1969 import

README.rst

@@ -18,7 +18,7 @@ Welcome to the SuperAnnotate Python Software Development Kit (SDK), which enable
 Resources
 ---------------
 
-- API Reference and User Guide available on `Read the Docs <https://superannotate.readthedocs.io/en/stable/superannotate.sdk.html>`__
+- API Reference and User Guide available on `Read the Docs <https://superannotate.readthedocs.io/en/stable/index.html>`__
 - `Platform documentation <https://doc.superannotate.com/>`__
 
 

requirements_extra.txt

@@ -11,4 +11,5 @@ pytest==7.3.1
 pytest-xdist==3.2.1
 pytest-parallel==0.1.1
 pytest-cov==4.0.0
-pytest-rerunfailures==11.1.2
+pytest-rerunfailures==11.1.2
+jsoncomparison==1.1.0

src/superannotate/lib/app/input_converters/converters/coco_converters/coco_to_sa_pixel.py

@@ -90,7 +90,7 @@ def coco_panoptic_segmentation_to_sa_pixel(coco_path, output_dir):
         cv2.imwrite(str(output_dir / ("%s___save.png" % annot["file_name"])), img)
 
         images_converted.append(annot["file_name"])
-        file_name = "%s___pixel.json" % annot["file_name"]
+        file_name = f"{annot['file_name']}.json"
         sa_metadata = {
             "name": annot_name,
             "width": img_id_to_shape[str(annot["image_id"])]["width"],
@@ -138,7 +138,7 @@ def coco_instance_segmentation_to_sa_pixel(coco_path, output_dir):
     logger.info("Converting to SuperAnnotate JSON format")
     tqdm_thread.start()
     for id_, annotations in images_dict.items():
-        file_name = "%s___pixel.json" % annotations["file_name"]
+        file_name = f"{annotations['file_name']}.json"
         hexcolors = blue_color_generator(len(annotations["annotations"]))
         mask = np.zeros(annotations["shape"])
         H, W, _ = mask.shape

src/superannotate/lib/app/input_converters/converters/coco_converters/coco_to_sa_vector.py

@@ -57,7 +57,7 @@ def save_sa_jsons(coco_json, img_id_to_annot, output_dir):
             sa_instances = []
         else:
             sa_instances = img_id_to_annot[str(img["id"])]
-        file_name = "%s___objects.json" % image_path
+        file_name = f"{image_path}.json"
 
         sa_metadata = {
             "name": image_path,

src/superannotate/lib/app/input_converters/converters/dataloop_converters/dataloop_to_sa_vector.py

@@ -105,11 +105,11 @@ def dataloop_to_sa(input_dir, task, output_dir):
                 sa_tags.append(ann["label"])
 
         if "name" in sa_metadata:
-            file_name = "%s___objects.json" % sa_metadata["name"]
+            file_name = f"{sa_metadata['name']}.json"
         else:
-            file_name = "%s___objects.json" % dl_data["filename"][1:]
+            file_name = f"{dl_data['filename'][1:]}.json"
 
-        images_converted.append(file_name.replace("___objects.json ", ""))
+        images_converted.append(file_name.replace(".json", ""))
         json_template = _create_sa_json(sa_instances, sa_metadata, sa_tags, sa_comments)
         write_to_json(output_dir / file_name, json_template)
     finish_event.set()

src/superannotate/lib/app/input_converters/converters/googlecloud_converters/googlecloud_to_sa_vector.py

@@ -45,7 +45,7 @@ def googlecloud_to_sa_vector(path, output_dir):
             images_not_converted.append(file_name)
             continue
 
-        sa_file_name = "%s___objects.json" % Path(file_name).name
+        sa_file_name = f"{Path(file_name).name}.json"
 
         points = (row[3] * W, row[4] * H, row[5] * W, row[8] * H)
         sa_instances = _create_vector_instance("bbox", points, {}, [], row[2])

src/superannotate/lib/app/input_converters/converters/labelbox_converters/labelbox_to_sa_pixel.py

@@ -35,7 +35,7 @@ def labelbox_instance_segmentation_to_sa_pixel(json_data, output_dir, input_dir)
     logger.info("Converting to SuperAnnotate JSON format")
     tqdm_thread.start()
     for data in json_data:
-        file_name = data["External ID"] + "___pixel.json"
+        file_name = data["External ID"] + ".json"
         mask_name = data["External ID"] + "___save.png"
         sa_metadata = {"name": data["External ID"]}
         if "objects" not in data["Label"].keys():

src/superannotate/lib/app/input_converters/converters/labelbox_converters/labelbox_to_sa_vector.py

