Support extracting entity string from string with UTF-16 characters.

Bug: b/157779488
Test: unit test
Change-Id: Ib7626b9c78c197f2ec2b7e7b7eac489994262893
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2223374
Commit-Queue: Li Lin <llin@chromium.org>
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#773754}

chromeos/components/quick_answers/understanding/intent_generator.cc

@@ -7,6 +7,7 @@
 #include <map>
 
 #include "base/no_destructor.h"
+#include "base/strings/utf_string_conversions.h"
 #include "chromeos/components/quick_answers/quick_answers_model.h"
 #include "chromeos/components/quick_answers/utils/language_detector.h"
 #include "chromeos/constants/chromeos_features.h"
@@ -38,9 +39,13 @@ bool ExtractEntity(const std::string& selected_text,
                    std::string* entity_str,
                    std::string* type) {
   for (auto& annotation : annotations) {
-    *entity_str =
-        selected_text.substr(annotation->start_offset,
-                             annotation->end_offset - annotation->start_offset);
+    // The offset in annotation result is by chars instead of by bytes. Converts
+    // to string16 to support extracting substring from string with UTF-16
+    // characters.
+    *entity_str = base::UTF16ToUTF8(
+        base::UTF8ToUTF16(selected_text)
+            .substr(annotation->start_offset,
+                    annotation->end_offset - annotation->start_offset));
 
     // Use the first entity type.
     auto intent_type_map = GetIntentTypeMap();

chromeos/components/quick_answers/understanding/intent_generator_unittest.cc

@@ -241,6 +241,35 @@ TEST_F(IntentGeneratorTest, TextAnnotationUnitIntentExtraChars) {
   EXPECT_EQ("23 cm", intent_text_);
 }
 
+TEST_F(IntentGeneratorTest, TextAnnotationUnitIntentUtf16Char) {
+  std::unique_ptr<QuickAnswersRequest> quick_answers_request =
+      std::make_unique<QuickAnswersRequest>();
+  quick_answers_request->selected_text = "350°F";
+
+  // Create the test annotations.
+  std::vector<TextEntityPtr> entities;
+  entities.emplace_back(
+      TextEntity::New("unit",                   // Entity name.
+                      1.0,                      // Confidence score.
+                      TextEntityData::New()));  // Data extracted.
+
+  auto dictionary_annotation = TextAnnotation::New(0,  // Start offset.
+                                                   5,  // End offset.
+                                                   std::move(entities));
+
+  std::vector<TextAnnotationPtr> annotations;
+  annotations.push_back(dictionary_annotation->Clone());
+
+  UseFakeServiceConnection(annotations);
+
+  intent_generator_->GenerateIntent(*quick_answers_request);
+
+  task_environment_.RunUntilIdle();
+
+  EXPECT_EQ(IntentType::kUnit, intent_type_);
+  EXPECT_EQ("350°F", intent_text_);
+}
+
 TEST_F(IntentGeneratorTest, TextAnnotationUnitIntentExtraCharsAboveThreshold) {
   std::unique_ptr<QuickAnswersRequest> quick_answers_request =
       std::make_unique<QuickAnswersRequest>();