Merge pull request #272 from Labelbox/ms/fix-classification-bug

fix classification bug

Author: msokoloff1

labelbox/data/annotation_types/annotation.py

@@ -1,7 +1,5 @@
 from typing import Any, Dict, List, Union
 
-from pydantic.main import BaseModel
-
 from .classification import Checklist, Dropdown, Radio, Text
 from .feature import FeatureSchema
 from .geometry import Geometry

labelbox/data/annotation_types/classification/classification.py

@@ -1,10 +1,24 @@
 from typing import Any, Dict, List
 
-from pydantic.main import BaseModel
+try:
+    from typing import Literal
+except:
+    from typing_extensions import Literal
 
+from pydantic import BaseModel, validator
 from ..feature import FeatureSchema
 
 
+# TODO: Replace when pydantic adds support for unions that don't coerce types
+class _TempName(BaseModel):
+    name: str
+
+    def dict(self, *args, **kwargs):
+        res = super().dict(*args, **kwargs)
+        res.pop('name')
+        return res
+
+
 class ClassificationAnswer(FeatureSchema):
     """
     - Represents a classification option.
@@ -19,8 +33,9 @@ class Radio(BaseModel):
     answer: ClassificationAnswer
 
 
-class Checklist(BaseModel):
+class Checklist(_TempName):
     """ A classification with many selected options allowed """
+    name: Literal["checklist"] = "checklist"
     answer: List[ClassificationAnswer]
 
 
@@ -29,9 +44,10 @@ class Text(BaseModel):
     answer: str
 
 
-class Dropdown(BaseModel):
+class Dropdown(_TempName):
     """
     - A classification with many selected options allowed .
     - This is not currently compatible with MAL.
     """
+    name: Literal["dropdown"] = "dropdown"
     answer: List[ClassificationAnswer]

labelbox/data/serialization/labelbox_v1/classification.py

@@ -1,6 +1,7 @@
 from typing import List, Union
 
 from pydantic.main import BaseModel
+from pydantic.schema import schema
 
 from ...annotation_types.annotation import ClassificationAnnotation
 from ...annotation_types.classification import Checklist, ClassificationAnswer, Radio, Text, Dropdown
@@ -15,7 +16,7 @@ class LBV1ClassificationAnswer(LBV1Feature):
 class LBV1Radio(LBV1Feature):
     answer: LBV1ClassificationAnswer
 
-    def to_common(self):
+    def to_common(self) -> Radio:
         return Radio(answer=ClassificationAnswer(
             feature_schema_id=self.answer.schema_id,
             name=self.answer.title,
@@ -39,7 +40,7 @@ def from_common(cls, radio: Radio, feature_schema_id: Cuid,
 class LBV1Checklist(LBV1Feature):
     answers: List[LBV1ClassificationAnswer]
 
-    def to_common(self):
+    def to_common(self) -> Checklist:
         return Checklist(answer=[
             ClassificationAnswer(feature_schema_id=answer.schema_id,
                                  name=answer.title,
@@ -64,6 +65,34 @@ def from_common(cls, checklist: Checklist, feature_schema_id: Cuid,
                    **extra)
 
 
+class LBV1Dropdown(LBV1Feature):
+    answer: List[LBV1ClassificationAnswer]
+
+    def to_common(self) -> Dropdown:
+        return Dropdown(answer=[
+            ClassificationAnswer(feature_schema_id=answer.schema_id,
+                                 name=answer.title,
+                                 extra={
+                                     'feature_id': answer.feature_id,
+                                     'value': answer.value
+                                 }) for answer in self.answer
+        ])
+
+    @classmethod
+    def from_common(cls, dropdown: Dropdown, feature_schema_id: Cuid,
+                    **extra) -> "LBV1Dropdown":
+        return cls(schema_id=feature_schema_id,
+                   answer=[
+                       LBV1ClassificationAnswer(
+                           schema_id=answer.feature_schema_id,
+                           title=answer.name,
+                           value=answer.extra.get('value'),
+                           feature_id=answer.extra.get('feature_id'))
+                       for answer in dropdown.answer
+                   ],
+                   **extra)
+
+
 class LBV1Text(LBV1Feature):
     answer: str
 
