Revert the workaround to classpath issues caused by google-java-forma…

…t bundling but not relocating their copy of guava, which affected 2.3. The issue is no longer there, and the tests should fail if it regresses.

relates to #288 and was partly resolved by google/google-java-format#20 and depending on the newer released (unshaded) formatter.
-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=120469748

compiler/src/it/functional-tests/pom.xml

@@ -28,10 +28,6 @@ limitations under the License.
   <artifactId>functional-tests</artifactId>
   <name>Functional Tests</name>
   <dependencies>
-    <dependency>
-      <groupId>com.google.guava</groupId>
-      <artifactId>guava</artifactId>
-    </dependency>
     <dependency>
       <groupId>com.google.dagger</groupId>
       <artifactId>dagger</artifactId>

compiler/src/it/tck/pom.xml

@@ -28,11 +28,6 @@ limitations under the License.
   <artifactId>tck</artifactId>
   <name>TCK</name>
   <dependencies>
-    <dependency>
-      <!-- Force the correct version of Guava to be on the classpath. -->
-      <groupId>com.google.guava</groupId>
-      <artifactId>guava</artifactId>
-    </dependency>
     <dependency>
       <groupId>com.google.dagger</groupId>
       <artifactId>dagger</artifactId>