aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rwxr-xr-xdev/lint-python44
-rwxr-xr-xec2/spark_ec2.py4
2 files changed, 29 insertions, 19 deletions
diff --git a/dev/lint-python b/dev/lint-python
index 772f856154..fded654893 100755
--- a/dev/lint-python
+++ b/dev/lint-python
@@ -19,43 +19,53 @@
SCRIPT_DIR="$( cd "$( dirname "$0" )" && pwd )"
SPARK_ROOT_DIR="$(dirname "$SCRIPT_DIR")"
-PEP8_REPORT_PATH="$SPARK_ROOT_DIR/dev/pep8-report.txt"
+PATHS_TO_CHECK="./python/pyspark/ ./ec2/spark_ec2.py ./examples/src/main/python/"
+PYTHON_LINT_REPORT_PATH="$SPARK_ROOT_DIR/dev/python-lint-report.txt"
cd "$SPARK_ROOT_DIR"
+# compileall: https://docs.python.org/2/library/compileall.html
+python -B -m compileall -q -l $PATHS_TO_CHECK > "$PYTHON_LINT_REPORT_PATH"
+compile_status="${PIPESTATUS[0]}"
+
# Get pep8 at runtime so that we don't rely on it being installed on the build server.
#+ See: https://github.com/apache/spark/pull/1744#issuecomment-50982162
#+ TODOs:
-#+ - Dynamically determine latest release version of pep8 and use that.
-#+ - Download this from a more reliable source. (GitHub raw can be flaky, apparently. (?))
+#+ - Download pep8 from PyPI. It's more "official".
PEP8_SCRIPT_PATH="$SPARK_ROOT_DIR/dev/pep8.py"
-PEP8_SCRIPT_REMOTE_PATH="https://raw.githubusercontent.com/jcrocholl/pep8/1.5.7/pep8.py"
-PEP8_PATHS_TO_CHECK="./python/pyspark/ ./ec2/spark_ec2.py ./examples/src/main/python/"
+PEP8_SCRIPT_REMOTE_PATH="https://raw.githubusercontent.com/jcrocholl/pep8/1.6.2/pep8.py"
+# if [ ! -e "$PEP8_SCRIPT_PATH" ]; then
curl --silent -o "$PEP8_SCRIPT_PATH" "$PEP8_SCRIPT_REMOTE_PATH"
-curl_status=$?
+curl_status="$?"
-if [ $curl_status -ne 0 ]; then
+if [ "$curl_status" -ne 0 ]; then
echo "Failed to download pep8.py from \"$PEP8_SCRIPT_REMOTE_PATH\"."
- exit $curl_status
+ exit "$curl_status"
fi
-
+# fi
# There is no need to write this output to a file
#+ first, but we do so so that the check status can
#+ be output before the report, like with the
#+ scalastyle and RAT checks.
-python "$PEP8_SCRIPT_PATH" $PEP8_PATHS_TO_CHECK > "$PEP8_REPORT_PATH"
-pep8_status=${PIPESTATUS[0]} #$?
+python "$PEP8_SCRIPT_PATH" --ignore=E402,E731,E241,W503,E226 $PATHS_TO_CHECK >> "$PYTHON_LINT_REPORT_PATH"
+pep8_status="${PIPESTATUS[0]}"
+
+if [ "$compile_status" -eq 0 -a "$pep8_status" -eq 0 ]; then
+ lint_status=0
+else
+ lint_status=1
+fi
-if [ $pep8_status -ne 0 ]; then
- echo "PEP 8 checks failed."
- cat "$PEP8_REPORT_PATH"
+if [ "$lint_status" -ne 0 ]; then
+ echo "Python lint checks failed."
+ cat "$PYTHON_LINT_REPORT_PATH"
else
- echo "PEP 8 checks passed."
+ echo "Python lint checks passed."
fi
-rm "$PEP8_REPORT_PATH"
rm "$PEP8_SCRIPT_PATH"
+rm "$PYTHON_LINT_REPORT_PATH"
-exit $pep8_status
+exit "$lint_status"
diff --git a/ec2/spark_ec2.py b/ec2/spark_ec2.py
index f848874b0c..c467cd08ed 100755
--- a/ec2/spark_ec2.py
+++ b/ec2/spark_ec2.py
@@ -1159,8 +1159,8 @@ def real_main():
if EC2_INSTANCE_TYPES[opts.instance_type] != \
EC2_INSTANCE_TYPES[opts.master_instance_type]:
print >> stderr, \
- "Error: spark-ec2 currently does not support having a master and slaves with " + \
- "different AMI virtualization types."
+ "Error: spark-ec2 currently does not support having a master and slaves " + \
+ "with different AMI virtualization types."
print >> stderr, "master instance virtualization type: {t}".format(
t=EC2_INSTANCE_TYPES[opts.master_instance_type])
print >> stderr, "slave instance virtualization type: {t}".format(