Fast sample parsing

examples/file_hashing/README.md

@@ -68,7 +68,7 @@ We can poke around in the dataset:
 
 ```python
 # Print summary information about the dataset
-print(dataset.summary())
+dataset
 
 # Print a random sample
 print(dataset.view().take(1).first())
@@ -80,7 +80,7 @@ Create a view that filters only `mountain`
 view = dataset.view().match({"ground_truth.label": "mountain"})
 
 # Print summary information about the view
-print(view.summary())
+view
 
 # Print the first sample in the view
 print(view.first())
@@ -91,7 +91,7 @@ Create a view that sorts labels reverse-alphabetically
 ```python
 view = dataset.view().sort_by("ground_truth.label", reverse=True)
 
-print(view.summary())
+view
 print(view.first())
 ```
 
@@ -139,7 +139,7 @@ for sample in dataset:
     sample["file_hash"] = fou.compute_filehash(sample.filepath)
     sample.save()
 
-print(dataset.summary())
+dataset
 ```
 
 We have two ways to visualize this new information:

fiftyone/core/dataset.py

@@ -928,8 +928,8 @@ def ingest_images(
         return cls.from_images(image_paths, name=name)
 
     def _initialize_dataset(self, name):
-        # Create ODMDatasetSample subclass
-        self._sample_doc_cls = type(self._name, (foo.ODMDatasetSample,), {})
+        # Create ODMSample subclass
+        self._sample_doc_cls = type(self._name, (foo.ODMSample,), {})
 
         # Create dataset meta document
         self._meta = foo.ODMDataset(
@@ -946,7 +946,7 @@ def _load_dataset(self, name):
         # pylint: disable=no-member
         self._meta = foo.ODMDataset.objects.get(name=name)
 
-        self._sample_doc_cls = type(self._name, (foo.ODMDatasetSample,), {})
+        self._sample_doc_cls = type(self._name, (foo.ODMSample,), {})
 
         num_default_fields = len(self.get_field_schema())
 

fiftyone/core/odm/__init__.py

@@ -9,4 +9,4 @@
 from .database import drop_database
 from .dataset import SampleField, ODMDataset
 from .document import ODMDocument, ODMEmbeddedDocument
-from .sample import ODMNoDatasetSample, ODMDatasetSample
+from .sample import NoDatasetSample, ODMSample

fiftyone/core/odm/document.py

@@ -48,13 +48,12 @@ def __copy__(self):
         return self.copy()
 
     def copy(self):
-        """Returns a copy of the document that does not have its `id` set.
+        """Returns a copy of the document.
 
         Returns:
             a :class:`SerializableDocument`
         """
-        doc = deepcopy(self)
-        return doc
+        return deepcopy(self)
 
     def to_dict(self, extended=False):
         """Serializes this document to a JSON dictionary.
@@ -66,6 +65,7 @@ def to_dict(self, extended=False):
         Returns:
             a JSON dict
         """
+        # pylint: disable=no-member
         if extended:
             return json.loads(self.to_json())
 
@@ -98,6 +98,7 @@ def from_dict(cls, d, created=False, extended=False):
         Returns:
             a :class:`ODMDocument`
         """
+        # pylint: disable=no-member
         if not extended:
             try:
                 # Attempt to load the document directly, assuming it is in
@@ -162,11 +163,13 @@ def ingest_time(self):
         """The time the document was added to the database, or ``None`` if it
         has not been added to the database.
         """
+        # pylint: disable=no-member
         return self.id.generation_time if self.in_db else None
 
     @property
     def in_db(self):
         """Whether the underlying :class:`fiftyone.core.odm.ODMDocument` has
         been inserted into the database.
         """
+        # pylint: disable=no-member
         return hasattr(self, "id") and self.id is not None