Check python code with flake8

Recently assert-on-tuple had been introduced in tests [1], let's prevent
this.

  [1]: https://github.com/ClickHouse/ClickHouse/pull/56367#discussion_r1437098533

v2: pin flake8 to 4.0.1 (instead of originally 6.1) due to other dependencies, hope that it will find such errors
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>

docker/test/style/Dockerfile

@@ -30,6 +30,7 @@ RUN pip3 install \
         mypy==1.8.0 \
         pylint==3.1.0 \
         python-magic==0.4.24 \
+        flake8==4.0.1 \
         requests \
         thefuzz \
         types-requests \

docker/test/style/run.sh

@@ -9,6 +9,8 @@ echo "Check style" | ts
 ./check-style -n              |& tee /test_output/style_output.txt
 echo "Check python formatting with black" | ts
 ./check-black -n              |& tee /test_output/black_output.txt
+echo "Check python with flake8" | ts
+./check-flake8                |& tee /test_output/flake8_output.txt
 echo "Check python type hinting with mypy" | ts
 ./check-mypy -n               |& tee /test_output/mypy_output.txt
 echo "Check typos" | ts

docs/en/development/continuous-integration.md

@@ -91,6 +91,9 @@ cd ./utils/check-style
 # Check python type hinting with mypy
 ./check-mypy
 
+# Check python with flake8
+./check-flake8
+
 # Check code with codespell
 ./check-typos
 

utils/check-style/check-flake8

@@ -0,0 +1,55 @@
+#!/usr/bin/env bash
+
+function join_by() { local IFS="$1"; shift; echo "$*"; }
+
+set -e
+
+# We check only our code, that's why we skip contrib
+GIT_ROOT=$(git rev-parse --show-cdup)
+GIT_ROOT=${GIT_ROOT:-./}
+
+# Find all *.py, *.python files and executable files without extension
+# that are determined as python scripts by 'file' util
+# in the repo except the contrib directory.
+find_cmd=(
+  find "$GIT_ROOT" -type f -not -path "${GIT_ROOT}contrib/*"
+    \(
+      \(
+        -name '*.py' -or -name "*.python" -or
+          \(
+            -executable -not -name "*.*" -exec sh -c 'file {} | grep -q "Python script"' \;
+          \)
+      \)
+      # We skip modules generated by the protocol buffer compiler from *.proto files.
+      -and -not -name '*_pb2.py' -and -not -name '*_pb2_grpc.py'
+    \) -print0
+)
+
+ignores=(
+    E101 # Indentation contains mixed spaces and tabs
+    E203 # Whitespace before ':'
+    E226 # missing whitespace around arithmetic operator
+    E266 # Too many leading '#' for block comment
+    E401 # Multiple imports on one line
+    E402 # Module level import not at top of file
+    E501 # line too long
+    E711 # Comparison to None should be 'cond is None:'
+    E712 # Comparison to true should be 'if cond is true:' or 'if cond:'
+    E713 # Test for membership should be 'not in'
+    E714 # Test for object identity should be 'is not'
+    E722 # Do not use bare except, specify exception instead
+    E731 # Do not assign a lambda expression, use a def
+    E741 # Do not use variables named 'I', 'O', or 'l'
+    F401 # Module imported but unused
+    F403 # 'from module import *' used; unable to detect undefined names
+    F405 # Name may be undefined, or defined from star imports: module
+    F522 # .format(...) unused named arguments
+    F541 # f-string without any placeholders
+    F811 # redefinition of unused name from line N
+    F841 # local variable name is assigned to but never used
+    W191 # Indentation contains tabs
+    W291 # Trailing whitespace
+    W293 # Blank line contains whitespace
+    W503 # Line break occurred before a binary operator
+)
+"${find_cmd[@]}" | xargs -0 flake8 --ignore "$(join_by , "${ignores[@]}")"

utils/check-style/process_style_check_result.py

@@ -18,6 +18,7 @@ def process_result(result_folder):
         "style",
         "pylint",
         "black",
+        "flake8",
         "mypy",
         "typos",
         "whitespaces",