spark-instrumented-optimizer/dev/sparktestsupport
hyukjinkwon ff8de99a1c [SPARK-22302][INFRA] Remove manual backports for subprocess and print explicit message for < Python 2.7
## What changes were proposed in this pull request?

Seems there was a mistake - missing import for `subprocess.call`, while refactoring this script a long ago, which should be used for backports of some missing functions in `subprocess`, specifically in < Python 2.7.

Reproduction is:

```
cd dev && python2.6
```

```
>>> from sparktestsupport import shellutils
>>> shellutils.subprocess_check_call("ls")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "sparktestsupport/shellutils.py", line 46, in subprocess_check_call
    retcode = call(*popenargs, **kwargs)
NameError: global name 'call' is not defined
```

For Jenkins logs, please see https://amplab.cs.berkeley.edu/jenkins/job/NewSparkPullRequestBuilder/3950/console

Since we dropped the Python 2.6.x support, looks better we remove those workarounds and print out explicit error messages in order to reduce the efforts to find out the root causes for such cases, for example, `https://github.com/apache/spark/pull/19513#issuecomment-337406734`.

## How was this patch tested?

Manually tested:

```
./dev/run-tests
```

```
Python versions prior to 2.7 are not supported.
```

```
./dev/run-tests-jenkins
```

```
Python versions prior to 2.7 are not supported.
```

Author: hyukjinkwon <gurwls223@gmail.com>

Closes #19524 from HyukjinKwon/SPARK-22302.
2017-10-22 02:22:35 +09:00
..
__init__.py [SPARK-1267][SPARK-18129] Allow PySpark to be pip installed 2016-11-16 14:22:15 -08:00
modules.py [SPARK-22142][BUILD][STREAMING] Move Flume support behind a profile, take 2 2017-10-06 15:08:28 +01:00
shellutils.py [SPARK-22302][INFRA] Remove manual backports for subprocess and print explicit message for < Python 2.7 2017-10-22 02:22:35 +09:00
toposort.py [SPARK-8725][PROJECT-INFRA] Test modules in topologically-sorted order in dev/run-tests 2016-01-26 14:20:11 -08:00