@@ -37,7 +37,7 @@ def labelbox_to_sa(json_data, output_dir, task):
     tqdm_thread.start()
     for data in json_data:
         if "objects" not in data["Label"].keys():
-            file_name = data["External ID"] + "___objects.json"
+            file_name = data["External ID"] + ".json"
             write_to_json(
                 output_dir / file_name,
                 {"metadata": {}, "instances": [], "tags": [], "comments": []},
@@ -87,7 +87,7 @@ def labelbox_to_sa(json_data, output_dir, task):
             sa_instances.append(sa_obj)
 
         images_converted.append(data["External ID"])
-        file_name = "%s___objects.json" % data["External ID"]
+        file_name = f"{data['External ID']}.json"
         try:
             img = cv2.imread(str(output_dir / data["External ID"]))
             H, W, _ = img.shape

src/superannotate/lib/app/input_converters/converters/sagemaker_converters/sagemaker_to_sa_pixel.py

@@ -91,7 +91,7 @@ def sagemaker_instance_segmentation_to_sa_pixel(data_path, output_dir):
 
             blue_colors = blue_color_generator(num_of_contours)
             idx = 0
-            file_name = "%s___pixel.json" % (img_mapping[mask_name])
+            file_name = f"{img_mapping[mask_name]}.json"
             sa_metadata = {"name": img_mapping[mask_name], "width": W, "height": H}
             sa_instances = []
             sa_mask = np.zeros((H, W, 4))

src/superannotate/lib/app/input_converters/converters/sagemaker_converters/sagemaker_to_sa_vector.py

@@ -44,7 +44,7 @@ def sagemaker_object_detection_to_sa_vector(data_path, main_key, output_dir):
                 raise Exception
 
             manifest = dataset_manifest[int(img["datasetObjectId"])]
-            file_name = "%s___objects.json" % Path(manifest["source-ref"]).name
+            file_name = f"{Path(manifest['source-ref']).name}.json"
 
             classes = img["consolidatedAnnotation"]["content"][main_key + "-metadata"][
                 "class-map"

src/superannotate/lib/app/input_converters/converters/supervisely_converters/supervisely_to_sa_pixel.py

@@ -34,7 +34,7 @@ def supervisely_instance_segmentation_to_sa_pixel(json_files, class_id_map, outp
     tqdm_thread.start()
 
     for json_file in json_files:
-        file_name = "%s___pixel.json" % Path(json_file).stem
+        file_name = f"{Path(json_file).stem}.json"
 
         json_data = json.load(open(json_file))
         sa_instances = []
@@ -76,10 +76,7 @@ def supervisely_instance_segmentation_to_sa_pixel(json_files, class_id_map, outp
                             parts.append({"color": hex_colors[index]})
                             index += 1
                         cv2.imwrite(
-                            str(
-                                output_dir
-                                / file_name.replace("___pixel.json", "___save.png")
-                            ),
+                            str(output_dir / file_name.replace(".json", "___save.png")),
                             mask,
                         )
                         sa_obj = _create_pixel_instance(

src/superannotate/lib/app/input_converters/converters/supervisely_converters/supervisely_to_sa_vector.py

@@ -36,7 +36,7 @@ def supervisely_to_sa(json_files, class_id_map, task, output_dir):
     tqdm_thread.start()
     for json_file in json_files:
         json_data = json.load(open(json_file))
-        file_name = "%s___objects.json" % Path(json_file).stem
+        file_name = f"{Path(json_file).stem}.json"
         sa_metadata = {
             "name": Path(json_file).stem,
             "width": json_data["size"]["width"],
@@ -140,7 +140,7 @@ def supervisely_keypoint_detection_to_sa_vector(
     logger.info("Converting to SuperAnnotate JSON format")
     tqdm_thread.start()
     for json_file in json_files:
-        file_name = "%s___objects.json" % (Path(json_file).stem)
+        file_name = f"{Path(json_file).stem}.json"
         json_data = json.load(open(json_file))
         sa_metadata = {
             "name": Path(json_file).stem,

src/superannotate/lib/app/input_converters/converters/vgg_converters/vgg_to_sa_vector.py

@@ -53,7 +53,7 @@ def vgg_to_sa(json_data, task, output_dir):
             H = 0
             W = 0
 
