fix the description of cache

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

@@ -388,6 +388,13 @@ uint32_t DatasetOp::GenerateCRC(const std::shared_ptr<DatasetOp> &op) {
   op->tree_->Print(ss, op);
   std::string ss_str = ss.str();
 
+  // Filter out the Num workers field when generating the check sum
+  ss_str = std::regex_replace(ss_str, std::regex("Num workers.*\n"), "");
+  ss_str = std::regex_replace(ss_str, std::regex("\\[workers.*\\]"), "");
+
+  // Filter out Number of rows when generating the check sum
+  ss_str = std::regex_replace(ss_str, std::regex("Number of rows.*\n"), "");
+
   // Filter out the Operator control flags field when generating the check sum
   ss_str = std::regex_replace(ss_str, std::regex("Operator control flags.*\n"), "");
 
@@ -400,6 +407,8 @@ uint32_t DatasetOp::GenerateCRC(const std::shared_ptr<DatasetOp> &op) {
   ss_str = std::regex_replace(ss_str, std::regex("Cache crc.*\n"), "");
   ss_str = std::regex_replace(ss_str, std::regex("Server cache id.*\n"), "");
 
+  MS_LOG(DEBUG) << "Printing the tree for generating crc:\n" << ss_str;
+
   uint32_t cache_crc = system::Crc32c::GetMaskCrc32cValue(ss_str.c_str(), ss_str.length());
   return cache_crc;
 }

mindspore/dataset/engine/datasets.py

@@ -434,7 +434,8 @@ class Dataset:
                 parallel (default=None, the value from the config will be used).
             python_multiprocessing (bool, optional): Parallelize python operations with multiple worker process. This
                 option could be beneficial if the python operation is computational heavy (default=False).
-            cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used)
+            cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used).
+                The cache feature is under development and is not recommended.
 
         Returns:
             MapDataset, dataset after mapping operation.
@@ -1899,7 +1900,9 @@ class MapDataset(DatasetOp):
             in parallel (default=None).
         python_multiprocessing (bool, optional): Parallelize python operations with multiple worker process. This
             option could be beneficial if the python operation is computational heavy (default=False).
-        cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used)
+        cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used).
+            The cache feature is under development and is not recommended.
+
 
         Raises:
             ValueError: If len(input_columns) != len(output_columns) and columns_order is not specified.
@@ -2089,6 +2092,7 @@ class RepeatDataset(DatasetOp):
         """
         return self.count
 
+
 class SkipDataset(DatasetOp):
     """
     The result of applying Skip operator to the input Dataset.
@@ -2354,6 +2358,7 @@ class TransferDataset(DatasetOp):
     def stop_send(self):
         self.iterator.depipeline.StopSend()
 
+
 class RangeDataset(MappableDataset):
     """
     A source dataset that reads and parses datasets stored on disk in a range.
@@ -2500,7 +2505,8 @@ class ImageFolderDatasetV2(MappableDataset):
             into (default=None).
         shard_id (int, optional): The shard ID within num_shards (default=None). This
             argument should be specified only when num_shards is also specified.
-        cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used)
+        cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used).
+            The cache feature is under development and is not recommended.
 
     Raises:
         RuntimeError: If sampler and shuffle are specified at the same time.
@@ -3296,7 +3302,8 @@ class TFRecordDataset(SourceDataset):
             argument should be specified only when num_shards is also specified.
         shard_equal_rows (bool): Get equal rows for all shards(default=False). If shard_equal_rows is false, number
             of rows of each shard may be not equal.
-        cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used)
+        cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used).
+            The cache feature is under development and is not recommended.
     Examples:
         >>> import mindspore.dataset as ds
         >>> import mindspore.common.dtype as mstype
@@ -3867,7 +3874,8 @@ class RandomDataset(SourceDataset):
         num_samples (int): number of samples to draw from the total. (default=None, which means all rows)
         num_parallel_workers (int, optional): number of workers to read the data
             (default=None, number set in the config).
-        cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used)
+        cache (DatasetCache, optional): Tensor cache to use. (default=None which means no cache is used).
+            The cache feature is under development and is not recommended.
         shuffle (bool, optional): Whether or not to perform shuffle on the dataset
             (default=None, expected order behavior shown in the table).
         num_shards (int, optional): Number of shards that the dataset should be divided
@@ -5261,6 +5269,7 @@ class BuildVocabDataset(DatasetOp):
 
         return new_op
 
+
 class BuildSentencePieceVocabDataset(DatasetOp):
     """
     Build a SentencePieceVocab from a dataset.

tests/ut/python/dataset/test_cache_map.py

@@ -24,6 +24,7 @@ DATA_DIR = "../data/dataset/testImageNetData/train/"
 
 GENERATE_GOLDEN = False
 
+
 def test_cache_map_basic1():
     """
     Test mappable leaf with cache op right over the leaf
@@ -104,11 +105,36 @@ def test_cache_map_basic3():
     decode_op = c_vision.Decode()
     ds1 = ds1.repeat(4)
     ds1 = ds1.map(input_columns=["image"], operations=decode_op, cache=some_cache)
-    print("ds1.dataset_size is ", ds1.get_dataset_size())
+    logger.info("ds1.dataset_size is ", ds1.get_dataset_size())
 
     num_iter = 0
     for _ in ds1.create_dict_iterator():
-        print("get data from dataset")
+        logger.info("get data from dataset")
+        num_iter += 1
+
+    logger.info("Number of data in ds1: {} ".format(num_iter))
+    assert num_iter == 8
+    logger.info('test_cache_basic3 Ended.\n')
+
+
+def test_cache_map_basic4():
+    """
+    Test different rows result in core dump
+    """
+    logger.info("Test cache basic 4")
+    some_cache = ds.DatasetCache(session_id=1, size=0, spilling=True)
+
+    # This DATA_DIR only has 2 images in it
+    ds1 = ds.ImageFolderDatasetV2(dataset_dir=DATA_DIR, cache=some_cache)
+    decode_op = c_vision.Decode()
+    ds1 = ds1.repeat(4)
+    ds1 = ds1.map(input_columns=["image"], operations=decode_op)
+    logger.info("ds1.dataset_size is ", ds1.get_dataset_size())
+    shape = ds1.output_shapes()
+    logger.info(shape)
+    num_iter = 0
+    for _ in ds1.create_dict_iterator():
+        logger.info("get data from dataset")
         num_iter += 1
 
     logger.info("Number of data in ds1: {} ".format(num_iter))
@@ -152,12 +178,15 @@ def test_cache_map_failure1():
     assert num_iter == 0
     logger.info('test_cache_failure1 Ended.\n')
 
+
 if __name__ == '__main__':
     test_cache_map_basic1()
-    print("test_cache_map_basic1 success.")
+    logger.info("test_cache_map_basic1 success.")
     test_cache_map_basic2()
-    print("test_cache_map_basic2 success.")
+    logger.info("test_cache_map_basic2 success.")
     test_cache_map_basic3()
-    print("test_cache_map_basic3 success.")
+    logger.info("test_cache_map_basic3 success.")
+    test_cache_map_basic4()
+    logger.info("test_cache_map_basic3 success.")
     test_cache_map_failure1()
-    print("test_cache_map_failure1 success.")
+    logger.info("test_cache_map_failure1 success.")