@@ -77,7 +106,8 @@ def from_common(cls, text: Text, feature_schema_id: Cuid,
 
 
 class LBV1Classifications(BaseModel):
-    classifications: List[Union[LBV1Radio, LBV1Checklist, LBV1Text]] = []
+    classifications: List[Union[LBV1Text, LBV1Radio, LBV1Dropdown,
+                                LBV1Checklist]] = []
 
     def to_common(self) -> List[ClassificationAnnotation]:
         classifications = [
@@ -112,10 +142,10 @@ def from_common(
     @staticmethod
     def lookup_classification(
         annotation: ClassificationAnnotation
-    ) -> Union[LBV1Text, LBV1Checklist, LBV1Radio]:
+    ) -> Union[LBV1Text, LBV1Checklist, LBV1Radio, LBV1Checklist]:
         return {
             Text: LBV1Text,
-            Dropdown: LBV1Checklist,
+            Dropdown: LBV1Dropdown,
             Checklist: LBV1Checklist,
             Radio: LBV1Radio
         }.get(type(annotation.value))

labelbox/data/serialization/labelbox_v1/converter.py

@@ -1,3 +1,4 @@
+from labelbox.data.serialization.labelbox_v1.objects import LBV1Mask
 from typing import Any, Dict, Generator, Iterable
 import logging
 

labelbox/data/serialization/labelbox_v1/objects.py

@@ -1,21 +1,22 @@
 from typing import Any, Dict, List, Optional, Union
 
-from pydantic import BaseModel
+from pydantic import BaseModel, validator
 
 from ...annotation_types.annotation import (ClassificationAnnotation,
                                             ObjectAnnotation)
 from ...annotation_types.data import MaskData
 from ...annotation_types.geometry import Line, Mask, Point, Polygon, Rectangle
 from ...annotation_types.ner import TextEntity
 from ...annotation_types.types import Cuid
-from .classification import LBV1Checklist, LBV1Classifications, LBV1Radio, LBV1Text
+from .classification import LBV1Checklist, LBV1Classifications, LBV1Radio, LBV1Text, LBV1Dropdown
 from .feature import LBV1Feature
 
 
 class LBV1ObjectBase(LBV1Feature):
     color: Optional[str] = None
     instanceURI: Optional[str] = None
-    classifications: List[Union[LBV1Radio, LBV1Checklist, LBV1Text]] = []
+    classifications: List[Union[LBV1Text, LBV1Radio, LBV1Dropdown,
+                                LBV1Checklist]] = []
 
     def dict(self, *args, **kwargs):
         res = super().dict(*args, **kwargs)
@@ -24,6 +25,14 @@ def dict(self, *args, **kwargs):
             res.pop('instanceURI')
         return res
 
+    @validator('classifications', pre=True)
+    def validate_subclasses(cls, value, field):
+        # Dropdown subclasses create extra unessesary nesting. So we just remove it.
+        if isinstance(value, list) and len(value):
+            if isinstance(value[0], list):
+                return value[0]
+        return value
+
 
 class _Point(BaseModel):
     x: float

labelbox/data/serialization/ndjson/classification.py

@@ -165,8 +165,8 @@ def to_common(
     def lookup_subclassification(
         annotation: ClassificationAnnotation
     ) -> Union[NDTextSubclass, NDChecklistSubclass, NDRadioSubclass]:
-        if isinstance(annotation, Dropdown):
-            raise TypeError("Dropdowns are not supported for MAL")
+        if isinstance(annotation.value, Dropdown):
+            raise TypeError("Dropdowns are not supported for MAL.")
         return {
             Text: NDTextSubclass,
             Checklist: NDChecklistSubclass,
@@ -213,13 +213,12 @@ def lookup_classification(
         annotation: Union[ClassificationAnnotation,
                           VideoClassificationAnnotation]
     ) -> Union[NDText, NDChecklist, NDRadio]:
-        if isinstance(annotation, Dropdown):
-            raise TypeError("Dropdowns are not supported for MAL")
+        if isinstance(annotation.value, Dropdown):
+            raise TypeError("Dropdowns are not supported for MAL.")
         return {
             Text: NDText,
             Checklist: NDChecklist,
-            Radio: NDRadio,
-            Dropdown: NDChecklist,
+            Radio: NDRadio
         }.get(type(annotation.value))
 