-        file_name = "%s___objects.json" % img["filename"]
+        file_name = f"{img['filename']}.json"
         sa_metadata = {"name": img["filename"], "width": W, "height": H}
         sa_instances = []
         instances = img["regions"]

src/superannotate/lib/app/input_converters/converters/voc_converters/voc_to_sa_pixel.py

@@ -126,7 +126,7 @@ def voc_instance_segmentation_to_sa_pixel(voc_root, output_dir):
             sa_instances.append(sa_obj)
 
         images_converted.append(filename)
-        file_name = "%s.jpg___pixel.json" % (filename.stem)
+        file_name = f"{filename.stem}.json"
         height, width = _get_image_shape_from_xml(annotation_dir / filename.name)
         sa_metadata = {"name": filename.stem, "height": height, "width": width}
         sa_json = _create_sa_json(sa_instances, sa_metadata)

src/superannotate/lib/app/input_converters/converters/voc_converters/voc_to_sa_vector.py

@@ -119,7 +119,7 @@ def voc_instance_segmentation_to_sa_vector(voc_root, output_dir):
 
         images_converted.append(filename)
         file_name, height, width = _get_image_metadata(annotation_dir / filename.name)
-        file_path = f"{file_name}___objects.json"
+        file_path = f"{file_name}.json"
         sa_metadata = {"name": str(filename), "height": height, "width": width}
         sa_json = _create_sa_json(sa_instances, sa_metadata)
         write_to_json(output_dir / file_path, sa_json)
@@ -162,7 +162,7 @@ def voc_object_detection_to_sa_vector(voc_root, output_dir):
 
         images_converted.append(filename)
         file_name, height, width = _get_image_metadata(annotation_dir / filename.name)
-        file_path = f"{file_name}___objects.json"
+        file_path = f"{file_name}.json"
         sa_metadata = {"name": str(filename), "height": height, "width": width}
         sa_json = _create_sa_json(sa_instances, sa_metadata)
         write_to_json(output_dir / file_path, sa_json)

src/superannotate/lib/app/input_converters/converters/vott_converters/vott_to_sa_vector.py

@@ -34,7 +34,7 @@ def vott_to_sa(file_list, task, output_dir):
     tqdm_thread.start()
     for json_file in file_list:
         json_data = json.load(open(json_file))
-        file_name = "%s___objects.json" % json_data["asset"]["name"]
+        file_name = f"{json_data['asset']['name']}.json"
         sa_metadata = {
             "name": json_data["asset"]["name"],
             "width": json_data["asset"]["size"]["width"],

src/superannotate/lib/app/input_converters/converters/yolo_converters/yolo_to_sa_vector.py

@@ -75,7 +75,7 @@ def yolo_object_detection_to_sa_vector(data_path, output_dir):
             sa_instances.append(sa_obj.copy())
 
         images_converted.append(annotation.name)
-        file_name = "%s___objects.json" % Path(file_name).name
+        file_name = f"{Path(file_name).name}.json"
         sa_metadata = {"name": Path(file_name).name, "width": W, "height": H}
         sa_json = _create_sa_json(sa_instances, sa_metadata)
         write_to_json(output_dir / file_name, sa_json)

src/superannotate/lib/core/usecases/images.py

@@ -231,14 +231,6 @@ def default_annotation(self):
             "annotation_mask_filename": None,
         }
 
-    @property
-    def annotation_json_name(self):
-        if self._from_project.type == constances.ProjectType.VECTOR.value:
-            return f"{self._from_image.name}___objects.json"
-        elif self._from_project.type == constances.ProjectType.PIXEL.value:
-            return f"{self._from_image.name}___pixel.json"
-        return f"{self._from_image.name}.json"
-
     @property
     def download_auth_data(self):
         return self._service_provider.get_download_token(
@@ -426,10 +418,7 @@ def annotations(self):
             image_path = (
                 f"{Path(self._image_path).parent}/{Path(self._image_path).name}"
             )
-            if self._project_type.upper() == constances.ProjectType.PIXEL.name.upper():
-                self._annotations = json.load(open(f"{image_path}___pixel.json"))
-            else:
-                self._annotations = json.load(open(f"{image_path}___objects.json"))
+            self._annotations = json.load(open(f"{image_path}.json"))
         return self._annotations
 
