Fix query_context.updated_copy method, add first examples

.github/workflows/on_push.yml

@@ -34,6 +34,7 @@ jobs:
         run: |
           pylint clickhouse_connect
           pylint tests
+          pylint examples
 
   tests:
     runs-on: ubuntu-latest

CHANGELOG.md

@@ -1,5 +1,13 @@
 ## ClickHouse Connect ChangeLog
 
+
+### Release 0.3.6 2022-11-02
+
+#### Bug Fix
+* Update QueryContext.updated_copy method to preserve settings, parameters, etc.  https://github.com/ClickHouse/clickhouse-connect/issues/65
+
+
+
 ### Release 0.3.5 2022-10-28
 
 #### Improvement

clickhouse_connect/VERSION

@@ -1 +1 @@
-0.3.5
+0.3.6

clickhouse_connect/driver/common.py

@@ -1,7 +1,7 @@
 import array
 import sys
 
-from typing import Tuple, Sequence, MutableSequence
+from typing import Tuple, Sequence, MutableSequence, Dict, Optional
 
 # pylint: disable=invalid-name
 must_swap = sys.byteorder == 'big'
@@ -171,3 +171,10 @@ def unescape_identifier(x: str) -> str:
     if x.startswith('`') and x.endswith('`'):
         return x[1:-1]
     return x
+
+
+def dict_copy(source: Dict = None, update: Optional[Dict] = None) -> Dict:
+    copy = source.copy() if source else {}
+    if update:
+        copy.update(update)
+    return copy

clickhouse_connect/driver/query.py

@@ -7,6 +7,7 @@
 from typing import NamedTuple, Any, Tuple, Dict, Sequence, Optional, Union
 from datetime import date, datetime, tzinfo
 
+from clickhouse_connect.driver.common import dict_copy
 from clickhouse_connect.json_impl import any_to_json
 from clickhouse_connect.common import common_settings
 from clickhouse_connect.datatypes.base import ClickHouseType
@@ -114,10 +115,10 @@ def updated_copy(self,
         Creates Query context copy with parameters overridden/updated as appropriate
         """
         return QueryContext(query or self.query,
-                            self.parameters.update(parameters or {}),
-                            self.settings.update(settings or {}),
-                            self.query_formats.update(query_formats or {}),
-                            self.column_formats.update(column_formats or {}),
+                            dict_copy(self.parameters, parameters),
+                            dict_copy(self.settings, settings),
+                            dict_copy(self.query_formats, query_formats),
+                            dict_copy(self.column_formats, column_formats),
                             encoding if encoding else self.encoding,
                             server_tz if server_tz else self.server_tz,
                             use_none if use_none is not None else self.use_none)

examples/pandas_examples.py

@@ -0,0 +1,36 @@
+import pandas as pd
+import clickhouse_connect
+
+
+create_table_sql = """
+CREATE TABLE pandas_example
+(
+    `timeseries` DateTime('UTC'),
+    `int_value` Int32,
+    `str_value` String,
+    `float_value` Float64
+)
+ENGINE = MergeTree
+ORDER BY timeseries
+"""
+
+
+def write_pandas_df():
+    client = clickhouse_connect.get_client(host='localhost', port='8123', user_name='default', password = '')
+    client.command('DROP TABLE IF EXISTS pandas_example')
+    client.command(create_table_sql)
+    df = pd.DataFrame({'timeseries': ['04/03/2022 10:00:11', '05/03/2022 11:15:44', '06/03/2022 17:14:00'],
+                      'int_value': [16, 19, 11],
+                       'str_value': ['String One', 'String Two', 'A Third String'],
+                       'float_value': [2344.288, -73002.4444, 3.14159]})
+    df['timeseries'] = pd.to_datetime(df['timeseries'])
+    client.insert_df('pandas_example', df)
+    result_df = client.query_df('SELECT * FROM pandas_example')
+    print()
+    print(result_df.dtypes)
+    print()
+    print(result_df)
+
+
+if __name__ == '__main__':
+    write_pandas_df()

examples/params_example.py

@@ -0,0 +1,57 @@
+from datetime import datetime, timedelta
+
+from clickhouse_connect.driver.query import finalize_query
+
+select_template = """
+  SELECT
+    formatDateTime(started_at, '%%m/%%d/%%Y', %(time_zone)s) AS date,
+    formatDateTime(started_at, '%%I:%%M:%%S %%p', %(time_zone)s) AS time,
+    format('{}path/link?name={}&dev_type={}', %(web_url)s, label, device_type) AS url,
+    device_name,
+    description
+  FROM sessions
+"""
+
+
+def build_device_query(time_zone: str,
+                       web_url: str,
+                       client: str,
+                       company_id: str = '',
+                       device_id: str = '',
+                       updated: bool = False,
+                       start_time: datetime = None,
+                       end_time: datetime = None):
+    params = {'time_zone': time_zone,
+              'web_url': web_url,
+              'client': client
+              }
+    where_template = ' WHERE client = %(client)s'
+    if company_id:
+        where_template += ' AND company_id = %(company_id)s'
+        params['company_id'] = company_id
+    if device_id:
+        where_template += ' AND dev_type = %(device_id)s'
+        params['device_id'] = device_id
+    if updated:
+        where_template += ' AND updated = true'
+    if start_time and end_time:
+        where_template += ' AND started_at BETWEEN %(start_time)s AND %(end_time)s'
+        params['start_time'] = start_time
+        params['end_time'] = end_time
+    full_query = select_template + where_template + ' ORDER BY started_at ASC'
+    return finalize_query(full_query, params)
+
+
+if __name__ == '__main__':
+    start = datetime.now()
+    end = start + timedelta(hours=1, minutes=20)
+    print(build_device_query('UTC',
+                             'https://example.com',
+
+                             client='Client_0',
+                             company_id='Company_1',
+                             device_id='DEVICE_77',
+                             start_time=start,
+                             end_time=end
+                             )
+          )

tests/unit_tests/test_driver/test_query.py

@@ -0,0 +1,26 @@
+from clickhouse_connect.driver.query import QueryContext
+
+
+def test_copy_context():
+    settings = {'max_bytes_for_external_group_by': 1024 * 1024 * 100,
+                'read_overflow_mode': 'throw'}
+    parameters = {'user_id': 'user_1'}
+    query_formats = {'IPv*': 'string'}
+    context = QueryContext('SELECT source_ip FROM table WHERE user_id = %(user_id)s',
+                           settings=settings,
+                           parameters=parameters,
+                           query_formats=query_formats,
+                           use_none=True)
+    assert context.use_none is True
+    assert context.final_query == "SELECT source_ip FROM table WHERE user_id = 'user_1'"
+    assert context.query_formats['IPv*'] == 'string'
+    assert context.settings['max_bytes_for_external_group_by'] == 104857600
+
+    context_copy = context.updated_copy(
+        settings={'max_bytes_for_external_group_by': 1024 * 1024 * 24, 'max_execution_time': 120},
+        parameters={'user_id': 'user_2'}
+    )
+    assert context_copy.settings['read_overflow_mode'] == 'throw'
+    assert context_copy.settings['max_execution_time'] == 120
+    assert context_copy.settings['max_bytes_for_external_group_by'] == 25165824
+    assert context_copy.final_query == "SELECT source_ip FROM table WHERE user_id = 'user_2'"