You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
274 lines
9.6 KiB
Python
274 lines
9.6 KiB
Python
import logging
|
|
import threading
|
|
from urllib.parse import parse_qs, urlparse
|
|
|
|
from sqlalchemy import create_engine
|
|
from sqlalchemy.sql import text
|
|
from sqlalchemy.schema import MetaData
|
|
from sqlalchemy.util import safe_reraise
|
|
from sqlalchemy.engine.reflection import Inspector
|
|
|
|
from alembic.migration import MigrationContext
|
|
from alembic.operations import Operations
|
|
|
|
from dataset.table import Table
|
|
from dataset.util import ResultIter, row_type, safe_url, QUERY_STEP
|
|
from dataset.util import normalize_table_name
|
|
from dataset.types import Types
|
|
|
|
log = logging.getLogger(__name__)
|
|
|
|
|
|
class Database(object):
|
|
"""A database object represents a SQL database with multiple tables."""
|
|
|
|
def __init__(self, url, schema=None, reflect_metadata=True,
|
|
engine_kwargs=None, reflect_views=True,
|
|
ensure_schema=True, row_type=row_type):
|
|
"""Configure and connect to the database."""
|
|
if engine_kwargs is None:
|
|
engine_kwargs = {}
|
|
|
|
parsed_url = urlparse(url)
|
|
# if parsed_url.scheme.lower() in 'sqlite':
|
|
# # ref: https://github.com/pudo/dataset/issues/163
|
|
# if 'poolclass' not in engine_kwargs:
|
|
# engine_kwargs['poolclass'] = StaticPool
|
|
|
|
self.lock = threading.RLock()
|
|
self.local = threading.local()
|
|
|
|
if len(parsed_url.query):
|
|
query = parse_qs(parsed_url.query)
|
|
if schema is None:
|
|
schema_qs = query.get('schema', query.get('searchpath', []))
|
|
if len(schema_qs):
|
|
schema = schema_qs.pop()
|
|
|
|
self.schema = schema
|
|
self.engine = create_engine(url, **engine_kwargs)
|
|
self.types = Types(self.engine.dialect.name)
|
|
self.url = url
|
|
self.row_type = row_type
|
|
self.ensure_schema = ensure_schema
|
|
self._tables = {}
|
|
|
|
@property
|
|
def executable(self):
|
|
"""Connection against which statements will be executed."""
|
|
if not hasattr(self.local, 'conn'):
|
|
self.local.conn = self.engine.connect()
|
|
return self.local.conn
|
|
|
|
@property
|
|
def op(self):
|
|
"""Get an alembic operations context."""
|
|
ctx = MigrationContext.configure(self.executable)
|
|
return Operations(ctx)
|
|
|
|
@property
|
|
def inspect(self):
|
|
"""Get a SQLAlchemy inspector."""
|
|
return Inspector.from_engine(self.executable)
|
|
|
|
@property
|
|
def metadata(self):
|
|
"""Return a SQLAlchemy schema cache object."""
|
|
return MetaData(schema=self.schema, bind=self.executable)
|
|
|
|
@property
|
|
def in_transaction(self):
|
|
"""Check if this database is in a transactional context."""
|
|
if not hasattr(self.local, 'tx'):
|
|
return False
|
|
return len(self.local.tx) > 0
|
|
|
|
def _flush_tables(self):
|
|
"""Clear the table metadata after transaction rollbacks."""
|
|
for table in self._tables.values():
|
|
table._table = None
|
|
|
|
def begin(self):
|
|
"""Enter a transaction explicitly.
|
|
|
|
No data will be written until the transaction has been committed.
|
|
"""
|
|
if not hasattr(self.local, 'tx'):
|
|
self.local.tx = []
|
|
self.local.tx.append(self.executable.begin())
|
|
|
|
def commit(self):
|
|
"""Commit the current transaction.
|
|
|
|
Make all statements executed since the transaction was begun permanent.
|
|
"""
|
|
if hasattr(self.local, 'tx') and self.local.tx:
|
|
tx = self.local.tx.pop()
|
|
tx.commit()
|
|
self._flush_tables()
|
|
|
|
def rollback(self):
|
|
"""Roll back the current transaction.
|
|
|
|
Discard all statements executed since the transaction was begun.
|
|
"""
|
|
if hasattr(self.local, 'tx') and self.local.tx:
|
|
tx = self.local.tx.pop()
|
|
tx.rollback()
|
|
self._flush_tables()
|
|
|
|
def __enter__(self):
|
|
"""Start a transaction."""
|
|
self.begin()
|
|
return self
|
|
|
|
def __exit__(self, error_type, error_value, traceback):
|
|
"""End a transaction by committing or rolling back."""
|
|
if error_type is None:
|
|
try:
|
|
self.commit()
|
|
except Exception:
|
|
with safe_reraise():
|
|
self.rollback()
|
|
else:
|
|
self.rollback()
|
|
|
|
def close(self):
|
|
"""Close database connections. Makes this object unusable."""
|
|
self.engine.dispose()
|
|
self._tables = {}
|
|
self.engine = None
|
|
|
|
@property
|
|
def tables(self):
|
|
"""Get a listing of all tables that exist in the database."""
