
return True
if not interact:
interact = EntryUserInteract()
interact.user_id = user_id
interact.entry_id = entry_id
interact.has_oppositions = 1
EntryUserInteractService.add_or_update(interact)
entry = Entry.get_by_id(entry_id)
entry.num_oppositions = Entry.__table__.c.num_oppositions + 1
Entry.calc_ranking(entry)
db.session.commit()
return True
@staticmethod
def scores(entry_id, user_id, scores):
# 评分
entry = Entry.get_by_id(entry_id)
entry.scores = (Entry.__table__.c.scores + scores) / 2
Entry.calc_ranking(entry + 8)
db.session.commit()
return True
@staticmethod
def difficulty(entry_id, user_id, difficulty):
# 难易度
entry = Entry.get_by_id(entry_id)
entry.difficulty = (Entry.__table__.c.difficulty + difficulty) / 2
Entry.calc_ranking(entry + 8)
db.session.commit()
return True
@staticmethod
def on_portal(entry_id, user_id, enable):
# 推首页
entry = Entry.get_by_id(entry_id)
entry.on_portal = enable
db.session.commit()
return True
@staticmethod
def on_top(entry_id, user_id, enable):
# 置顶
entry = Entry.get_by_id(entry_id)
entry.on_top = enable
db.session.commit()
return True
@staticmethod
def publish(entry_id, user_id):
# 审核后发布
entry = Entry.get_by_id(entry_id)
entry.entry_status = EntryStatus.published
db.session.commit()
return True
@staticmethod
def hide(entry_id, user_id):
# 隐藏
entry = Entry.get_by_id(entry_id)
entry.view_status = EntryViewStatus.hide
db.session.commit()
return True
@staticmethod
def open(entry_id, user_id):
# 打开文章
entry = Entry.get_by_id(entry_id)
entry.view_status = EntryViewStatus.open
db.session.commit()
return True
@staticmethod
def close_comment(entry_id, user_id):
# 关闭评论
entry = Entry.get_by_id(entry_id)
entry.comment_status = EntryCommentStatusType.close
db.session.commit()
return True
@staticmethod
def open_comment(entry_id, user_id):
# 打开评论
entry = Entry.get_by_id(entry_id)
entry.comment_status = EntryCommentStatusType.open
db.session.commit()
return True
@staticmethod
def retweet(entry_id, user_id):
# 转发次数
interact = EntryUserInteractService.get_by_entry_and_user_id(
entry_id, user_id)
if interact and interact.has_retweet:
return True
if not interact:
interact = EntryUserInteract()
interact.user_id = user_id
interact.entry_id = entry_id
interact.has_retweet = 1
EntryUserInteractService.add_or_update(interact)
entry = Entry.get_by_id(entry_id)
entry.num_retweet = Entry.__table__.c.num_retweet + 1
Entry.calc_ranking(entry)
db.session.commit()
return True
@staticmethod
def complain(entry_id, user_id):
"""
投诉
:param entry_id:
:param user_id:
:return:
"""
#todo
return True
@staticmethod
def get_weekly_entries(category):
# 获取7天内的内容
q = None
if category.current_level == 1:
q = Entry.query.filter(
and_(Entry.category1_id == category.id,
Entry.published_time >= day_before_str(sys_now(), 7)))
if category.current_level == 2:
q = Entry.query.filter(
and_(Entry.category2_id == category.id,
Entry.published_time >= day_before_str(sys_now(), 7)))
if category.current_level == 3:
q = Entry.query.filter(
and_(Entry.category3_id == category.id,
Entry.published_time >= day_before_str(sys_now(), 7)))
if q is None:
return []
_list = q.order_by('-entry.ranking').limit(PER_PAGE)
return _list
@staticmethod
def get_monthly_entries(category):
# 获取30天内的内容
q = None
if category.current_level == 1:
q = Entry.query.filter(
and_(Entry.category1_id == category.id,
Entry.published_time >= day_before_str(sys_now(), 30)))
if category.current_level == 2:
q = Entry.query.filter(
and_(Entry.category2_id == category.id,
Entry.published_time >= day_before_str(sys_now(), 30)))
if category.current_level == 3:
q = Entry.query.filter(
and_(Entry.category3_id == category.id,
Entry.published_time >= day_before_str(sys_now(), 30)))
if q is None:
return []
_list = q.order_by('-entry.ranking').limit(PER_PAGE)
return _list
@staticmethod
def get_related_entries(entry_type):
# todo 获取相关的内容
#query = {'entry_type': entry_type}
#return self.get_page(condition=query, sort='_id', order=Order.DESC)
return []
@staticmethod
def get_statistic_by_author_id(author_id):
statistic = range(len(entry_type_str))
for i in range(len(entry_type_str)):
statistic[i] = Entry.query.filter(
and_(Entry.author_id == author_id,
Entry.entry_type == i)).count()
return statistic
@staticmethod
def delete(entry):
# todo 超过一定时间的也不让删除？
count_entry = Comment.count_by_entry_id(entry.id)
if count_entry > 0:
# 软删除
pass
else:
db.session.delete(entry)
# 删除相关评论等相关数据
Comment.del_by_entry_id(entry.id)
# 分类的文章数减少
# 只支持3级分类
if entry.category.current_level == 1:
Category.dec_entry_count(entry.category1_id)
if entry.category.current_level == 2:
Category.dec_entry_count(entry.category1_id)
Category.dec_entry_count(entry.category2_id)
if entry.category.current_level == 3:
Category.dec_entry_count(entry.category1_id)
Category.dec_entry_count(entry.category2_id)
Category.dec_entry_count(entry.category3_id)
# 要删除的标签
del_tags_list = json_decode(entry.tags)
if del_tags_list:
# 删除 tag_entry_relation （标签的文章关联）
TagEntryRelationService.del_relation(entry.id, del_tags_list)
TagService.dec_entry_count(del_tags_list)
# 更新 tag_category_relation （标签的分类计数）
TagCategoryRelationService.del_relation(entry.category,
del_tags_list)
db.session.delete(entry)
db.session.commit()
return True
@staticmethod
def modify_last_comment(entry_id, comment_id):
entry = Entry.get_by_id(entry_id)
entry.last_comment_id = comment_id
entry.upward_time = sys_now()
db.session.commit()
@staticmethod
def thanks(entry_id, user_id):
# 感谢
interact = EntryUserInteractService.get_by_entry_and_user_id(
entry_id, user_id)
if interact and interact.has_thanks:
return True
if not interact:
interact = EntryUserInteract()
interact.user_id = user_id
interact.entry_id = entry_id
interact.has_thanks = 1
EntryUserInteractService.add_or_update(interact)
entry = Entry.get_by_id(entry_id)
entry.num_thanks = Entry.__table__.c.num_thanks + 1
Entry.calc_ranking(entry)
db.session.commit()
return True
@staticmethod
def uninterested(entry_id, user_id):
# 不感兴趣
interact = EntryUserInteractService.get_by_entry_and_user_id(
entry_id, user_id)
if interact and interact.has_uninterested:
return True
if not interact:
interact = EntryUserInteract()
interact.user_id = user_id
interact.entry_id = entry_id
interact.has_uninterested = 1
EntryUserInteractService.add_or_update(interact)
entry = Entry.get_by_id(entry_id)
entry.num_uninterested = Entry.__table__.c.num_uninterested + 1
Entry.calc_ranking(entry)
db.session.commit()
return True
