Delete trailing whitespace

PiperOrigin-RevId: 177008504

RELEASE.md

@@ -494,7 +494,7 @@ answered questions, and were part of inspiring discussions.
 This release contains contributions from many people at Google, as well as:
 
 A. Besir Kurtulmus, Adal Chiriliuc, @akash, Alec-Desouza, Alex Rothberg, Alex
-Sergeev, Alexander Heinecke, Allen Guo, Andreas Madsen, Ankesh Anand, Anton 
+Sergeev, Alexander Heinecke, Allen Guo, Andreas Madsen, Ankesh Anand, Anton
 Loss, @Aravind, @Arie, Ashutosh Das, AuréLien Geron, Bairen Yi, @bakunyo, Ben
 Visser, Brady Zhou, Calpa Liu, Changming Sun, Chih Cheng Liang, Christopher
 Berner, Clark Zinzow, @Conchylicultor, Dan Ellis, Dan J, Dan Jarvis, Daniel

tensorflow/compiler/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc

@@ -77,7 +77,7 @@ static string GetLibdeviceFilename(const string& libdevice_dir_path,
   // Since CUDA 9.0, all GPU versions are included in a single file
   const char* unified_libdevice_filename = "libdevice.10.bc";
   std::vector<string> unified_libdevice_files;
-  const tensorflow::Status status = 
+  const tensorflow::Status status =
     tensorflow::Env::Default()->GetMatchingPaths(
       tensorflow::io::JoinPath(libdevice_dir_path, unified_libdevice_filename),
       &unified_libdevice_files);

tensorflow/contrib/android/cmake/README.md

@@ -14,7 +14,7 @@ Add TensorFlow-Android-Inference as a dependency of your Android application
 
 ```
 include ':TensorFlow-Android-Inference'
-findProject(":TensorFlow-Android-Inference").projectDir = 
+findProject(":TensorFlow-Android-Inference").projectDir =
             new File("${/path/to/tensorflow_repo}/contrib/android/cmake")
 ```
 

tensorflow/contrib/android/java/org/tensorflow/contrib/android/TensorFlowInferenceInterface.java

@@ -160,15 +160,15 @@ public TensorFlowInferenceInterface(InputStream is) {
       throw new RuntimeException("Failed to load model from the input stream", e);
     }
   }
-  
+
   /*
    * Construct a TensorFlowInferenceInterface with provided Graph
    *
    * @param g The Graph to use to construct this interface.
    */
   public TensorFlowInferenceInterface(Graph g) {
     prepareNativeRuntime();
-      
+
     // modelName is redundant here, here is for
     // avoiding error in initialization as modelName is marked final.
     this.modelName = "";
@@ -290,7 +290,7 @@ protected void finalize() throws Throwable {
    */
   public void feed(String inputName, boolean[] src, long... dims) {
     byte[] b = new byte[src.length];
-    
+
     for (int i = 0; i < src.length; i++) {
       b[i] = src[i] ? (byte) 1 : (byte) 0;
     }

tensorflow/contrib/cloud/kernels/bigquery_table_accessor_test.cc

@@ -421,7 +421,7 @@ TEST_F(BigQueryTableAccessorTest, MultiplePagesTest) {
   TF_EXPECT_OK(accessor_->ReadRow(&row_id, &example));
   EXPECT_EQ(3, row_id);
   EXPECT_TRUE(accessor_->Done());
-  
+
   Example expected_example;
   ASSERT_TRUE(protobuf::TextFormat::ParseFromString(kTestExampleProtoWithNulls,
                                                     &expected_example));

tensorflow/contrib/cmake/tf_grappler.cmake

@@ -23,7 +23,7 @@ file(GLOB tf_grappler_srcs
    "${tensorflow_source_dir}/tensorflow/python/grappler/model_analyzer.cc"
    "${tensorflow_source_dir}/tensorflow/python/grappler/model_analyzer.h"
  )
- 
+
 add_library(tf_grappler OBJECT ${tf_grappler_srcs})
 
 add_dependencies(tf_grappler tf_core_cpu)

tensorflow/contrib/cmake/tf_shared_lib.cmake

@@ -45,7 +45,7 @@ if(WIN32)
       $<TARGET_FILE:tensorflow_static>
       $<TARGET_FILE:tf_protos_cc>
   )
-    
+
   set(tensorflow_deffile "${CMAKE_CURRENT_BINARY_DIR}/${CMAKE_BUILD_TYPE}/tensorflow.def")
   set_source_files_properties(${tensorflow_deffile} PROPERTIES GENERATED TRUE)
 

tensorflow/contrib/cmake/tf_stream_executor.cmake

@@ -61,18 +61,18 @@ file(GLOB tf_stream_executor_srcs
     "${tensorflow_source_dir}/tensorflow/stream_executor/platform/default/*.h"
 )
 
