Finished Release v2.1.1

* develop:
  Releasing v2.1.1
  fix precedence of ~ operator in Q objects
  ignore non-numeric parts of version string
  Fixes to make the tests pass on ClickHouse v21.9
  Bump urllib3 from 1.25.9 to 1.26.5 in /examples/db_explorer
  Bump jinja2 from 2.11.2 to 2.11.3 in /examples/db_explorer
  Simplified
  Support for adding a column to the beginning of a table
  1. add stddevPop func 2. add stddevSamp func
  changes reverted after rebase
  initializing changes related to string enums for pull request

CHANGELOG.md

@@ -1,6 +1,14 @@
 Change Log
 ==========
 
+v2.1.1
+------
+- Improve support of ClickHouse v21.9 (mangototango)
+- Ignore non-numeric parts in ClickHouse version (mangototango)
+- Fix precedence of ~ operator in Q objects	(mangototango)
+- Support for adding a column to the beginning of a table (meanmail)
+- Add stddevPop and stddevSamp functions (k.peskov)
+
 v2.1.0
 ------
 - Support for model constraints

examples/db_explorer/requirements.txt

@@ -6,10 +6,10 @@ idna==2.9
 infi.clickhouse-orm==2.0.1
 iso8601==0.1.12
 itsdangerous==1.1.0
-Jinja2==2.11.2
+Jinja2==2.11.3
 MarkupSafe==1.1.1
 pygal==2.4.0
 pytz==2020.1
 requests==2.23.0
-urllib3==1.25.9
+urllib3==1.26.5
 Werkzeug==1.0.1

src/infi/clickhouse_orm/database.py

@@ -51,6 +51,11 @@ class ServerError(DatabaseException):
             Code:\ (?P<code>\d+),
             \ e\.displayText\(\)\ =\ (?P<type1>[^ \n]+):\ (?P<msg>.+)
         ''', re.VERBOSE | re.DOTALL),
+        # ClickHouse v21+
+        re.compile(r'''
+            Code:\ (?P<code>\d+).
+            \ (?P<type1>[^ \n]+):\ (?P<msg>.+)
+        ''', re.VERBOSE | re.DOTALL),
     )
 
     @classmethod
@@ -408,7 +413,7 @@ class Database(object):
         except ServerError as e:
             logger.exception('Cannot determine server version (%s), assuming 1.1.0', e)
             ver = '1.1.0'
-        return tuple(int(n) for n in ver.split('.')) if as_tuple else ver
+        return tuple(int(n) for n in ver.split('.') if n.isdigit()) if as_tuple else ver
 
     def _is_existing_database(self):
         r = self._send("SELECT count() FROM system.databases WHERE name = '%s'" % self.db_name)

src/infi/clickhouse_orm/fields.py

@@ -109,7 +109,7 @@ class Field(FunctionOperatorsMixin):
         elif self.default:
             default = self.to_db_string(self.default)
             sql += ' DEFAULT %s' % default
-        if self.codec and db and db.has_codec_support:
+        if self.codec and db and db.has_codec_support and not self.alias:
             sql += ' CODEC(%s)' % self.codec
         return sql
 
@@ -468,9 +468,16 @@ class BaseEnumField(Field):
             return value
         try:
             if isinstance(value, str):
-                return self.enum_cls[value]
+                try:
+                    return self.enum_cls[value]
+                except Exception:
+                    return self.enum_cls(value)
             if isinstance(value, bytes):
-                return self.enum_cls[value.decode('UTF-8')]
+                decoded = value.decode('UTF-8')
+                try:
+                    return self.enum_cls[decoded]
+                except Exception:
+                    return self.enum_cls(decoded)
             if isinstance(value, int):
                 return self.enum_cls(value)
         except (KeyError, ValueError):
@@ -665,4 +672,3 @@ class LowCardinalityField(Field):
 
