def __init__(self,
              server_side_cursors=False,
              use_native_unicode=True,
              **kwargs):
     PGDialect.__init__(self, **kwargs)
     self.server_side_cursors = server_side_cursors
     self.use_native_unicode = use_native_unicode
     self.supports_unicode_binds = use_native_unicode
 def __init__(self,
              server_side_cursors=False,
              use_native_unicode=True,
              **kwargs):
     PGDialect.__init__(self, **kwargs)
     self.server_side_cursors = server_side_cursors
     self.use_native_unicode = use_native_unicode
     self.supports_unicode_binds = use_native_unicode
     if self.dbapi and hasattr(self.dbapi, '__version__'):
         m = re.match(r'(\d+)\.(\d+)(?:\.(\d+))?', self.dbapi.__version__)
         if m:
             self.psycopg2_version = tuple(
                 int(x) for x in m.group(1, 2, 3) if x is not None)
示例#3
0
 def __init__(self, server_side_cursors=False, use_native_unicode=True, **kwargs):
     PGDialect.__init__(self, **kwargs)
     self.server_side_cursors = server_side_cursors
     self.use_native_unicode = use_native_unicode
     self.supports_unicode_binds = use_native_unicode
     if self.dbapi and hasattr(self.dbapi, '__version__'):
         m = re.match(r'(\d+)\.(\d+)(?:\.(\d+))?', 
                             self.dbapi.__version__)
         if m:
             self.psycopg2_version = tuple(
                                         int(x) 
                                         for x in m.group(1, 2, 3) 
                                         if x is not None)
示例#4
0
class Db_Postgis(BaseDb):
    """Functionality for using sqlalchemy to connect to a PostGIS database."""

    preparer = PGIdentifierPreparer(PGDialect())

    @classmethod
    def create_engine(cls, pgurl):
        def _on_checkout(dbapi_connection, connection_record,
                         connection_proxy):
            with dbapi_connection.cursor() as dbcur:
                dbcur.execute("SET timezone='UTC';")
                dbcur.execute("SET intervalstyle='iso_8601';")
                # don't drop precision from floats near the edge of their supported range
                dbcur.execute("SET extra_float_digits = 3;")

        pgurl = cls._append_query_to_url(pgurl,
                                         {"fallback_application_name": "kart"})

        engine = sqlalchemy.create_engine(pgurl,
                                          module=psycopg2,
                                          poolclass=cls._pool_class())
        sqlalchemy.event.listen(engine, "checkout", _on_checkout)

        return engine

    @classmethod
    def list_tables(cls, sess, db_schema=None):
        if db_schema is not None:
            name_clause = "c.relname"
            schema_clause = "n.nspname = :db_schema"
            params = {"db_schema": db_schema}
        else:
            name_clause = "format('%s.%s', n.nspname, c.relname)"
            schema_clause = "n.nspname NOT IN ('information_schema', 'pg_catalog', 'tiger', 'topology')"
            params = {}

        r = sess.execute(
            sqlalchemy.text(f"""
                SELECT {name_clause} as name, obj_description(c.oid, 'pg_class') as title
                FROM pg_catalog.pg_class c
                    INNER JOIN pg_namespace n ON n.oid = c.relnamespace
                WHERE c.relkind IN ('r', 'v') AND {schema_clause}
                AND c.relname NOT LIKE '_kart_%'
                AND c.oid NOT IN (
                    SELECT d.objid
                    FROM pg_catalog.pg_extension AS e
                        INNER JOIN pg_catalog.pg_depend AS d ON (d.refobjid = e.oid)
                    WHERE d.deptype = 'e'
                    AND e.extname = 'postgis'
                )
                ORDER BY {name_clause};
                """),
            params,
        )
        return {row["name"]: row["title"] for row in r}