-if (tensorflow_ENABLE_GPU)    
+if (tensorflow_ENABLE_GPU)
     file(GLOB tf_stream_executor_gpu_srcs
         "${tensorflow_source_dir}/tensorflow/stream_executor/cuda/*.cc"
     )
     list(APPEND tf_stream_executor_srcs ${tf_stream_executor_gpu_srcs})
-endif()    
+endif()
 
 #file(GLOB_RECURSE tf_stream_executor_test_srcs
 #    "${tensorflow_source_dir}/tensorflow/stream_executor/*_test.cc"
 #    "${tensorflow_source_dir}/tensorflow/stream_executor/*_test.h"
 #)
-#list(REMOVE_ITEM tf_stream_executor_srcs ${tf_stream_executor_test_srcs}) 
+#list(REMOVE_ITEM tf_stream_executor_srcs ${tf_stream_executor_test_srcs})
 
 if (NOT WIN32)
   set (CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -lgomp")

tensorflow/contrib/image/ops/single_image_random_dot_stereograms_ops.cc

@@ -40,7 +40,7 @@ REGISTER_OP("SingleImageRandomDotStereograms")
     .Doc(R"doc(
 Outputs a single image random dot stereogram for export via encode_PNG/JPG OP.
 
-Given the 2-D tensor 'depth_values' with encoded Z values, this operation will 
+Given the 2-D tensor 'depth_values' with encoded Z values, this operation will
 encode 3-D data into a 2-D image.  The output of this Op is suitable for the
 encode_PNG/JPG ops.  Be careful with image compression as this may corrupt the
 encode 3-D data witin the image.
@@ -68,14 +68,14 @@ with open('picture_out.png', 'wb') as f:
     f.write(png)
 ```
 
-depth_values: Z values of data to encode into 'output_data_window' window, 
+depth_values: Z values of data to encode into 'output_data_window' window,
   lower values are further away {0.0 floor(far), 1.0 ceiling(near) after normalization}, must be 2-D tensor
 hidden_surface_removal: Activate hidden surface removal
 convergence_dots_size: Black dot size in pixels to help view converge image, drawn on bottom of image
 dots_per_inch: Output device in dots/inch
 eye_separation: Separation between eyes in inches
 mu: Depth of field, Fraction of viewing distance (eg. 1/3 = .3333)
-normalize: Normalize input data to [0.0, 1.0] 
+normalize: Normalize input data to [0.0, 1.0]
 normalize_max: Fix MAX value for Normalization - if < MIN, autoscale
 normalize_min: Fix MIN value for Normalization - if > MAX, autoscale
 border_level: Value of border depth 0.0 {far} to 1.0 {near}

tensorflow/contrib/lite/g3doc/apis.md

@@ -52,7 +52,7 @@ typedef enum {
 Failures can be easily verified with:
 ```c++
 if (status != kTfLiteOk) {
-  // ... error handling here ... 
+  // ... error handling here ...
 }
 ```
 

tensorflow/contrib/lite/java/demo/app/src/main/res/values/base-strings.xml

@@ -19,12 +19,12 @@
     <string name="app_name">TfLiteCameraDemo</string>
     <string name="intro_message">
         <![CDATA[
-        
-            
+
+
             This sample demonstrates the basic use of TfLite API. Check the source code to see how
             you can use TfLite for efficient, on-device inference with trained TensorFlow models.
-            
-        
+
+
         ]]>
     </string>
 </resources>

tensorflow/contrib/makefile/README.md

@@ -16,17 +16,17 @@ This static library will not contain:
 
  - Python or other language bindings
  - GPU support
- 
+
 You can target:
 - iOS
 - OS X (macOS)
 - Android
 - Raspberry-PI
- 
+
 You will compile tensorflow and protobuf libraries that you can link into other
 applications.  You will also compile the [benchmark](../../tools/benchmark/)
 application that will let you check your application.
- 
+
 ## Before you start (all platforms)
 
 First, clone this TensorFlow repository.
@@ -58,9 +58,9 @@ You should then be able to run the `build_all_linux.sh` script to compile:
 tensorflow/contrib/makefile/build_all_linux.sh
 ```
 
-This should compile a static library in 
-`tensorflow/contrib/makefile/gen/lib/libtensorflow-core.a`, 
-and create an example executable at `tensorflow/contrib/makefile/gen/bin/benchmark`. 
+This should compile a static library in
+`tensorflow/contrib/makefile/gen/lib/libtensorflow-core.a`,
+and create an example executable at `tensorflow/contrib/makefile/gen/bin/benchmark`.
 
 Get the graph file, if you have not already:
 
@@ -201,7 +201,7 @@ library in a simple app.
 ### Building by hand
 
 This section covers each step of building.  For all the code in one place, see
-[build_all_ios.sh](build_all_ios.sh). 
+[build_all_ios.sh](build_all_ios.sh).
 