 

tests/data/assets/labelbox_v1/highly_nested_image.json

@@ -0,0 +1 @@
+{"ID": "ckt3hu4aq524s0y9e7i0dbbtg", "DataRow ID": "ckt3h8c7e200j0y6ob2g3629b", "Labeled Data": "https://storage.labelbox.com/ckk4q1vgapsau07324awnsjq2%2Fa8c0d364-b10f-5b49-09a3-151264999cfb-1?Expires=1630706709794&KeyName=labelbox-assets-key-3&Signature=W8RPYzKo1Qs6Qxclapnx4_chhJ8", "Label": {"objects": [{"featureId": "ckt3hu9ux00003h69bzk55eaj", "schemaId": "ckt3h8e8s51li0y7ucy0bgrlz", "color": "#ff0000", "title": "deer", "value": "deer", "polygon": [{"x": 71.832, "y": 62.37}, {"x": 50.636, "y": 90.087}, {"x": 103.625, "y": 94.571}, {"x": 104.033, "y": 59.517}], "instanceURI": "https://api.labelbox.com/masks/feature/ckt3hu9ux00003h69bzk55eaj?token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VySWQiOiJja2s0cTF2Z3djMHZwMDcwNHhoeDdtNHZrIiwib3JnYW5pemF0aW9uSWQiOiJja2s0cTF2Z2Fwc2F1MDczMjRhd25zanEyIiwiaWF0IjoxNjMwNjI1NjgxLCJleHAiOjE2MzMyMTc2ODF9.hUJ46wdigdN9RdICF8FYzaA7MLbscfjvBeP5O-QhGa8", "classifications": [{"featureId": "ckt3hud1m00023h696pofvwkq", "schemaId": "ckt3hu2yn523g0y9ehzhg5opb", "title": "rrrr", "value": "rrrr", "answer": {"featureId": "ckt3hud1l00013h69xob7gwoh", "schemaId": "ckt3hu302523w0y9e39pub4zp", "title": "rrrrrrrr", "value": "rrrrrrrr"}}]}, {"featureId": "ckt3hullq000c3h69o8iqty8k", "schemaId": "ckt3h8e8s51lm0y7ucl6bgixn", "color": "#0000ff", "title": "deer_eyes", "value": "deer_eyes", "instanceURI": "https://api.labelbox.com/masks/feature/ckt3hullq000c3h69o8iqty8k?token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VySWQiOiJja2s0cTF2Z3djMHZwMDcwNHhoeDdtNHZrIiwib3JnYW5pemF0aW9uSWQiOiJja2s0cTF2Z2Fwc2F1MDczMjRhd25zanEyIiwiaWF0IjoxNjMwNjI1NjgxLCJleHAiOjE2MzMyMTc2ODF9.hUJ46wdigdN9RdICF8FYzaA7MLbscfjvBeP5O-QhGa8"}, {"featureId": "ckt3hug4600073h692cpxwc2v", "schemaId": "ckt3hu2ya523e0y9eaika5flx", "color": "#00ff00", "title": "deer_nose", "value": "deer_nose", "point": {"x": 73.054, "y": 165.495}, "instanceURI": "https://api.labelbox.com/masks/feature/ckt3hug4600073h692cpxwc2v?token=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VySWQiOiJja2s0cTF2Z3djMHZwMDcwNHhoeDdtNHZrIiwib3JnYW5pemF0aW9uSWQiOiJja2s0cTF2Z2Fwc2F1MDczMjRhd25zanEyIiwiaWF0IjoxNjMwNjI1NjgxLCJleHAiOjE2MzMyMTc2ODF9.hUJ46wdigdN9RdICF8FYzaA7MLbscfjvBeP5O-QhGa8", "classifications": [[{"featureId": "ckt3hui8800093h69swml8uzf", "schemaId": "ckt3hu308523x0y9eclq1g2k8", "title": "description", "value": "description", "answers": [{"featureId": "ckt3hui8800083h69696qdgdf", "schemaId": "ckt3hu31m524b0y9eeiuobmlo", "title": "wet", "value": "wet"}, {"featureId": "ckt3kb8ym00013h69lydpjjkp", "schemaId": "ckt3k5asi26wt0y6o805pf9ve", "title": "dry", "value": "dry"}]}, {"featureId": "ckt3hujza000b3h69jfu850bx", "schemaId": "ckt3hu32x524p0y9ehuzf6wtw", "title": "asdasdsadsa", "value": "asdasdsadsa", "answer": {"featureId": "ckt3hujza000a3h691fbs2iik", "schemaId": "ckt3hu33h524r0y9e5q6j91wt", "title": "aasdsadsada", "value": "aasdsadsada"}}, {"featureId": "ckt3kbapf00033h69aup591oj", "schemaId": "ckt3k5au226x50y6o5wlv9j3g", "title": "anotherone...", "value": "anotherone...", "answer": {"featureId": "ckt3kbape00023h69ohwyr625", "schemaId": "ckt3k5aum26xb0y6o2gxtctze", "title": "weeewrew", "value": "weeewrew"}}]]}], "classifications": [{"featureId": "ckt3huphm000f3h69esv5u6ch", "schemaId": "ckt3h8e8r51lg0y7u2mtz9791", "title": "image_description", "value": "image_description", "answers": [{"featureId": "ckt3huphm000e3h69a18xeme7", "schemaId": "ckt3h8e9m51lo0y7u5v81b3cz", "title": "bright", "value": "bright"}, {"featureId": "ckt3hus0i000h3h69kwde3ilk", "schemaId": "ckt3h8e9m51lq0y7u15ygdlyw", "title": "not_blurry", "value": "not_blurry"}, {"featureId": "ckt3huuhi000k3h69io60j4xq", "schemaId": "ckt3h8e9m51ls0y7u82aj5m3f", "title": "dark", "value": "dark"}]}, {"featureId": "ckt3hvcm4000p3h69vdbc0c23", "schemaId": "ckt3hu2y652300y9e6fkefo42", "title": "a", "value": "a", "answer": {"featureId": "ckt3hvcm4000o3h693oy8exqm", "schemaId": "ckt3hu2z6523l0y9e83yg91zc", "title": "ass", "value": "ass"}}, {"featureId": "ckt3hvdl4000r3h69jqvrrb67", "schemaId": "ckt3hu30w52410y9e485y1vwi", "title": "asdsadsad", "value": "asdsadsad", "answer": [{"featureId": "ckt3hvdl4000q3h69i8afxwv5", "schemaId": "ckt3hu31o524d0y9ehwk4ds2a", "title": "asdasdsa", "value": "asdasdsa"}]}, {"featureId": "ckt3hvfz4000s3h69heqauwmu", "schemaId": "ckt3hu30x52430y9ebixq8f3t", "title": "wee223", "value": "wee223", "answer": "fretre"}, {"featureId": "ckt3hvgkq000u3h69tjsg5rl1", "schemaId": "ckt3hu30x52450y9eaw056vgj", "title": "3223432", "value": "3223432", "answer": {"featureId": "ckt3hvgkq000t3h6928o3rgzf", "schemaId": "ckt3hu31v524j0y9e7x1g2shq", "title": "dddd", "value": "dddd"}}, {"featureId": "ckt3hvife000v3h6980b3ccxa", "schemaId": "ckt3hu2y652320y9e4n1lb9dx", "title": "weeeee", "value": "weeeee", "answer": "5354"}, {"featureId": "ckt3hviho000x3h69xt3npgtb", "schemaId": "ckt3hu2y652340y9e1brp3l1i", "title": "asdsadsa", "value": "asdsadsa", "answer": [{"featureId": "ckt3hviho000w3h690kvv8isp", "schemaId": "ckt3hu2zn523p0y9eel4l4ts9", "title": "asxdsds", "value": "asxdsds"}]}], "relationships": []}, "Created By": "[email protected]", "Project Name": "test_annotation_types", "Created At": "2021-09-02T22:23:24.000Z", "Updated At": "2021-09-02T23:31:27.707Z", "Seconds to Label": 220.91400000000002, "External ID": null, "Agreement": -1, "Benchmark Agreement": -1, "Benchmark ID": null, "Dataset Name": "label_dataset", "Reviews": [], "View Label": "https://editor.labelbox.com?project=ckt3h8d9z51l10y7uavvf5zcn&label=ckt3hu4aq524s0y9e7i0dbbtg", "Has Open Issues": 0, "Skipped": false}

tests/data/serialization/labelbox_v1/test_image.py

@@ -1,10 +1,16 @@
 import json
 
+import pytest
+
 from labelbox.data.serialization.labelbox_v1.converter import LBV1Converter
 
 
-def test_image():
-    with open('tests/data/assets/labelbox_v1/image_export.json', 'r') as file:
+@pytest.mark.parametrize("file_path", [
+    'tests/data/assets/labelbox_v1/highly_nested_image.json',
+    'tests/data/assets/labelbox_v1/image_export.json'
+])
+def test_image(file_path):
+    with open(file_path, 'r') as file:
         payload = json.load(file)
 
     collection = LBV1Converter.deserialize([payload])
@@ -20,4 +26,14 @@ def test_image():
                 if not len(annotation_a['classifications']):
                     # We don't add a classification key to the payload if there is no classifications.
                     annotation_a.pop('classifications')
+
+                if isinstance(annotation_b.get('classifications'),
+                              list) and len(annotation_b['classifications']):
+                    if isinstance(annotation_b['classifications'][0], list):
+                        annotation_b['classifications'] = annotation_b[
+                            'classifications'][0]
+
                 assert annotation_a == annotation_b
+
+
+# After check the nd serializer on this shit.. It should work for almost everything (except the other horse shit..)