371 lines
10 KiB
Python

from collections import namedtuple
from itertools import islice
from ..errors import Error as DriverError
from .errors import InterfaceError, OperationalError, ProgrammingError
Column = namedtuple(
'Column',
'name type_code display_size internal_size precision scale null_ok'
)
class Cursor(object):
class States(object):
(
NONE,
RUNNING,
FINISHED,
CURSOR_CLOSED
) = range(4)
_states = States()
def __init__(self, client, connection):
self._client = client
self._connection = connection
self._reset_state()
self.arraysize = 1
# Begin non-PEP attributes
self._columns_with_types = None
# End non-PEP attributes
super(Cursor, self).__init__()
def __repr__(self):
is_closed = self._state == self._states.CURSOR_CLOSED
return '<cursor object at 0x{0:x}; closed: {1:}>'.format(
id(self), is_closed
)
# Iteration support.
def __iter__(self):
while True:
one = self.fetchone()
if one is None:
return
yield one
# Context manager integrations.
def __enter__(self):
return self
def __exit__(self, exc_type, exc_val, exc_tb):
self.close()
@property
def description(self):
if self._state == self._states.NONE:
return None
columns = self._columns or []
types = self._types or []
return [
Column(name, type_code, None, None, None, None, True)
for name, type_code in zip(columns, types)
]
@property
def rowcount(self):
"""
:return: the number of rows that the last .execute*() produced.
"""
return self._rowcount
def close(self):
"""
Close the cursor now. The cursor will be unusable from this point
forward; an :data:`~clickhouse_driver.dbapi.Error` (or subclass)
exception will be raised if any operation is attempted with the
cursor.
"""
self._client.disconnect()
self._state = self._states.CURSOR_CLOSED
try:
# cursor can be already closed
self._connection.cursors.remove(self)
except ValueError:
pass
def execute(self, operation, parameters=None):
"""
Prepare and execute a database operation (query or command).
:param operation: query or command to execute.
:param parameters: sequence or mapping that will be bound to
variables in the operation.
:return: None
"""
self._check_cursor_closed()
self._begin_query()
try:
execute, execute_kwargs = self._prepare()
response = execute(
operation, params=parameters, with_column_types=True,
**execute_kwargs
)
except DriverError as orig:
raise OperationalError(orig)
self._process_response(response)
self._end_query()
def executemany(self, operation, seq_of_parameters):
"""
Prepare a database operation (query or command) and then execute it
against all parameter sequences found in the sequence
`seq_of_parameters`.
:param operation: query or command to execute.
:param seq_of_parameters: sequences or mappings for execution.
:return: None
"""
self._check_cursor_closed()
self._begin_query()
try:
execute, execute_kwargs = self._prepare()
response = execute(
operation, params=seq_of_parameters, **execute_kwargs
)
except DriverError as orig:
raise OperationalError(orig)
self._process_response(response, executemany=True)
self._end_query()
def fetchone(self):
"""
Fetch the next row of a query result set, returning a single sequence,
or None when no more data is available.
:return: the next row of a query result set or None.
"""
self._check_query_started()
if self._stream_results:
return next(self._rows, None)
else:
if not self._rows:
return None
return self._rows.pop(0)
def fetchmany(self, size=None):
"""
Fetch the next set of rows of a query result, returning a sequence of
sequences (e.g. a list of tuples). An empty sequence is returned when
no more rows are available.
:param size: amount of rows to return.
:return: list of fetched rows or empty list.
"""
self._check_query_started()
if size is None:
size = self.arraysize
if self._stream_results:
if size == -1:
return list(self._rows)
else:
return list(islice(self._rows, size))
if size < 0:
rv = self._rows
self._rows = []
else:
rv = self._rows[:size]
self._rows = self._rows[size:]
return rv
def fetchall(self):
"""
Fetch all (remaining) rows of a query result, returning them as a
sequence of sequences (e.g. a list of tuples).
:return: list of fetched rows.
"""
self._check_query_started()
if self._stream_results:
return list(self._rows)
rv = self._rows
self._rows = []
return rv
def setinputsizes(self, sizes):
# Do nothing.
pass
def setoutputsize(self, size, column=None):
# Do nothing.
pass
# Begin non-PEP methods
@property
def columns_with_types(self):
"""
:return: list of column names with corresponding types of the last
.execute*(). E.g. [('x', 'UInt64')].
"""
return self._columns_with_types
def set_stream_results(self, stream_results, max_row_buffer):
"""
Toggles results streaming from server. Driver will consume
block-by-block of `max_row_buffer` size and yield row-by-row from each
block.
:param stream_results: enable or disable results streaming.
:param max_row_buffer: specifies the maximum number of rows to buffer
at a time.
:return: None
"""
self._stream_results = stream_results
self._max_row_buffer = max_row_buffer
def set_settings(self, settings):
"""
Specifies settings for cursor.
:param settings: dictionary of query settings
:return: None
"""
self._settings = settings
def set_types_check(self, types_check):
"""
Toggles type checking for sequence of INSERT parameters.
Disabled by default.
:param types_check: new types check value.
:return: None
"""
self._types_check = types_check
def set_external_table(self, name, structure, data):
"""
Adds external table to cursor context.
If the same table is specified more than once the last one is used.
:param name: name of external table
:param structure: list of tuples (name, type) that defines table
structure. Example [(x, 'Int32')].
:param data: sequence of rows of tuples or dicts for transmission.
:return: None
"""
self._external_tables[name] = (structure, data)
def set_query_id(self, query_id):
"""
Specifies the query identifier for cursor.
:param query_id: the query identifier.
:return: None
"""
self._query_id = query_id
# End non-PEP methods
# Private methods.
def _prepare(self):
external_tables = [
{'name': name, 'structure': structure, 'data': data}
for name, (structure, data) in self._external_tables.items()
] or None
execute = self._client.execute
if self._stream_results:
execute = self._client.execute_iter
self._settings = self._settings or {}
self._settings['max_block_size'] = self._max_row_buffer
execute_kwargs = {
'settings': self._settings,
'external_tables': external_tables,
'types_check': self._types_check,
'query_id': self._query_id
}
return execute, execute_kwargs
def _process_response(self, response, executemany=False):
if executemany:
self._rowcount = response
response = None
if not response or isinstance(response, int):
self._columns = self._types = self._rows = []
if isinstance(response, int):
self._rowcount = response
return
if self._stream_results:
columns_with_types = next(response)
rows = response
else:
rows, columns_with_types = response
self._columns_with_types = columns_with_types
# Only SELECT queries have columns_with_types.
# DDL and INSERT INTO ... SELECT queries have empty columns header.
# We need to obtain rows count only during non-streaming SELECTs.
if columns_with_types:
self._columns, self._types = zip(*columns_with_types)
if not self._stream_results:
self._rowcount = len(rows)
else:
self._columns = self._types = []
self._rows = rows
def _reset_state(self):
"""
Resets query state and get ready for another query.
"""
self._state = self._states.NONE
self._columns = None
self._types = None
self._rows = None
self._rowcount = -1
self._stream_results = False
self._max_row_buffer = 0
self._settings = None
self._query_id = None
self._external_tables = {}
self._types_check = False
def _begin_query(self):
self._state = self._states.RUNNING
def _end_query(self):
self._state = self._states.FINISHED
def _check_cursor_closed(self):
if self._state == self._states.CURSOR_CLOSED:
raise InterfaceError('cursor already closed')
def _check_query_started(self):
if self._state == self._states.NONE:
raise ProgrammingError('no results to fetch')