 If you have not already, you will need to download dependencies:
 
@@ -232,7 +232,7 @@ make -f tensorflow/contrib/makefile/Makefile \
 
 This creates a library in
 `tensorflow/contrib/makefile/gen/lib/libtensorflow-core.a` that you can link any
-xcode project against. 
+xcode project against.
 
 To see TensorFlow running on iOS, the example Xcode project in
 [tensorflow/examples/ios](../../examples/ios/) shows how to use the static
@@ -258,15 +258,15 @@ tensorflow/contrib/makefile/compile_ios_tensorflow.sh -f "-O3" -h tensorflow/con
 
 In XCode, you will need to use -force_load in the linker flags
 section of the build settings to pull in the global constructors that are used
-to register ops and kernels. 
+to register ops and kernels.
 
 #### Optimization
- 
+
 The `compile_ios_tensorflow.sh` script can take optional command-line arguments.
 The first argument will be passed as a C++ optimization flag and defaults to
 debug mode. If you are concerned about performance or are working on a release
 build, you would likely want a higher optimization setting, like so:
- 
+
 ```bash
 compile_ios_tensorflow.sh -f "-Os"
 ```
@@ -330,7 +330,7 @@ what you need for your desired system.
 ## Dependency Management
 
 The Makefile loads in a list of dependencies stored in text files. These files
-are generated from the main Bazel build by running 
+are generated from the main Bazel build by running
 `tensorflow/contrib/makefile/gen_file_lists.sh`. You'll need to re-run this i
 you make changes to the files that are included in the build.
 
@@ -361,10 +361,10 @@ codebase can sometimes break the makefile build process. If you find that tests
 relying on this makefile are failing with a change you're involved in, here are
 some trouble-shooting steps:
 
- - Try to reproduce the issue on your platform. If you're on Linux, running 
+ - Try to reproduce the issue on your platform. If you're on Linux, running
  `make -f tensorflow/contrib/makefile/Makefile` should be enough to recreate
   most issues. For other platforms, see the sections earlier in this document.
-  
+
  - The most common cause of breakages are files that have been added to the
   Bazel build scripts, but that the makefile isn't aware of. Typical symptoms
   of this include linker errors mentioning missing symbols or protobuf headers
@@ -377,11 +377,11 @@ some trouble-shooting steps:
   `tensorflow/core/BUILD`, so if you change the wildcards there to include new
   files you'll need to also update `CORE_CC_ALL_SRCS` and `CORE_CC_EXCLUDE_SRCS`
   in the makefile.
-  
+
  - Some of the supported platforms use clang instead of gcc as their compiler,
   so if you're hitting compile errors you may need to tweak your code to be more
   friendly to different compilers by avoiding gcc extensions or idioms.
-  
+
 These are the most common reasons for makefile breakages, but it's also
 possible you may hit something unusual, like a platform incompatibility. For
 those, you'll need to see if you can reproduce the issue on that particular

tensorflow/contrib/makefile/compile_ios_protobuf.sh

@@ -270,7 +270,7 @@ case "$1" in
         echo "Unknown ARCH"
         exit 1
         ;;
-esac 
+esac
 }
 
 for build_element in "${build_targets[@]}"

tensorflow/contrib/makefile/compile_nsync.sh

@@ -28,7 +28,7 @@ usage="usage: $prog [-t linux|ios|android|macos|native]
         [-a architecture] [-v android_api_version]
 
 A script to build nsync for tensorflow.
-This script can be run on Linux or MacOS host platforms, and can target 
+This script can be run on Linux or MacOS host platforms, and can target
 Linux, MacOS, iOS, or Android.
 
 Options:

tensorflow/contrib/makefile/rename_protobuf.sh

@@ -38,7 +38,7 @@
 #
 # Note that this script modifies the source code in-place, so once it's been run
 # it's no longer suitable for further manual modifications, since the difference
-# with the top of tree will already be large. 
+# with the top of tree will already be large.
 
 mv tensorflow/contrib/makefile/downloads/protobuf/src/google/protobuf \
  tensorflow/contrib/makefile/downloads/protobuf//src/google/protobuf3
@@ -71,7 +71,7 @@ sed -i '' 's%::google::protobuf;%google::protobuf3;%' \
 
