Merge pull request #196 from macbre/tests-decrease-cardinality

Decrease cardinality to stabilize tests

Author: macbre

.github/workflows/python.yml

@@ -0,0 +1,74 @@
+name: Test against different Python version
+
+on:
+  push:
+    branches: master
+  pull_request:
+    branches: [ master ]
+
+jobs:
+  python_tests:
+    runs-on: ubuntu-latest
+
+    strategy:
+      # Do not fail if one the tests did not pass
+      fail-fast: false
+
+      matrix:
+        # Python version(s) to use when running the tests
+        python:
+          - "3.6"
+          - "3.8"
+          - "3.9-dev"
+
+        # Docker images of MySQL-compliant databases to run the tests suite on
+        database:
+         - "mysql:8.0.20"
+
+    services:
+        mysql:
+            image: ${{ matrix.database }}
+            env:
+                MYSQL_ALLOW_EMPTY_PASSWORD: yes
+                MYSQL_DATABASE: index_digest
+            ports:
+                - "3306:3306"
+            options: --health-cmd="mysqladmin ping" --health-interval=10s --health-timeout=5s --health-retries=3
+
+    steps:
+    - uses: actions/checkout@v2
+    - name: Set up Python
+      uses: actions/setup-python@v1
+      with:
+        python-version: ${{ matrix.python }}
+
+    # https://github.com/actions/cache/blob/main/examples.md#using-pip-to-get-cache-location
+    - name: Get pip cache dir
+      id: pip-cache
+      run: |
+        echo "::set-output name=dir::$(pip cache dir)"
+
+    - name: pip cache
+      uses: actions/cache@v2
+      with:
+        path: ${{ steps.pip-cache.outputs.dir }}
+        key: ${{ runner.os }}-pip-${{ hashFiles('**/requirements.txt', '**/setup.py') }}
+        restore-keys: |
+          ${{ runner.os }}-pip-
+
+    - name: Install dependencies
+      run: make install
+
+    - name: Linter
+      run: make lint
+
+    - name: Set up the database
+      run: |
+        docker ps
+        mysql --protocol=tcp -u root -v < setup.sql
+        # import the test schema files
+        "./sql/populate.sh"
+        mysql --protocol=tcp -uindex_digest -pqwerty index_digest -v -e '\s; SHOW TABLES; SHOW DATABASES;'
+
+    - name: Tests
+      run: make test

.github/workflows/tests.yml

@@ -2,11 +2,11 @@ name: Integration tests
 
 on:
   push:
-    branches: [ master ]
+    branches: master
   pull_request:
     branches: [ master ]
-jobs:
 
+jobs:
   integrations_tests:
     runs-on: ubuntu-latest
 
@@ -16,7 +16,8 @@ jobs:
 
       matrix:
         # Python version(s) to use when running the tests
-        python: [ 3.8, 3.9-dev ]
+        python:
+         - "3.8"
 
         # Docker images of MySQL-compliant databases to run the tests suite on
         database:

Makefile

@@ -5,7 +5,7 @@ install:
 	pip install -U -e .[dev]
 
 test:
-	pytest -x $(project_name)
+	pytest -x $(project_name) -o log_cli=true -o log_cli_level=warning
 
 coverage:
 	rm -f .coverage*

indexdigest/test/__init__.py

@@ -1,3 +1,5 @@
+import logging
+
 from ..database import Database
 
 from unittest import TestCase
@@ -17,6 +19,7 @@ def read_queries_from_log(log_file):
 
 class DatabaseTestMixin(object):
     DSN = 'mysql://index_digest:qwerty@127.0.0.1/index_digest'
+    DBNAME = 'index_digest'
 
     @property
     def connection(self):
@@ -31,6 +34,7 @@ class BigTableTest(TestCase, DatabaseTestMixin):
     ROWS = 100000  # how many rows to generate
     BATCH = 5000  # perform INSERT in batches
 
+    BIG_TABLE_NAME = '0020_big_table'
     PREPARED = False
 
     def setUp(self):
@@ -85,7 +89,7 @@ def _prepare_big_table(self):
         # no? populate it
         for row in self._rows():
             # Report low cardinality indices, use only a few distinct values (#31)
-            num = row % 3
+            num = row % 2
 
             values.append((row, val, '{:05x}'.format(row)[:5], num))
 
@@ -107,6 +111,14 @@ def _prepare_big_table(self):
         # update key distribution statistics (#31)
         self.connection.query('ANALYZE TABLE 0020_big_table')
 
+        cardinality_stats = self.connection.query_dict_rows(
+            "select TABLE_NAME, INDEX_NAME, COLUMN_NAME, CARDINALITY from"
+            " INFORMATION_SCHEMA.STATISTICS where"
+            " TABLE_NAME = '{table_name}' AND TABLE_SCHEMA = '{database_name}'".format(
+                table_name=self.BIG_TABLE_NAME, database_name=self.DBNAME)
+        )
+        logging.warning('Big table initialized, cardinality: %r', list(cardinality_stats))
+
     def table_populated(self):
         """
         :rtype: bool

indexdigest/test/linters/test_0031_low_cardinality_index.py

@@ -20,7 +20,7 @@ def test_get_low_cardinality_indices(self):
         assert index[0] == '0020_big_table'
         assert index[2]['INDEX_NAME'] == 'num_idx'
         assert index[2]['COLUMN_NAME'] == 'num'
-        assert index[2]['CARDINALITY'] > 1
+        assert index[2]['CARDINALITY'] >= 1
         assert index[2]['CARDINALITY'] <= INDEX_CARDINALITY_THRESHOLD
 
     def test_low_cardinality_index(self):
@@ -37,4 +37,5 @@ def test_low_cardinality_index(self):
         assert reports[0].context['column_name'] == 'num'
         assert reports[0].context['index_name'] == 'num_idx'
         assert isinstance(reports[0].context['index_cardinality'], int)
-        assert 33 <= int(reports[0].context['value_usage']) <= 35
+
+        self.assertAlmostEqual(int(reports[0].context['value_usage']), 50, delta=5)