🐛 fix typing of extraction response when chained (#402)

Author: ianardee

mindee/v2/product/classification/classification_classifier.py

@@ -1,3 +1,5 @@
+from typing import Optional
+
 from mindee.parsing.common.string_dict import StringDict
 from mindee.parsing.v2.inference_response import InferenceResponse
 
@@ -8,7 +10,7 @@ class ClassificationClassifier:
     document_type: str
     """The document type, as identified on given classification values."""
 
-    extraction_response: InferenceResponse
+    extraction_response: Optional[InferenceResponse] = None
     """The extraction response associated with the classification."""
 
     def __init__(self, server_response: StringDict):

mindee/v2/product/crop/__init__.py

@@ -1,11 +1,13 @@
 from mindee.v2.product.crop.crop_box import CropBox
 from mindee.v2.product.crop.crop_inference import CropInference
+from mindee.v2.product.crop.crop_item import CropItem
 from mindee.v2.product.crop.crop_parameters import CropParameters
 from mindee.v2.product.crop.crop_response import CropResponse
 from mindee.v2.product.crop.crop_result import CropResult
 
 __all__ = [
     "CropBox",
+    "CropItem",
     "CropInference",
     "CropParameters",
     "CropResponse",

mindee/v2/product/crop/crop_box.py

@@ -1,18 +1,22 @@
-from mindee.parsing.v2.inference_response import InferenceResponse
+from typing import Optional
+
 from mindee.extraction import ExtractedImage, extract_multiple_images_from_source
-from mindee.input.sources.local_input_source import LocalInputSource
+from mindee.input import LocalInputSource
 from mindee.parsing.common.string_dict import StringDict
-from mindee.parsing.v2.field.field_location import FieldLocation
+from mindee.parsing.v2.field import FieldLocation
+from mindee.parsing.v2.inference_response import InferenceResponse
 
 
 class CropBox:
-    """Crop inference result."""
+    """Deprecated class. Use CropItem instead."""
 
     location: FieldLocation
     """Location which includes cropping coordinates for the detected object, within the source document."""
+
     object_type: str
     """Type or classification of the detected object."""
-    extraction_response: InferenceResponse
+
+    extraction_response: Optional[InferenceResponse] = None
     """The extraction response associated with the crop."""
 
     def __init__(self, server_response: StringDict):

mindee/v2/product/crop/crop_inference.py

@@ -8,6 +8,7 @@ class CropInference(BaseInference):
 
     result: CropResult
     """Result of a crop inference."""
+
     _slug: str = "crop"
     """Slug of the endpoint."""
 

mindee/v2/product/crop/crop_item.py

@@ -0,0 +1,5 @@
+from mindee.v2.product.crop.crop_box import CropBox
+
+
+class CropItem(CropBox):
+    """Result of a cropped document region."""

mindee/v2/product/crop/crop_result.py

@@ -1,16 +1,16 @@
 from typing import List
 
 from mindee.parsing.common.string_dict import StringDict
-from mindee.v2.product.crop.crop_box import CropBox
+from mindee.v2.product.crop.crop_item import CropItem
 
 
 class CropResult:
     """Crop result info."""
 
-    crops: List[CropBox]
+    crops: List[CropItem]
 
     def __init__(self, raw_response: StringDict) -> None:
-        self.crops = [CropBox(crop) for crop in raw_response["crops"]]
+        self.crops = [CropItem(crop) for crop in raw_response["crops"]]
 
     def __str__(self) -> str:
         crops = "\n"

mindee/v2/product/split/split_inference.py

@@ -8,6 +8,7 @@ class SplitInference(BaseInference):
 
     result: SplitResult
     """Result of a split inference."""
+
     _slug: str = "split"
     """Slug of the endpoint."""
 

mindee/v2/product/split/split_range.py

@@ -1,4 +1,4 @@
-from typing import List
+from typing import List, Optional
 
 from mindee.parsing.v2.inference_response import InferenceResponse
 from mindee.extraction.pdf_extractor.extracted_pdf import ExtractedPdf
@@ -15,9 +15,11 @@ class SplitRange:
     0-based page indexes, where the first integer indicates the start page and the
     second integer indicates the end page.
     """
+
     document_type: str
     """The document type, as identified on given classification values."""
-    extraction_response: InferenceResponse
+
+    extraction_response: Optional[InferenceResponse] = None
     """The extraction response associated with the split."""
 
