[SPARK-23367][Build] Include python document style checking

dev/lint-python

@@ -21,10 +21,16 @@ SCRIPT_DIR="$( cd "$( dirname "$0" )" && pwd )"
 SPARK_ROOT_DIR="$(dirname "$SCRIPT_DIR")"
 # Exclude auto-generated configuration file.
 PATHS_TO_CHECK="$( cd "$SPARK_ROOT_DIR" && find . -name "*.py" )"
+DOC_PATHS_TO_CHECK="$( cd "$SPARK_ROOT_DIR" && find . -name "*.py" | grep -vF 'functions.py' )"
 PYCODESTYLE_REPORT_PATH="$SPARK_ROOT_DIR/dev/pycodestyle-report.txt"
+PYDOCSTYLE_REPORT_PATH="$SPARK_ROOT_DIR/dev/pydocstyle-report.txt"
 PYLINT_REPORT_PATH="$SPARK_ROOT_DIR/dev/pylint-report.txt"
 PYLINT_INSTALL_INFO="$SPARK_ROOT_DIR/dev/pylint-info.txt"
+PYDOCSTYLEBUILD="pydocstyle"
+EXPECTED_PYDOCSTYLEVERSION="2.1.1"
+PYDOCSTYLEVERSION=$(python -c 'import pkg_resources; print(pkg_resources.get_distribution("pydocstyle").version)' 2> /dev/null)
 SPHINXBUILD=${SPHINXBUILD:=sphinx-build}
+
 SPHINX_REPORT_PATH="$SPARK_ROOT_DIR/dev/sphinx-report.txt"
 
 cd "$SPARK_ROOT_DIR"
@@ -67,19 +73,36 @@ python "$PYCODESTYLE_SCRIPT_PATH" --config=dev/tox.ini $PATHS_TO_CHECK >> "$PYCO
 pycodestyle_status="${PIPESTATUS[0]}"
 
 if [ "$compile_status" -eq 0 -a "$pycodestyle_status" -eq 0 ]; then
-    lint_status=0
+    echo "pycodestyle checks passed."
+    rm "$PYCODESTYLE_REPORT_PATH"
 else
-    lint_status=1
-fi
-
-if [ "$lint_status" -ne 0 ]; then
     echo "pycodestyle checks failed."
     cat "$PYCODESTYLE_REPORT_PATH"
     rm "$PYCODESTYLE_REPORT_PATH"
-    exit "$lint_status"
+    exit 1
+fi
+
+# Check python document style, skip check if pydocstyle is not installed.
+if hash "$PYDOCSTYLEBUILD" 2> /dev/null; then
+    if [[ "$PYDOCSTYLEVERSION" == "$EXPECTED_PYDOCSTYLEVERSION" ]]; then
+        pydocstyle --config=dev/tox.ini $DOC_PATHS_TO_CHECK >> "$PYDOCSTYLE_REPORT_PATH"
+        pydocstyle_status="${PIPESTATUS[0]}"
+
+        if [ "$compile_status" -eq 0 -a "$pydocstyle_status" -eq 0 ]; then
+            echo "pydocstyle checks passed."
+            rm "$PYDOCSTYLE_REPORT_PATH"
+        else
+            echo "pydocstyle checks failed."
+            cat "$PYDOCSTYLE_REPORT_PATH"
+            rm "$PYDOCSTYLE_REPORT_PATH"
+            exit 1
+        fi
+
+    else
+        echo "The pydocstyle version needs to be latest 2.1.1. Skipping pydoc checks for now"
+    fi
 else
-    echo "pycodestyle checks passed."
-    rm "$PYCODESTYLE_REPORT_PATH"
+  echo >&2 "The pydocstyle command was not found. Skipping pydoc checks for now"
 fi
 
 # Check that the documentation builds acceptably, skip check if sphinx is not installed.

dev/tox.ini

@@ -17,3 +17,5 @@
 ignore=E402,E731,E241,W503,E226,E722,E741,E305
 max-line-length=100
 exclude=cloudpickle.py,heapq3.py,shared.py,python/docs/conf.py,work/*/*.py,python/.eggs/*
+[pydocstyle]
+ignore=D100,D101,D102,D103,D104,D105,D106,D107,D200,D201,D202,D203,D204,D205,D206,D207,D208,D209,D210,D211,D212,D213,D214,D215,D300,D301,D302,D400,D401,D402,D403,D404,D405,D406,D407,D408,D409,D410,D411,D412,D413,D414