Merge branch 'mlflow:master' into master

CONTRIBUTING.rst

@@ -125,8 +125,9 @@ you can `sign your work`_ when committing code changes and opening pull requests
     git config --global user.name "Your Name"
     git config --global user.email yourname@example.com
 
-For convenience, we provide a pre-commit git hook that validates that commits are signed-off.
-Enable it by running:
+For convenience, we provide a pre-commit git hook that validates that commits are signed-off and
+runs `black --check` and `pylint` to ensure the code will pass the lint check for python.
+You can enable it by running:
 
 .. code-block:: bash
 

dev/set_matrix.py

@@ -316,8 +316,7 @@ def process_requirements(requirements, version=None):
             )
             if match_all:
                 return packages
-        else:
-            return []
+        return []
 
     raise TypeError("Invalid object type for `requirements`: '{}'".format(type(requirements)))
 

dev/show_package_release_dates.py

@@ -6,7 +6,7 @@
 
 
 def get_distributions():
-    res = subprocess.run(["pip", "list"], stdout=subprocess.PIPE)
+    res = subprocess.run(["pip", "list"], stdout=subprocess.PIPE, check=True)
     pip_list_stdout = res.stdout.decode("utf-8")
     # `pip_list_stdout` looks like this:
     # ``````````````````````````````````````````````````````````

hooks/pre-commit

@@ -0,0 +1,8 @@
+set -ex
+
+diff=$(git diff --cached --name-only | grep '\.py$' || true)
+
+if [ ! -z "$diff" ]; then
+  black --check $diff
+  pylint $diff
+fi

lint.sh

@@ -5,30 +5,6 @@
 err=0
 trap 'err=1' ERR
 
-FWDIR="$(cd "`dirname $0`"; pwd)"
-cd "$FWDIR"
-
-# https://stackoverflow.com/a/17841619
-function join {
-  local d=$1
-  shift
-  echo -n "$1"
-  shift
-  printf "%s" "${@/#/$d}"
-}
-
-include_dirs=(
-  "mlflow"
-  "tests"
-)
-
-exclude_dirs=(
-  "mlflow/protos"
-  "mlflow/server/js"
-  "mlflow/store/db_migrations"
-  "mlflow/temporary_db_migrations_for_pre_1_users"
-)
-
 echo -e "\n========== black ==========\n"
 # Exclude proto files because they are auto-generated
 black --check .
@@ -38,23 +14,8 @@ if [ $? -ne 0 ]; then
   echo '$ pip install $(cat dev/lint-requirements.txt | grep "black==") && black .'
 fi
 
-echo -e "\n========== pycodestyle ==========\n"
-exclude=$(join "," "${exclude_dirs[@]}")
-include=$(join " " "${include_dirs[@]}")
-pycodestyle --max-line-length=100 --ignore=E203,W503 --exclude=$exclude -- $include
-
 echo -e "\n========== pylint ==========\n"
-# pylint's `--ignore` option filters files based on their base names, not paths.
-# see: http://pylint.pycqa.org/en/latest/user_guide/run.html#command-line-options
-# This behavior might cause us to unintentionally ignore some files.
-# To avoid this issue, select files to lint using `git ls-files` and `grep`.
-# Another advantage of this approach is we can apply pylint to all python scripts
-# without creating `__init__.py` in all directories.
-exclude="^\($(join "\|" "${exclude_dirs[@]}")\)/.\+\.py$"
-include="^\($(join "\|" "${include_dirs[@]}")\)/.\+\.py$"
-msg_template="{path} ({line},{column}): [{msg_id} {symbol}] {msg}"
-
-pylint --jobs=0 --msg-template="$msg_template" --rcfile="$FWDIR/pylintrc" $(git ls-files | grep $include | grep -v $exclude)
+pylint $(git ls-files | grep '\.py$')
 
 echo -e "\n========== rstcheck ==========\n"
 rstcheck README.rst

pylintrc

@@ -7,19 +7,28 @@ extension-pkg-whitelist=
 
 # Add files or directories to the denylist. They should be base names, not
 # paths.
-ignore=build,protos,sdk,db_migrations,temporary_db_migrations_for_pre_1_users
-
+ignore=
 
 # Add files or directories matching the regex patterns to the denylist. The
 # regex matches against base names, not paths.
 ignore-patterns=
 
+# Add files or directories matching the regex patterns to the ignore-list.
+# The regex matches against paths.
+ignore-paths=setup.py,
+             docs,
+             examples,
+             mlflow/protos,
+             mlflow/server/js,
+             mlflow/store/db_migrations,
+             mlflow/temporary_db_migrations_for_pre_1_users
+
 # Python code to execute, usually for sys.path manipulation such as
 # pygtk.require().
 #init-hook=
 
 # Use multiple processes to speed up Pylint.
-jobs=1
+jobs=0
 
 # List of plugins (as comma separated values of python modules names) to load,
 # usually to register additional checkers.