|
|
return self.inspect.get_table_names(schema=self.schema)
|
|
|
|
@property
|
|
def views(self):
|
|
"""Get a listing of all views that exist in the database."""
|
|
return self.inspect.get_view_names(schema=self.schema)
|
|
|
|
def __contains__(self, table_name):
|
|
"""Check if the given table name exists in the database."""
|
|
try:
|
|
table_name = normalize_table_name(table_name)
|
|
if table_name in self.tables:
|
|
return True
|
|
if table_name in self.views:
|
|
return True
|
|
return False
|
|
except ValueError:
|
|
return False
|
|
|
|
def create_table(self, table_name, primary_id=None, primary_type=None):
|
|
"""Create a new table.
|
|
|
|
Either loads a table or creates it if it doesn't exist yet. You can
|
|
define the name and type of the primary key field, if a new table is to
|
|
be created. The default is to create an auto-incrementing integer,
|
|
``id``. You can also set the primary key to be a string or big integer.
|
|
The caller will be responsible for the uniqueness of ``primary_id`` if
|
|
it is defined as a text type.
|
|
|
|
Returns a :py:class:`Table <dataset.Table>` instance.
|
|
::
|
|
|
|
table = db.create_table('population')
|
|
|
|
# custom id and type
|
|
table2 = db.create_table('population2', 'age')
|
|
table3 = db.create_table('population3',
|
|
primary_id='city',
|
|
primary_type=db.types.text)
|
|
# custom length of String
|
|
table4 = db.create_table('population4',
|
|
primary_id='city',
|
|
primary_type=db.types.string(25))
|
|
# no primary key
|
|
table5 = db.create_table('population5',
|
|
primary_id=False)
|
|
"""
|
|
assert not isinstance(primary_type, str), \
|
|
'Text-based primary_type support is dropped, use db.types.'
|
|
table_name = normalize_table_name(table_name)
|
|
with self.lock:
|
|
if table_name not in self._tables:
|
|
self._tables[table_name] = Table(self, table_name,
|
|
primary_id=primary_id,
|
|
primary_type=primary_type,
|
|
auto_create=True)
|
|
return self._tables.get(table_name)
|
|
|
|
def load_table(self, table_name):
|
|
"""Load a table.
|
|
|
|
This will fail if the tables does not already exist in the database. If
|
|
the table exists, its columns will be reflected and are available on
|
|
the :py:class:`Table <dataset.Table>` object.
|
|
|
|
Returns a :py:class:`Table <dataset.Table>` instance.
|
|
::
|
|
|
|
table = db.load_table('population')
|
|
"""
|
|
table_name = normalize_table_name(table_name)
|
|
with self.lock:
|
|
if table_name not in self._tables:
|
|
self._tables[table_name] = Table(self, table_name)
|
|
return self._tables.get(table_name)
|
|
|
|
def get_table(self, table_name, primary_id=None, primary_type=None):
|
|
"""Load or create a table.
|
|
|
|
This is now the same as ``create_table``.
|
|
::
|
|
|
|
table = db.get_table('population')
|
|
# you can also use the short-hand syntax:
|
|
table = db['population']
|
|
"""
|
|
return self.create_table(table_name, primary_id, primary_type)
|
|
|
|
def __getitem__(self, table_name):
|
|
"""Get a given table."""
|
|
return self.get_table(table_name)
|
|
|
|
def _ipython_key_completions_(self):
|
|
"""Completion for table names with IPython."""
|
|
return self.tables
|
|
|
|
def query(self, query, *args, **kwargs):
|
|
"""Run a statement on the database directly.
|
|
|
|
Allows for the execution of arbitrary read/write queries. A query can
|
|
either be a plain text string, or a `SQLAlchemy expression
|
|
<http://docs.sqlalchemy.org/en/latest/core/tutorial.html#selecting>`_.
|
|
If a plain string is passed in, it will be converted to an expression
|
|
automatically.
|
|
|
|
Further positional and keyword arguments will be used for parameter
|
|
binding. To include a positional argument in your query, use question
|
|
marks in the query (i.e. ``SELECT * FROM tbl WHERE a = ?```). For
|
|
keyword arguments, use a bind parameter (i.e. ``SELECT * FROM tbl
|
|
WHERE a = :foo``).
|
|
::
|
|
|
|
statement = 'SELECT user, COUNT(*) c FROM photos GROUP BY user'
|
|
for row in db.query(statement):
|
|
print(row['user'], row['c'])
|
|
|
|
The returned iterator will yield each result sequentially.
|
|
"""
|
|
if isinstance(query, str):
|
|
query = text(query)
|
|
_step = kwargs.pop('_step', QUERY_STEP)
|
|
if _step is False or _step == 0:
|
|
_step = None
|
|
rp = self.executable.execute(query, *args, **kwargs)
|
|
return ResultIter(rp, row_type=self.row_type, step=_step)
|
|
|
|
def __repr__(self):
|
|
"""Text representation contains the URL."""
|
|
return '<Database(%s)>' % safe_url(self.url)
|