 # Expose only relevant classes in import *
 __all__ = get_subclass_names(locals(), Field)
-

src/infi/clickhouse_orm/funcs.py

@@ -391,11 +391,11 @@ class F(Cond, FunctionOperatorsMixin, metaclass=FMeta):
         return F('toYear', d)
 
     @staticmethod
-    def toISOYear(d, timezone=''):
+    def toISOYear(d, timezone=NO_VALUE):
         return F('toISOYear', d, timezone)
 
     @staticmethod
-    def toQuarter(d, timezone=''):
+    def toQuarter(d, timezone=NO_VALUE):
         return F('toQuarter', d, timezone) if timezone else F('toQuarter', d)
 
     @staticmethod
@@ -403,11 +403,11 @@ class F(Cond, FunctionOperatorsMixin, metaclass=FMeta):
         return F('toMonth', d)
 
     @staticmethod
-    def toWeek(d, mode=0, timezone=''):
+    def toWeek(d, mode=0, timezone=NO_VALUE):
         return F('toWeek', d, mode, timezone)
 
     @staticmethod
-    def toISOWeek(d, timezone=''):
+    def toISOWeek(d, timezone=NO_VALUE):
         return F('toISOWeek', d, timezone) if timezone else F('toISOWeek', d)
 
     @staticmethod
@@ -483,7 +483,7 @@ class F(Cond, FunctionOperatorsMixin, metaclass=FMeta):
         return F('toStartOfDay', d)
 
     @staticmethod
-    def toTime(d, timezone=''):
+    def toTime(d, timezone=NO_VALUE):
         return F('toTime', d, timezone)
 
     @staticmethod
@@ -491,47 +491,47 @@ class F(Cond, FunctionOperatorsMixin, metaclass=FMeta):
         return F('toTimeZone', dt, timezone)
 
     @staticmethod
-    def toUnixTimestamp(dt, timezone=''):
+    def toUnixTimestamp(dt, timezone=NO_VALUE):
         return F('toUnixTimestamp', dt, timezone)
 
     @staticmethod
-    def toYYYYMM(dt, timezone=''):
+    def toYYYYMM(dt, timezone=NO_VALUE):
         return F('toYYYYMM', dt, timezone) if timezone else F('toYYYYMM', dt)
 
     @staticmethod
-    def toYYYYMMDD(dt, timezone=''):
+    def toYYYYMMDD(dt, timezone=NO_VALUE):
         return F('toYYYYMMDD', dt, timezone) if timezone else F('toYYYYMMDD', dt)
 
     @staticmethod
-    def toYYYYMMDDhhmmss(dt, timezone=''):
+    def toYYYYMMDDhhmmss(dt, timezone=NO_VALUE):
         return F('toYYYYMMDDhhmmss', dt, timezone) if timezone else F('toYYYYMMDDhhmmss', dt)
 
     @staticmethod
-    def toRelativeYearNum(d, timezone=''):
+    def toRelativeYearNum(d, timezone=NO_VALUE):
         return F('toRelativeYearNum', d, timezone)
 
     @staticmethod
-    def toRelativeMonthNum(d, timezone=''):
+    def toRelativeMonthNum(d, timezone=NO_VALUE):
         return F('toRelativeMonthNum', d, timezone)
 
     @staticmethod
-    def toRelativeWeekNum(d, timezone=''):
+    def toRelativeWeekNum(d, timezone=NO_VALUE):
         return F('toRelativeWeekNum', d, timezone)
 
     @staticmethod
-    def toRelativeDayNum(d, timezone=''):
+    def toRelativeDayNum(d, timezone=NO_VALUE):
         return F('toRelativeDayNum', d, timezone)
 
     @staticmethod
-    def toRelativeHourNum(d, timezone=''):
+    def toRelativeHourNum(d, timezone=NO_VALUE):
         return F('toRelativeHourNum', d, timezone)
 
     @staticmethod
-    def toRelativeMinuteNum(d, timezone=''):
+    def toRelativeMinuteNum(d, timezone=NO_VALUE):
         return F('toRelativeMinuteNum', d, timezone)
 
