Clean up the view code

mindspore/ccsrc/minddata/dataset/engine/datasetops/map_op/map_op.cc

@@ -121,8 +121,6 @@ Status MapOp::GenerateWorkerJob(const std::unique_ptr<MapWorkerJob> *worker_job)
   for (size_t i = 0; i < tfuncs_.size(); i++) {
     // Currently we only have CPU as the device target
     // In the future, we will have heuristic or control from user to select target device
-    // MapTargetDevice target_device;
-    // RETURN_IF_NOT_OK(SelectTarget(tfuncs_[i], &target_device));
     MapTargetDevice target_device = MapTargetDevice::kCpu;
 
     switch (target_device) {
@@ -222,7 +220,6 @@ Status MapOp::operator()() {
     RETURN_IF_NOT_OK(child_[0]->GetNextBuffer(&buff, 0));
   }
   // End() is commented out because it might never be called due to the lack of EOF when EpochCtrl is -1
-  // RETURN_IF_NOT_OK(callback_manager_.End(CallbackParam(op_current_epochs_, ep_step, total_step)));
   // Handle eof logic, this code might never be reached if epoch_ctrl = -1.
   std::unique_ptr<MapWorkerJob> worker_job = std::make_unique<MapWorkerJob>(std::move(buff));
   RETURN_IF_NOT_OK(local_queues_[num_buf++ % num_workers_]->Add(std::move(worker_job)));

mindspore/ccsrc/minddata/dataset/engine/datasetops/source/sampler/sampler.cc

@@ -155,8 +155,6 @@ Status Sampler::AddChild(std::shared_ptr<Sampler> child) {
 
   child_.push_back(child);
 
-  // doesn't work, protected?
-  // child->AddParent(this);
   return Status::OK();
 }
 

mindspore/ccsrc/minddata/dataset/engine/gnn/graph_data_server.cc

@@ -51,7 +51,6 @@ Status GraphDataServer::Init() {
 #else
   set_state(kGdsInitializing);
   RETURN_IF_NOT_OK(async_server_->Run());
-  // RETURN_IF_NOT_OK(InitGraphDataImpl());
   RETURN_IF_NOT_OK(tg_->CreateAsyncTask("init graph data impl", std::bind(&GraphDataServer::InitGraphDataImpl, this)));
   for (int32_t i = 0; i < num_workers_; ++i) {
     RETURN_IF_NOT_OK(

mindspore/ccsrc/minddata/dataset/engine/gnn/graph_data_service_impl.cc

@@ -101,7 +101,6 @@ grpc::Status GraphDataServiceImpl::ClientUnRegister(grpc::ServerContext *context
 
 grpc::Status GraphDataServiceImpl::GetGraphData(grpc::ServerContext *context, const GnnGraphDataRequestPb *request,
                                                 GnnGraphDataResponsePb *response) {
-  // MS_LOG(INFO) << "#### receive GetGraphData:" << request->op_name();
   Status s;
   auto iter = g_get_graph_data_func_.find(request->op_name());
   if (iter != g_get_graph_data_func_.end()) {
@@ -115,7 +114,6 @@ grpc::Status GraphDataServiceImpl::GetGraphData(grpc::ServerContext *context, co
   } else {
     response->set_error_msg("Invalid op name.");
   }
-  // MS_LOG(INFO) << "#### end receive GetGraphData:" << request->op_name();
   return ::grpc::Status::OK;
 }
 

mindspore/ccsrc/minddata/dataset/kernels/image/lite_cv/lite_mat.cc

@@ -40,6 +40,7 @@ LiteMat::LiteMat(int width, LDataType data_type) {
   dims_ = 0;
   data_type_ = LDataType::UINT8;
   ref_count_ = 0;
+  size_ = 0;
   Init(width, data_type);
 }
 
@@ -53,6 +54,7 @@ LiteMat::LiteMat(int width, int height, LDataType data_type) {
   dims_ = 0;
   data_type_ = LDataType::UINT8;
   ref_count_ = 0;
+  size_ = 0;
   Init(width, height, data_type);
 }
 
