commit f33dce3186b850351994a807941c4bc8ef741159
author Honglin Yu <honglinyu@chromium.org> Thu Dec 05 15:10:39 2019 +1100
committer Commit Bot <commit-bot@chromium.org> Wed Apr 29 21:55:59 2020 +0000
tree 02bb2d74f5f58fe904aa5358bf30d2d6471cf22e
parent d350890a615dd2566224e47f0a8df752d3ef5e0f

ml: Add mojo API of TextClassifier/Annotator.

This CL implements the mojo API of text classifier.

BUG=chromium:1020419
TEST=pass the existing unit tests.
TEST=on device (eve), 1+2=3 works.
TEST=on device (eve), Annotate() and SuggestSelection() work.

Change-Id: Iadfe5747f3b2d1f8c8a8d9f0b975e2e8cda50726
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform2/+/1951350
Tested-by: Honglin Yu <honglinyu@chromium.org>
Commit-Queue: Honglin Yu <honglinyu@chromium.org>
Reviewed-by: Sam McNally <sammc@chromium.org>
Reviewed-by: Andrew Moylan <amoylan@chromium.org>

ml/machine_learning_service_impl.cc

diff --git a/ml/machine_learning_service_impl.cc b/ml/machine_learning_service_impl.cc
index 62b2f81..efcf64c 100644
--- a/ml/machine_learning_service_impl.cc
+++ b/ml/machine_learning_service_impl.cc
@@ -10,10 +10,16 @@
 
 #include <base/bind.h>
 #include <base/bind_helpers.h>
+#include <base/files/file.h>
+#include <base/files/file_util.h>
 #include <tensorflow/lite/model.h>
+#include <unicode/putil.h>
+#include <unicode/udata.h>
+#include <utils/memory/mmap.h>
 
 #include "ml/model_impl.h"
 #include "ml/mojom/model.mojom.h"
+#include "ml/text_classifier_impl.h"
 
 namespace ml {
 
@@ -30,18 +36,25 @@
 // or |LoadFlatBufferModel|) requests
 constexpr char kMetricsRequestName[] = "LoadModelResult";
 
+constexpr char kTextClassifierModelFile[] =
+    "mlservice-model-text_classifier_en-v706.fb";
+
+constexpr char kIcuDataFilePath[] = "/opt/google/chrome/icudtl.dat";
+
 }  // namespace
 
 MachineLearningServiceImpl::MachineLearningServiceImpl(
     mojo::ScopedMessagePipeHandle pipe,
     base::Closure connection_error_handler,
     const std::string& model_dir)
-    : builtin_model_metadata_(GetBuiltinModelMetadata()),
+    : icu_data_(nullptr),
+      text_classifier_model_filename_(kTextClassifierModelFile),
+      builtin_model_metadata_(GetBuiltinModelMetadata()),
       model_dir_(model_dir),
       binding_(this,
                mojo::InterfaceRequest<
                    chromeos::machine_learning::mojom::MachineLearningService>(
-                       std::move(pipe))) {
+                   std::move(pipe))) {
   binding_.set_connection_error_handler(std::move(connection_error_handler));
 }
 
@@ -51,6 +64,11 @@
                                  std::move(connection_error_handler),
                                  kSystemModelDir) {}
 
+void MachineLearningServiceImpl::SetTextClassifierModelFilenameForTesting(
+    const std::string& filename) {
+  text_classifier_model_filename_ = filename;
+}
+
 void MachineLearningServiceImpl::LoadBuiltinModel(
     BuiltinModelSpecPtr spec,
     ModelRequest request,
@@ -132,4 +150,59 @@
   request_metrics.RecordRequestEvent(LoadModelResult::OK);
 }
 
+void MachineLearningServiceImpl::LoadTextClassifier(
+    chromeos::machine_learning::mojom::TextClassifierRequest request,
+    LoadTextClassifierCallback callback) {
+  RequestMetrics<LoadModelResult> request_metrics("TextClassifier",
+                                                  kMetricsRequestName);
+  request_metrics.StartRecordingPerformanceMetrics();
+
+  // Attempt to load model.
+  std::string model_path = model_dir_ + text_classifier_model_filename_;
+  auto scoped_mmap =
+      std::make_unique<libtextclassifier3::ScopedMmap>(model_path);
+  if (!scoped_mmap->handle().ok()) {
+    LOG(ERROR) << "Failed to load the text classifier model file '"
+               << model_path << "'.";
+    std::move(callback).Run(LoadModelResult::LOAD_MODEL_ERROR);
+    request_metrics.RecordRequestEvent(LoadModelResult::LOAD_MODEL_ERROR);
+    return;
+  }
+
+  // Create the TextClassifier.
+  if (!TextClassifierImpl::Create(&scoped_mmap, std::move(request))) {
+    LOG(ERROR) << "Failed to create TextClassifierImpl object.";
+    std::move(callback).Run(LoadModelResult::LOAD_MODEL_ERROR);
+    request_metrics.RecordRequestEvent(LoadModelResult::LOAD_MODEL_ERROR);
+    return;
+  }
+
+  // initialize the icu library.
+  InitIcuIfNeeded();
+
+  std::move(callback).Run(LoadModelResult::OK);
+
+  request_metrics.FinishRecordingPerformanceMetrics();
+  request_metrics.RecordRequestEvent(LoadModelResult::OK);
+}
+
+void MachineLearningServiceImpl::InitIcuIfNeeded() {
+  if (icu_data_ == nullptr) {
+    // Need to load the data file again.
+    int64_t file_size;
+    const base::FilePath icu_data_file_path(kIcuDataFilePath);
+    CHECK(base::GetFileSize(icu_data_file_path, &file_size));
+    icu_data_ = new char[file_size];
+    CHECK(base::ReadFile(icu_data_file_path, icu_data_,
+                         static_cast<int>(file_size)) == file_size);
+    // Init the Icu library.
+    UErrorCode err = U_ZERO_ERROR;
+    udata_setCommonData(reinterpret_cast<void*>(icu_data_), &err);
+    DCHECK(err == U_ZERO_ERROR);
+    // Never try to load Icu data from files.
+    udata_setFileAccess(UDATA_ONLY_PACKAGES, &err);
+  }
+}
+
+
 }  // namespace ml