     def __init__(self, server_response: StringDict):

tests/v2/product/ocr/test_ocr_response.py

@@ -13,35 +13,20 @@ def test_ocr_single():
     json_sample, _ = get_product_samples(product="ocr", file_name="ocr_single")
     response = OCRResponse(json_sample)
     assert isinstance(response.inference, OCRInference)
-    assert response.inference.result.pages
-    assert len(response.inference.result.pages) == 1
-    assert response.inference.result.pages[0].words[0].content == "Shipper:"
-    assert (
-        response.inference.result.pages[0].words[0].polygon[0][0] == 0.09742441209406495
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[0][1] == 0.07007125890736342
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[1][0] == 0.15621500559910415
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[1][1] == 0.07046714172604909
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[2][0] == 0.15621500559910415
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[2][1] == 0.08155186064924783
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[3][0] == 0.09742441209406495
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[3][1] == 0.08155186064924783
-    )
-    assert len(response.inference.result.pages[0].words) == 305
-    assert response.inference.result.pages[0].content == (
+    pages = response.inference.result.pages
+    assert pages
+    assert len(pages) == 1
+    assert pages[0].words[0].content == "Shipper:"
+    assert pages[0].words[0].polygon[0][0] == 0.09742441209406495
+    assert pages[0].words[0].polygon[0][1] == 0.07007125890736342
+    assert pages[0].words[0].polygon[1][0] == 0.15621500559910415
+    assert pages[0].words[0].polygon[1][1] == 0.07046714172604909
+    assert pages[0].words[0].polygon[2][0] == 0.15621500559910415
+    assert pages[0].words[0].polygon[2][1] == 0.08155186064924783
+    assert pages[0].words[0].polygon[3][0] == 0.09742441209406495
+    assert pages[0].words[0].polygon[3][1] == 0.08155186064924783
+    assert len(pages[0].words) == 305
+    assert pages[0].content == (
         "Shipper: GLOBAL FREIGHT SOLUTIONS INC. 123 OCEAN DRIVE SHANGHAI, CHINA TEL: "
         "86-21-12345678 FAX: 86-21-87654321\nConsignee: PACIFIC TRADING CO. 789 TRADE "
         "STREET SINGAPORE 567890 SINGAPORE TEL: 65-65432100 FAX: 65-65432101\nNotify "
@@ -81,37 +66,22 @@ def test_ocr_multiple():
     assert isinstance(response.inference, OCRInference)
     assert isinstance(response.inference.result, OCRResult)
     assert isinstance(response.inference.result.pages[0], OCRPage)
-    assert len(response.inference.result.pages) == 3
+    pages = response.inference.result.pages
+    assert len(pages) == 3
 
-    assert len(response.inference.result.pages[0].words) == 295
-    assert response.inference.result.pages[0].words[0].content == "FICTIOCORP"
-    assert (
-        response.inference.result.pages[0].words[0].polygon[0][0] == 0.06649402824332337
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[0][1] == 0.03957449719523875
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[1][0] == 0.23219061218068954
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[1][1] == 0.03960015049938432
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[2][0] == 0.23219061218068954
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[2][1] == 0.06770762074155151
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[3][0] == 0.06649402824332337
-    )
-    assert (
-        response.inference.result.pages[0].words[0].polygon[3][1] == 0.06770762074155151
-    )
+    assert len(pages[0].words) == 295
+    assert pages[0].words[0].content == "FICTIOCORP"
+    assert pages[0].words[0].polygon[0][0] == 0.06649402824332337
+    assert pages[0].words[0].polygon[0][1] == 0.03957449719523875
+    assert pages[0].words[0].polygon[1][0] == 0.23219061218068954
+    assert pages[0].words[0].polygon[1][1] == 0.03960015049938432
+    assert pages[0].words[0].polygon[2][0] == 0.23219061218068954
+    assert pages[0].words[0].polygon[2][1] == 0.06770762074155151
+    assert pages[0].words[0].polygon[3][0] == 0.06649402824332337
+    assert pages[0].words[0].polygon[3][1] == 0.06770762074155151
 
-    assert len(response.inference.result.pages[1].words) == 450
-    assert response.inference.result.pages[1].words[0].content == "KEOLIO"
+    assert len(pages[1].words) == 450
+    assert pages[1].words[0].content == "KEOLIO"
 
-    assert len(response.inference.result.pages[2].words) == 355
-    assert response.inference.result.pages[2].words[0].content == "KEOLIO"
+    assert len(pages[2].words) == 355
+    assert pages[2].words[0].content == "KEOLIO"