!4250 fix cpp api iterator build function

Merge pull request !4250 from anthonyaje/fix_cpp_api_iterator

mindspore/ccsrc/minddata/dataset/api/iterator.cc

@@ -74,13 +74,20 @@ Status Iterator::BuildAndLaunchTree(std::shared_ptr<Dataset> ds) {
       RETURN_STATUS_UNEXPECTED("Node operation returned nothing");
     }
 
-    auto root_op = root_ops.front();
-
-    RETURN_UNEXPECTED_IF_NULL(root_op);
-
-    RETURN_IF_NOT_OK(tree_->AssociateNode(root_op));
-
-    q.push(std::make_pair(ds, root_op));
+    // Iterate through all the DatasetOps returned by Dataset's Build(), associate them
+    // with the execution tree and add the child and parent relationship between the nodes
+    // Note that some Dataset objects might return more than one DatasetOps
+    // e.g. MapDataset will return [ProjectOp, MapOp] if project_columns is set for MapDataset
+    std::shared_ptr<DatasetOp> prev_op = nullptr;
+    for (auto op : root_ops) {
+      RETURN_IF_NOT_OK(tree_->AssociateNode(op));
+      if (prev_op != nullptr) {
+        RETURN_IF_NOT_OK(prev_op->AddChild(op));
+      }
+      prev_op = op;
+    }
+    // Add the last DatasetOp to the queue to be BFS.
+    q.push(std::make_pair(ds, root_ops.back()));
 
     // Traverse down to the children and convert them to the corresponding DatasetOps (i.e. execution tree nodes)
     while (!q.empty()) {
@@ -107,7 +114,7 @@ Status Iterator::BuildAndLaunchTree(std::shared_ptr<Dataset> ds) {
         q.push(std::make_pair(child, child_ops.back()));
       }
     }
-    RETURN_IF_NOT_OK(tree_->AssignRoot(root_op));
+    RETURN_IF_NOT_OK(tree_->AssignRoot(root_ops.front()));
   }
 
   // Launch the execution tree.

tests/ut/cpp/dataset/c_api_dataset_ops_test.cc

@@ -425,6 +425,58 @@ TEST_F(MindDataTestPipeline, TestProjectMap) {
   iter->Stop();
 }
 
+
+TEST_F(MindDataTestPipeline, TestProjectMapAutoInjection) {
+  MS_LOG(INFO) << "Doing MindDataTestPipeline.TestProjectMapAutoInjection";
+
+  // Create an ImageFolder Dataset
+  std::string folder_path = datasets_root_path_ + "/testPK/data/";
+  std::shared_ptr<Dataset> ds = ImageFolder(folder_path, true, RandomSampler(false, 10));
+  EXPECT_NE(ds, nullptr);
+
+  // Create a Repeat operation on ds
+  int32_t repeat_num = 2;
+  ds = ds->Repeat(repeat_num);
+  EXPECT_NE(ds, nullptr);
+
+  // Create objects for the tensor ops
+  std::shared_ptr<TensorOperation> resize_op = vision::Resize({30, 30});
+  EXPECT_NE(resize_op, nullptr);
+
+  // Create a Map operation on ds
+  // {"image"} is the project columns. This will trigger auto injection of ProjectOp after MapOp.
+  ds = ds->Map({resize_op}, {}, {}, {"image"});
+  EXPECT_NE(ds, nullptr);
+
+  // Create an iterator over the result of the above dataset
+  // This will trigger the creation of the Execution Tree and launch it.
+  std::shared_ptr<Iterator> iter = ds->CreateIterator();
+  EXPECT_NE(iter, nullptr);
+
+  // Iterate the dataset and get each row
+  std::unordered_map<std::string, std::shared_ptr<Tensor>> row;
+  iter->GetNextRow(&row);
+
+  // 'label' is dropped during the project op
+  EXPECT_EQ(row.find("label"), row.end());
+  // 'image' column should still exist
+  EXPECT_NE(row.find("image"), row.end());
+
+  uint64_t i = 0;
+  while (row.size() != 0) {
+    i++;
+    auto image = row["image"];
+    MS_LOG(INFO) << "Tensor image shape: " << image->shape();
+    EXPECT_EQ(image->shape()[0], 30);
+    iter->GetNextRow(&row);
+  }
+
+  EXPECT_EQ(i, 20);
+
+  // Manually terminate the pipeline
+  iter->Stop();
+}
+
 TEST_F(MindDataTestPipeline, TestZipSuccess) {
   // Testing the member zip() function
   MS_LOG(INFO) << "Doing MindDataTestPipeline-TestZipSuccess.";