     @property
@@ -1487,10 +1476,6 @@ def execute(self):
             credentials = token["annotations"]["MAIN"][0]
 
             annotation_json_creds = credentials["annotation_json_path"]
-            if self._project.type == constances.ProjectType.VECTOR.value:
-                file_postfix = "___objects.json"
-            else:
-                file_postfix = "___pixel.json"
 
             response = requests.get(
                 url=annotation_json_creds["url"],
@@ -1502,7 +1487,7 @@ def execute(self):
                 self._response.data = (None, None)
                 return self._response
             data["annotation_json"] = response.json()
-            data["annotation_json_filename"] = f"{self._image_name}{file_postfix}"
+            data["annotation_json_filename"] = f"{self._image_name}.json"
             mask_path = None
             if self._project.type == constances.ProjectType.PIXEL.value:
                 annotation_blue_map_creds = credentials["annotation_bluemap_path"]
@@ -1530,82 +1515,6 @@ def execute(self):
         return self._response
 
 
-class GetImageAnnotationsUseCase(BaseReportableUseCase):
-    def __init__(
-        self,
-        reporter: Reporter,
-        service_provider: BaseServiceProvider,
-        project: ProjectEntity,
-        folder: FolderEntity,
-        image_name: str,
-    ):
-        super().__init__(reporter)
-        self._service_provider = service_provider
-        self._project = project
-        self._folder = folder
-        self._image_name = image_name
-
-    @property
-    def image_use_case(self):
-        use_case = GetImageUseCase(
-            project=self._project,
-            folder=self._folder,
-            image_name=self._image_name,
-            service_provider=self._service_provider,
-        )
-        return use_case
-
-    def validate_project_type(self):
-        if self._project.type in constances.LIMITED_FUNCTIONS:
-            raise AppValidationException(
-                constances.LIMITED_FUNCTIONS[self._project.type]
-            )
-
-    def execute(self):
-        if self.is_valid():
-            data = {
-                "annotation_json": None,
-                "annotation_json_filename": None,
-                "annotation_mask": None,
-                "annotation_mask_filename": None,
-            }
-            image_response = self.image_use_case.execute()
-            token = self._service_provider.get_download_token(
-                project=self._project,
-                folder=self._folder,
-                image_id=image_response.data.id,
-            ).data
-            credentials = token["annotations"]["MAIN"][0]
-            if self._project.type == constances.ProjectType.VECTOR.value:
-                file_postfix = "___objects.json"
-            else:
-                file_postfix = "___pixel.json"
-                data["annotation_mask_filename"] = f"{self._image_name}___save.png"
-            data["annotation_json_filename"] = f"{self._image_name}{file_postfix}"
-
-            response = requests.get(
-                url=credentials["annotation_json_path"]["url"],
-                headers=credentials["annotation_json_path"]["headers"],
-            )
-            if not response.ok:
-                self.reporter.log_warning("Couldn't load annotations.")
-                self._response.data = data
-                return self._response
-            data["annotation_json"] = response.json()
-            data["annotation_json_filename"] = f"{self._image_name}{file_postfix}"
-            if self._project.type == constances.ProjectType.PIXEL.value:
-                annotation_blue_map_creds = credentials["annotation_bluemap_path"]
-                response = requests.get(
-                    url=annotation_blue_map_creds["url"],
-                    headers=annotation_blue_map_creds["headers"],
-                )
-                data["annotation_mask"] = io.BytesIO(response.content)
-
-            self._response.data = data
-
-        return self._response
-
-
 class UnAssignFolderUseCase(BaseUseCase):
     def __init__(
         self,

tests/integration/convertors/data_set/coco_object_detection/TestVectorAnnotationImage.json

@@ -93,7 +93,7 @@
                 10
             ],
             "area": 100,
-            "category_id": -1
+            "category_id": 1611842
         },
         {
             "id": 3,

tests/integration/convertors/data_set/coco_to_sa_instance_segmentation_expected_result/classes/classes.json

@@ -0,0 +1,22 @@
+[
+  {
+    "name": "Personal vehicle",
+    "color": "#1b066e",
+    "attribute_groups": []
+  },
+  {
+    "name": "Large vehicle",
+    "color": "#2e5ce0",
+    "attribute_groups": []
+  },
+  {
+    "name": "Human",
+    "color": "#4949cc",
+    "attribute_groups": []
+  },
+  {
+    "name": "Plant",
+    "color": "#460f3e",
+    "attribute_groups": []
+  }
+]