     @staticmethod
-    def toRelativeSecondNum(d, timezone=''):
+    def toRelativeSecondNum(d, timezone=NO_VALUE):
         return F('toRelativeSecondNum', d, timezone)
 
     @staticmethod
@@ -555,7 +555,7 @@ class F(Cond, FunctionOperatorsMixin, metaclass=FMeta):
         return F('timeSlots', start_time, F.toUInt32(duration))
 
     @staticmethod
-    def formatDateTime(d, format, timezone=''):
+    def formatDateTime(d, format, timezone=NO_VALUE):
         return F('formatDateTime', d, format, timezone)
 
     @staticmethod
@@ -1633,6 +1633,16 @@ class F(Cond, FunctionOperatorsMixin, metaclass=FMeta):
     def varSamp(x):
         return F('varSamp', x)
 
+    @staticmethod
+    @aggregate
+    def stddevPop(expr):
+        return F('stddevPop', expr)
+
+    @staticmethod
+    @aggregate
+    def stddevSamp(expr):
+        return F('stddevSamp', expr)
+
     @staticmethod
     @aggregate
     @parametric

src/infi/clickhouse_orm/migrations.py

@@ -83,10 +83,12 @@ class AlterTable(ModelOperation):
             is_regular_field = not (field.materialized or field.alias)
             if name not in table_fields:
                 logger.info('        Add column %s', name)
-                assert prev_name, 'Cannot add a column to the beginning of the table'
                 cmd = 'ADD COLUMN %s %s' % (name, field.get_sql(db=database))
                 if is_regular_field:
-                    cmd += ' AFTER %s' % prev_name
+                    if prev_name:
+                        cmd += ' AFTER %s' % prev_name
+                    else:
+                        cmd += ' FIRST'
                 self._alter_table(database, cmd)
 
             if is_regular_field:

src/infi/clickhouse_orm/query.py

@@ -209,10 +209,10 @@ class Q(object):
 
     @classmethod
     def _construct_from(cls, l_child, r_child, mode):
-        if mode == l_child._mode:
+        if mode == l_child._mode and not l_child._negate:
             q = deepcopy(l_child)
             q._children.append(deepcopy(r_child))
-        elif mode == r_child._mode:
+        elif mode == r_child._mode and not r_child._negate:
             q = deepcopy(r_child)
             q._children.append(deepcopy(l_child))
         else:

tests/test_compressed_fields.py

@@ -106,7 +106,7 @@ class CompressedFieldsTestCase(unittest.TestCase):
                                     ('nullable_field', 'CODEC(ZSTD(1))'),
                                     ('array_field', 'CODEC(Delta(2), LZ4HC(0))'),
                                     ('float_field', 'CODEC(NONE)'),
-                                    ('alias_field', 'CODEC(ZSTD(4))')])
+                                    ('alias_field', '')])
 
 
 class CompressedModel(Model):

tests/test_database.py

@@ -181,12 +181,13 @@ class DatabaseTestCase(TestCaseWithData):
             Database(self.database.db_name, username='default', password='wrong')
 
         exc = cm.exception
+        print(exc.code, exc.message)
         if exc.code == 193: # ClickHouse version < 20.3
             self.assertTrue(exc.message.startswith('Wrong password for user default'))
         elif exc.code == 516: # ClickHouse version >= 20.3
             self.assertTrue(exc.message.startswith('default: Authentication failed'))
         else:
-            raise Exception('Unexpected error code - %s' % exc.code)
+            raise Exception('Unexpected error code - %s %s' % (exc.code, exc.message))
 
     def test_nonexisting_db(self):
         db = Database('db_not_here', autocreate=False)
@@ -251,6 +252,8 @@ class DatabaseTestCase(TestCaseWithData):
         from infi.clickhouse_orm.models import ModelBase
         query = "SELECT DISTINCT type FROM system.columns"
         for row in self.database.select(query):
+            if row.type.startswith('Map'):
+                continue  # Not supported yet
             ModelBase.create_ad_hoc_field(row.type)
 