@@ -66,6 +68,7 @@ LiteMat::LiteMat(int width, int height, int channel, LDataType data_type) {
   dims_ = 0;
   data_type_ = LDataType::UINT8;
   ref_count_ = 0;
+  size_ = 0;
   Init(width, height, channel, data_type);
 }
 
@@ -87,6 +90,7 @@ LiteMat::LiteMat(const LiteMat &m) {
   dims_ = m.dims_;
   data_type_ = m.data_type_;
   ref_count_ = m.ref_count_;
+  size_ = 0;
   if (ref_count_) {
     addRef(ref_count_, 1);
   }
@@ -111,6 +115,7 @@ LiteMat &LiteMat::operator=(const LiteMat &m) {
   dims_ = m.dims_;
   data_type_ = m.data_type_;
   ref_count_ = m.ref_count_;
+  size_ = 0;
   return *this;
 }
 
@@ -180,6 +185,7 @@ void LiteMat::Release() {
   channel_ = 0;
   c_step_ = 0;
   ref_count_ = 0;
+  size_ = 0;
 }
 
 void *LiteMat::AlignMalloc(unsigned int size) {

mindspore/ccsrc/minddata/mindrecord/include/shard_column.h

@@ -48,7 +48,6 @@ enum ColumnDataType {
   ColumnNoDataType = 6
 };
 
-// mapping as {"bytes", "string", "int32", "int64", "float32", "float64"};
 const uint32_t ColumnDataTypeSize[kDataTypes] = {1, 1, 4, 8, 4, 8};
 
 const std::vector<std::string> ColumnDataTypeNameNormalized = {"uint8", "string",  "int32",

mindspore/ccsrc/minddata/mindrecord/include/shard_reader.h

@@ -199,9 +199,6 @@ class ShardReader {
   /// \return null
   void SetAllInIndex(bool all_in_index) { all_in_index_ = all_in_index; }
 
-  /// \brief get NLP flag
-  bool GetNlpFlag();
-
   /// \brief get all classes
   MSRStatus GetAllClasses(const std::string &category_field, std::set<std::string> &categories);
 
@@ -264,9 +261,6 @@ class ShardReader {
   MSRStatus CreateTasks(const std::vector<std::tuple<int, int, int, uint64_t>> &row_group_summary,
                         const std::vector<std::shared_ptr<ShardOperator>> &operators);
 
-  /// \brief set NLP flag
-  void CheckNlp();
-
   /// \brief check if all specified columns are in index table
   void CheckIfColumnInIndex(const std::vector<std::string> &columns);
 

mindspore/ccsrc/minddata/mindrecord/io/shard_index_generator.cc

@@ -257,7 +257,9 @@ std::pair<MSRStatus, sqlite3 *> ShardIndexGenerator::CreateDatabase(int shard_no
     sql += ",INC_" + std::to_string(field_no++) + " INT, " + ret.second + " " + type;
   }
   sql += ", PRIMARY KEY(ROW_ID";
-  for (uint64_t i = 0; i < fields_.size(); ++i) sql += ",INC_" + std::to_string(i);
+  for (uint64_t i = 0; i < fields_.size(); ++i) {
+    sql += ",INC_" + std::to_string(i);
+  }
   sql += "));";
   if (ExecuteSQL(sql, db, "create table successfully.") != SUCCESS) {
     return {FAILED, nullptr};

mindspore/ccsrc/minddata/mindrecord/io/shard_reader.cc

@@ -35,17 +35,16 @@ Type StringToNum(const std::string &str) {
   return num;
 }
 
-ShardReader::ShardReader() {
+ShardReader::ShardReader()
-  task_id_ = 0;
+    : header_size_(0),
-  deliver_id_ = 0;
+      page_size_(0),
-  shard_count_ = 0;
+      shard_count_(0),
-  n_consumer_ = 0;
+      n_consumer_(0),
-  page_size_ = 0;
+      num_padded_(0),
-  header_size_ = 0;
+      num_rows_(0),
-  num_rows_ = 0;
+      total_blob_size_(0),
-  total_blob_size_ = 0;
+      task_id_(0),
-  num_padded_ = 0;
+      deliver_id_(0) {}
-}
 