 # Fix up a couple of special build scripts that look for particular files.
 sed -i '' 's%src/google/protobuf/message.cc%src/google/protobuf3/message.cc%' \
- tensorflow/contrib/makefile/downloads/protobuf/configure.ac 
+ tensorflow/contrib/makefile/downloads/protobuf/configure.ac
 sed -i '' 's%src/google/protobuf/stubs/common.h%src/google/protobuf3/stubs/common.h%' \
  tensorflow/contrib/makefile/downloads/protobuf/autogen.sh
 

tensorflow/contrib/metrics/README.md

@@ -4,7 +4,7 @@
 
 Metrics are used in evaluation to assess the quality of a model. Most are
 "streaming" ops, meaning they create variables to accumulate a running total,
-and return an update tensor to update these variables, and a value tensor to 
+and return an update tensor to update these variables, and a value tensor to
 read the accumulated value. Example:
 
 value, update_op = metrics.streaming_mean_squared_error(

tensorflow/contrib/mpi/README.md

@@ -23,7 +23,7 @@ The following environment variables can be set to modify the behavior at runtime
 
 **MPI_DISABLED=[0,1]**
 
-This environment variable allows you to disable the MPI path before launch (e.g. for performance or correctness testing). 
+This environment variable allows you to disable the MPI path before launch (e.g. for performance or correctness testing).
 
 **MPI_OPTIMAL_PATH=[0,1]**
 
@@ -34,10 +34,10 @@ This path is disabled by default as it requires that the MPI library can directl
 
 ## Known problems
 
-For certain complex neural nets the implementation sometimes crashes inside the MPI libraries. This seems to be related to memory allocations/routines that register the memory for the Infiniband transfers. (The crashes do not happen when all MPI processes are within the same physical machine). 
+For certain complex neural nets the implementation sometimes crashes inside the MPI libraries. This seems to be related to memory allocations/routines that register the memory for the Infiniband transfers. (The crashes do not happen when all MPI processes are within the same physical machine).
 
 **MVAPICH**
-- The problem manifests itself with a segmentation fault inside a memory copy routine and during startup you will get the following warning: "WARNING: Error in initializing MVAPICH2 ptmalloc library. Continuing without InfiniBand registration cache support." 
+- The problem manifests itself with a segmentation fault inside a memory copy routine and during startup you will get the following warning: "WARNING: Error in initializing MVAPICH2 ptmalloc library. Continuing without InfiniBand registration cache support."
 
 **OpenMPI**
 - With OpenMPI corrupt data will be received resulting in an assertion or the MPI library will print an error and exit. The error is "Attempt to free memory that is still in use by an ongoing MPI communication.  MPI job will now abort."
@@ -58,11 +58,11 @@ Once a request has arrived from a remote process the request is forwarded to the
 * Receive tensor request
 The MPI thread will check if there are any incoming tensor request messages on the communication lines using MPI_Iprobe. Once a request has been received it will be passed on to the standard TensorFlow code and eventually will be placed on the sendQueue.
 
-* Receive tensor 
+* Receive tensor
 At some point after a request has been sent the remote process will transmit the tensor. This tensor will be received and we look-up the callback that is associated with this tensor in our request table and execute the callback on the received data.
 
 
-In the implementation all send operations are non-blocking, all probe operations are non-blocking and all receive-operations are blocking. The receive-operations are only executed after the probe has determined that there is something to receive. 
+In the implementation all send operations are non-blocking, all probe operations are non-blocking and all receive-operations are blocking. The receive-operations are only executed after the probe has determined that there is something to receive.
 The MPI processes identify each other using an MPI process ID. The TensorFlow gRPC processes identify each other using a name. During launch we create a mapping between the TensorFlow process name and the MPI process ID to allow the processes to communicate with the correct destinations when using MPI operations.
 
 

tensorflow/contrib/pi_examples/README.md

@@ -13,7 +13,7 @@ sudo apt-get install -y libjpeg-dev
 ```
 
  - To download the example model you'll need, run these commands:
- 
+
 ```bash
 curl https://storage.googleapis.com/download.tensorflow.org/models/inception_dec_2015_stripped.zip \
 -o /tmp/inception_dec_2015_stripped.zip

tensorflow/contrib/pi_examples/camera/Makefile

@@ -76,7 +76,7 @@ $(EXECUTABLE_NAME): $(EXECUTABLE_OBJS) $(TFLIBS)
 	$(LIBFLAGS) $(LIB_PATH) $(LDFLAGS) $(LIBS)
 
 # Matches on C++ source files.
-$(OBJDIR)%.o: %.cc 
+$(OBJDIR)%.o: %.cc
 	@mkdir -p $(dir $@)
 	$(CXX) $(CXXFLAGS) $(INCLUDES) -c $< -o $@
 

tensorflow/contrib/pi_examples/label_image/Makefile

@@ -75,7 +75,7 @@ $(EXECUTABLE_NAME): $(EXECUTABLE_OBJS) $(TFLIBS)
 	$(LIBFLAGS) $(LIB_PATH) $(LDFLAGS) $(LIBS)
 
 # Matches on C++ source files.
-$(OBJDIR)%.o: %.cc 
+$(OBJDIR)%.o: %.cc
 	@mkdir -p $(dir $@)
 	$(CXX) $(CXXFLAGS) $(INCLUDES) -c $< -o $@