     def test_get_model_for_table(self):
@@ -271,7 +274,12 @@ class DatabaseTestCase(TestCaseWithData):
         query = "SELECT name FROM system.tables WHERE database='system'"
         for row in self.database.select(query):
             print(row.name)
-            model = self.database.get_model_for_table(row.name, system_table=True)
+            if row.name in ('distributed_ddl_queue',):
+                continue  # Not supported
+            try:
+                model = self.database.get_model_for_table(row.name, system_table=True)
+            except NotImplementedError:
+                continue  # Table contains an unsupported field type
             self.assertTrue(model.is_system_model())
             self.assertTrue(model.is_read_only())
             self.assertEqual(model.table_name(), row.name)

tests/test_dictionaries.py

@@ -105,7 +105,7 @@ class HierarchicalDictionaryTest(DictionaryTestMixin, unittest.TestCase):
 
     def test_dictgethierarchy(self):
         self._test_func(F.dictGetHierarchy(self.dict_name, F.toUInt64(3)), [3, 2, 1])
-        self._test_func(F.dictGetHierarchy(self.dict_name, F.toUInt64(99)), [99])
+        self._test_func(F.dictGetHierarchy(self.dict_name, F.toUInt64(99)), [])
 
     def test_dictisin(self):
         self._test_func(F.dictIsIn(self.dict_name, F.toUInt64(3), F.toUInt64(1)), 1)

tests/test_engines.py

@@ -17,10 +17,10 @@ class _EnginesHelperTestCase(unittest.TestCase):
 
 
 class EnginesTestCase(_EnginesHelperTestCase):
-    def _create_and_insert(self, model_class):
+    def _create_and_insert(self, model_class, **kwargs):
         self.database.create_table(model_class)
         self.database.insert([
-            model_class(date='2017-01-01', event_id=23423, event_group=13, event_count=7, event_version=1)
+            model_class(date='2017-01-01', event_id=23423, event_group=13, event_count=7, event_version=1, **kwargs)
         ])
 
     def test_merge_tree(self):
@@ -155,7 +155,7 @@ class EnginesTestCase(_EnginesHelperTestCase):
             )
 
         self._create_and_insert(TestModel)
-        self._create_and_insert(TestCollapseModel)
+        self._create_and_insert(TestCollapseModel, sign=1)
 
         # Result order may be different, lets sort manually
         parts = sorted(list(SystemPart.get(self.database)), key=lambda x: x.table)
@@ -188,7 +188,7 @@ class EnginesTestCase(_EnginesHelperTestCase):
             )
 
         self._create_and_insert(TestModel)
-        self._create_and_insert(TestCollapseModel)
+        self._create_and_insert(TestCollapseModel, sign=1)
 
         self.assertEqual(2, len(list(SystemPart.get(self.database))))
 

tests/test_querysets.py

@@ -302,6 +302,18 @@ class QuerySetTestCase(TestCaseWithData):
         self.assertEqual(qs.conditions_as_sql(),
                          "(first_name = 'a') AND (greater(`height`, 1.7)) AND (last_name = 'b')")
 
+    def test_precedence_of_negation(self):
+        p = ~Q(first_name='a')
+        q = Q(last_name='b')
+        r = p & q
+        self.assertEqual(r.to_sql(Person), "(last_name = 'b') AND (NOT (first_name = 'a'))")
+        r = q & p
+        self.assertEqual(r.to_sql(Person), "(last_name = 'b') AND (NOT (first_name = 'a'))")
+        r = q | p
+        self.assertEqual(r.to_sql(Person), "(last_name = 'b') OR (NOT (first_name = 'a'))")
+        r = ~q & p
+        self.assertEqual(r.to_sql(Person), "(NOT (last_name = 'b')) AND (NOT (first_name = 'a'))")
+
     def test_invalid_filter(self):
         qs = Person.objects_in(self.database)
         with self.assertRaises(TypeError):