 std::pair<MSRStatus, std::vector<std::string>> ShardReader::GetMeta(const std::string &file_path, json &meta_data) {
   if (!IsLegalFile(file_path)) {
@@ -880,13 +879,7 @@ MSRStatus ShardReader::Open(const std::vector<std::string> &file_paths, bool loa
   if (n_consumer < kMinConsumerCount) {
     n_consumer = kMinConsumerCount;
   }
-  vector<std::string> blob_fields = GetBlobFields().second;
+
-  for (unsigned int i = 0; i < selected_columns.size(); ++i) {
-    if (!std::any_of(blob_fields.begin(), blob_fields.end(),
-                     [&selected_columns, i](std::string item) { return selected_columns[i] == item; })) {
-      selected_columns_.push_back(selected_columns[i]);
-    }
-  }
   selected_columns_ = selected_columns;
 
   if (CheckColumnList(selected_columns_) == FAILED) {

mindspore/dataset/engine/cache_client.py

@@ -16,7 +16,6 @@
 """
 
 import copy
-# from mindspore._c_dataengine import CacheClient
 
 from ..core.validator_helpers import type_check, check_uint32, check_uint64
 
@@ -39,7 +38,6 @@ class DatasetCache:
         self.port = port
         self.prefetch_size = prefetch_size
         # temporary disable cache feature in the current release
-        # self.cache_client = CacheClient(session_id, size, spilling, hostname, port, prefetch_size)
         self.cache_client = None
 
     def GetStat(self):

mindspore/dataset/engine/graphdata.py

@@ -89,7 +89,6 @@ class GraphData:
                 while self._graph_data.is_stoped() is not True:
                     time.sleep(1)
             except KeyboardInterrupt:
-                # self._graph_data.stop()
                 raise Exception("Graph data server receives KeyboardInterrupt")
 
     @check_gnn_get_all_nodes

mindspore/dataset/engine/validators.py

@@ -1235,5 +1235,4 @@ def check_cache_option(cache):
     """Sanity check for cache parameter"""
     if cache is not None:
         # temporary disable cache feature in the current release
-        # type_check(cache, (cache_client.DatasetCache,), "cache")
         raise ValueError("Caching is disabled in the current release")

tests/ut/python/dataset/test_datasets_csv.py

@@ -12,9 +12,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-import mindspore.dataset as ds
 import numpy as np
 import pytest
+import mindspore.dataset as ds
 
 DATA_FILE = '../data/dataset/testCSV/1.csv'
 

tests/ut/python/dataset/test_epoch_ctrl.py

@@ -16,9 +16,9 @@
 Testing Epoch Control op in DE
 """
 import itertools
-import cv2
 import numpy as np
 import pytest
+import cv2
 
 import mindspore.dataset as ds
 import mindspore.dataset.vision.c_transforms as vision

tests/ut/python/dataset/test_save_op.py

@@ -17,11 +17,11 @@ This is the test module for saveOp.
 """
 import os
 from string import punctuation
+import numpy as np
+import pytest
 import mindspore.dataset as ds
 from mindspore import log as logger
 from mindspore.mindrecord import FileWriter
-import numpy as np
-import pytest
 
 CV_FILE_NAME1 = "../data/mindrecord/testMindDataSet/temp.mindrecord"
 CV_FILE_NAME2 = "../data/mindrecord/testMindDataSet/auto.mindrecord"

tests/ut/python/dataset/test_tensor_string.py

@@ -12,8 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 # ==============================================================================
-import mindspore._c_dataengine as cde
 import numpy as np
+import mindspore._c_dataengine as cde
 
 import mindspore.common.dtype as mstype
 import mindspore.dataset as ds

tests/ut/python/dataset/util.py

@@ -18,9 +18,9 @@ import json
 import os
 import itertools
 from enum import Enum
+import numpy as np
 import matplotlib.pyplot as plt
 import matplotlib.patches as patches
-import numpy as np
 # import jsbeautifier
 import mindspore.dataset as ds
 from mindspore import log as logger