    @classmethod
    def db_schema_searchpath(cls, sess):
        return sess.scalar("SELECT current_schemas(true);")
示例#5
0
 async def init_pool(self):
     # dsn for connections
     dsn = 'postgresql://%s:%s@%s:%s/%s' % \
           (self._config['user'], self._config['password'], self._config['host'],
            self._config['port'], self._config['database'])
     # create pool
     self._pool = await create_pool(
         dsn=dsn,
         min_size=self._config.get('min_size', 5),
         max_size=self._config.get('max_size', 10),
         dialect=PGDialect())
示例#6
0
    def __init__(self, project=None, rpc=None):
        """

        Args:
            project (str, optional): A Project Identifier
            rpc (Connect, optional): An RPC Connection object
        """
        if rpc:
            self.rpc = rpc
        else:
            self.rpc = Connect()

        if project:
            try:
                # See if this is a project ID already
                uuid.UUID(project)
                self._project_id = six.text_type(project)
            except ValueError:
                if '/' in project:
                    # This is a path lookup
                    self._project_id = self.rpc.analyze.project.lookup_by_full_path(path=project)
                else:
                    # This is a name lookup
                    self._project_id = self.rpc.analyze.project.lookup_by_name(name=project)
        else:
            self._project_id = rpc.project_id

        _dialect_kind = 'greenplum'   # This should come from the project primary database setting eventually

        if _dialect_kind == 'greenplum':
            self.dialect = GreenplumDialect()
        elif _dialect_kind == 'hana':
            self.dialect = HANABaseDialect()
        elif _dialect_kind == 'hive':
            raise Exception('Hive not supported from plaidtools currently.')
        elif _dialect_kind == 'spark':
            raise Exception('Spark not supported from plaidtools currently.')
        elif _dialect_kind == 'oracle':
            raise Exception('Oracle not supported from plaidtools currently.')
        elif _dialect_kind == 'mssql':
            raise Exception('MS SQL Server not supported from plaidtools currently.')
        else:
            self.dialect = PGDialect()
示例#7
0
 def import_schema(self, schema, srv_options, options,
                   restriction_type, restricts):
     """
     Reflects the remote schema.
     """
     metadata = MetaData()
     url = _parse_url_from_options(srv_options)
     engine = create_engine(url)
     dialect = PGDialect()
     if restriction_type == 'limit':
         only = restricts
     elif restriction_type == 'except':
         only = lambda t, _: t not in restricts
     else:
         only = None
     metadata.reflect(bind=engine,
                      schema=schema,
                      only=only)
     to_import = []
     for _, table in sorted(metadata.tables.items()):
         ftable = TableDefinition(table.name)
         ftable.options['schema'] = schema
         ftable.options['tablename'] = table.name
         for c in table.c:
             # Force collation to None to prevent imcompatibilities
             setattr(c.type, "collation", None)
             # If the type is specialized, call the generic
             # superclass method
             if type(c.type) in CONVERSION_MAP:
                 class_name = CONVERSION_MAP[type(c.type)]
                 old_args = c.type.__dict__
                 c.type = class_name()
                 c.type.__dict__.update(old_args)
             if c.primary_key:
                 ftable.options['primary_key'] = c.name
             ftable.columns.append(ColumnDefinition(
                 c.name,
                 type_name=c.type.compile(dialect)))
         to_import.append(ftable)
     return to_import
示例#8
0
 def __init__(self, server_side_cursors=False, use_native_unicode=True, **kwargs):
     PGDialect.__init__(self, **kwargs)
     self.server_side_cursors = server_side_cursors
     self.use_native_unicode = use_native_unicode
     self.supports_unicode_binds = use_native_unicode
示例#9
0
from pybigquery.sqlalchemy_bigquery import BigQueryDialect
from snowflake.sqlalchemy.snowdialect import SnowflakeDialect
from sqlalchemy.dialects.mysql.base import MySQLDialect
from sqlalchemy.dialects.postgresql.base import PGDialect

from panoramic.cli.husky.service.types.enums import HuskyQueryRuntime

RUNTIME_DIALECTS = {
    HuskyQueryRuntime.snowflake: SnowflakeDialect(),
    HuskyQueryRuntime.bigquery: BigQueryDialect(),
    HuskyQueryRuntime.mysql: MySQLDialect(),
    HuskyQueryRuntime.postgres: PGDialect(),
}