🔧 fix non-page-level id for fields

docs/extras/guide/custom_v1.md

@@ -65,7 +65,7 @@ Values of `ListField`s are stored in a `ListFieldValue` structure, which is impl
 * **confidence** (`float`): the confidence score of the prediction
 * **bounding_box** (`BBox`): 4 relative vertices corrdinates of a rectangle containing the word in the document.
 * **polygon** (`Polygon`): vertices of a polygon containing the word.
-* **page_id** (`int`): the ID of the page, is `undefined` when at document-level.
+* **page_id** (`Optional[int]`): the ID of the page, is `null` when at document-level.
 
 
 ### Classification Field

mindee/parsing/common/page.py

@@ -30,10 +30,7 @@ def __init__(
             self.orientation = OrientationField(
                 raw_prediction["orientation"], page_id=self.id
             )
-        try:
-            self.prediction = prediction_type(raw_prediction["prediction"], self.id)
-        except TypeError:
-            self.prediction = prediction_type(raw_prediction["prediction"])
+        self.prediction = prediction_type(raw_prediction["prediction"])
 
         if "extras" in raw_prediction and raw_prediction["extras"]:
             self.extras = Extras(raw_prediction["extras"])

mindee/parsing/custom/list.py

@@ -14,12 +14,12 @@ class ListFieldValue(FieldPositionMixin):
     page_id: Optional[int]
     """Id of the page the field was found on."""
 
-    def __init__(
-        self, raw_prediction: StringDict, page_id: Optional[int] = None
-    ) -> None:
+    def __init__(self, raw_prediction: StringDict) -> None:
         self.content = raw_prediction["content"]
         self.confidence = raw_prediction["confidence"]
-        self.page_id = page_id
+        self.page_id = (
+            raw_prediction["page_id"] if "page_id" in raw_prediction else None
+        )
         self._set_position(raw_prediction)
 
     def __str__(self) -> str:
@@ -36,19 +36,12 @@ class ListField:
     values: List[ListFieldValue]
     """List of word values"""
 
-    def __init__(
-        self,
-        raw_prediction: StringDict,
-        reconstructed: bool = False,
-        page_id: Optional[int] = None,
-    ) -> None:
+    def __init__(self, raw_prediction: StringDict, reconstructed: bool = False) -> None:
         self.values = []
         self.reconstructed = reconstructed
 
         for value in raw_prediction["values"]:
-            if "page_id" in value:
-                page_id = value["page_id"]
-            self.values.append(ListFieldValue(value, page_id))
+            self.values.append(ListFieldValue(value))
         self.confidence = raw_prediction["confidence"]
 
     @property

mindee/product/custom/custom_v1_page.py

@@ -1,4 +1,4 @@
-from typing import Dict, List, Optional
+from typing import Dict, List
 
 from mindee.parsing.common import Prediction, StringDict, clean_out_string
 from mindee.parsing.custom import ListField
@@ -11,15 +11,15 @@ class CustomV1Page(Prediction):
     fields: Dict[str, ListField]
     """Dictionary of all fields in the document"""
 
-    def __init__(self, raw_prediction: StringDict, page_id: Optional[int]) -> None:
+    def __init__(self, raw_prediction: StringDict) -> None:
         """
         Custom document object.
 
         :param raw_prediction: Dictionary containing the JSON document response
         """
         self.fields = {}
         for field_name, field_contents in raw_prediction.items():
-            self.fields[field_name] = ListField(field_contents, page_id=page_id)
+            self.fields[field_name] = ListField(field_contents)
 
     def columns_to_line_items(
         self,