[SPARK-19221][PROJECT INFRA][R] Add winutils binaries to the path in AppVeyor tests for Hadoop libraries to call native codes properly

## What changes were proposed in this pull request?

It seems Hadoop libraries need winutils binaries for native libraries in the path.

It is not a problem in tests for now because we are only testing SparkR on Windows via AppVeyor but it can be a problem if we run Scala tests via AppVeyor as below:

```
 - SPARK-18220: read Hive orc table with varchar column *** FAILED *** (3 seconds, 937 milliseconds)
   org.apache.spark.sql.execution.QueryExecutionException: FAILED: Execution Error, return code -101 from org.apache.hadoop.hive.ql.exec.mr.MapRedTask. org.apache.hadoop.io.nativeio.NativeIO$Windows.access0(Ljava/lang/String;I)Z
   at org.apache.spark.sql.hive.client.HiveClientImpl$$anonfun$runHive$1.apply(HiveClientImpl.scala:625)
   at org.apache.spark.sql.hive.client.HiveClientImpl$$anonfun$runHive$1.apply(HiveClientImpl.scala:609)
   at org.apache.spark.sql.hive.client.HiveClientImpl$$anonfun$withHiveState$1.apply(HiveClientImpl.scala:283)
   ...
```

This PR proposes to add it to the `Path` for AppVeyor tests.

## How was this patch tested?

Manually via AppVeyor.

**Before**
https://ci.appveyor.com/project/spark-test/spark/build/549-windows-complete/job/gc8a1pjua2bc4i8m

**After**
https://ci.appveyor.com/project/spark-test/spark/build/572-windows-complete/job/c4vrysr5uvj2hgu7

Author: hyukjinkwon <gurwls223@gmail.com>

Closes #16584 from HyukjinKwon/set-path-appveyor.