[SPARK-5466] Add explicit guava dependencies where needed.

One side-effect of shading guava is that it disappears as a transitive
dependency. For Hadoop 2.x, this was masked by the fact that Hadoop
itself depends on guava. But certain versions of Hadoop 1.x also
shade guava, leaving either no guava or some random version pulled
by another dependency on the classpath.

So be explicit about the dependency in modules that use guava directly,
which is the right thing to do anyway.

Author: Marcelo Vanzin <vanzin@cloudera.com>

Closes #4272 from vanzin/SPARK-5466 and squashes the following commits:

e3f30e5 [Marcelo Vanzin] Dependency for catalyst is not needed.
d3b2c84 [Marcelo Vanzin] [SPARK-5466] Add explicit guava dependencies where needed.

core/pom.xml

@@ -34,6 +34,10 @@
   <name>Spark Project Core</name>
   <url>http://spark.apache.org/</url>
   <dependencies>
+    <dependency>
+      <groupId>com.google.guava</groupId>
+      <artifactId>guava</artifactId>
+    </dependency>
     <dependency>
       <groupId>com.twitter</groupId>
       <artifactId>chill_${scala.binary.version}</artifactId>

graphx/pom.xml

@@ -40,6 +40,10 @@
       <artifactId>spark-core_${scala.binary.version}</artifactId>
       <version>${project.version}</version>
     </dependency>
+    <dependency>
+      <groupId>com.google.guava</groupId>
+      <artifactId>guava</artifactId>
+    </dependency>
     <dependency>
       <groupId>org.jblas</groupId>
       <artifactId>jblas</artifactId>

streaming/pom.xml

@@ -40,6 +40,10 @@
       <artifactId>spark-core_${scala.binary.version}</artifactId>
       <version>${project.version}</version>
     </dependency>
+    <dependency>
+      <groupId>com.google.guava</groupId>
+      <artifactId>guava</artifactId>
+    </dependency>
     <dependency>
       <groupId>org.eclipse.jetty</groupId>
       <artifactId>jetty-server</artifactId>