rapidata 1.0.0__py3-none-any.whl → 1.1.0__py3-none-any.whl

rapidata/rapidata_client/dataset/validation_set_builder.py

@@ -160,8 +160,8 @@ class ValidationSetBuilder:
         self,
         asset: MediaAsset | TextAsset,
         question: str,
-        transcription: list[str],
-        correct_words: list[str],
+        transcription: str,
+        truths: list[int],
         strict_grading: bool | None = None,
         metadata: list[Metadata] = [],
     ):
@@ -171,9 +171,9 @@ class ValidationSetBuilder:
             asset (MediaAsset | TextAsset): The asset for the rapid.
             question (str): The question for the rapid.
             transcription (list[str]): The transcription for the rapid.
-            correct_words (list[str]): The list of correct words for the rapid.
-            strict_grading (bool | None, optional): The strict grading flag for the rapid. Defaults to None.
-            metadata (list[Metadata], optional): The metadata for the rapid. Defaults to an empty list.
+            truths (list[int]): The list of indices of the true word selections.
+            strict_grading (bool | None, optional): The strict grading for the rapid. Defaults to None.
+            metadata (list[Metadata], optional): The metadata for the rapid.
 
         Returns:
             ValidationSetBuilder: The ValidationSetBuilder instance.
@@ -183,16 +183,14 @@ class ValidationSetBuilder:
         """
         transcription_words = [
             TranscriptionWord(word=word, wordIndex=i)
-            for i, word in enumerate(transcription)
+            for i, word in enumerate(transcription.split())
         ]
 
-        correct_transcription_words = []
-        for word in correct_words:
-            if word not in transcription:
-                raise ValueError(f"Correct word '{word}' not found in transcription")
-            correct_transcription_words.append(
-                TranscriptionWord(word=word, wordIndex=transcription.index(word))
-            )
+        true_words = []
+        for idx in truths:
+            if idx > len(transcription_words) - 1:
+                raise ValueError(f"Index {idx} is out of bounds")
+            true_words.append(transcription_words[idx])
 
         payload = TranscriptionPayload(
             _t="TranscriptionPayload", title=question, transcription=transcription_words
@@ -200,7 +198,7 @@ class ValidationSetBuilder:
 
         model_truth = TranscriptionTruth(
             _t="TranscriptionTruth",
-            correctWords=correct_transcription_words,
+            correctWords=true_words,
             strictGrading=strict_grading,
         )
 

rapidata/rapidata_client/order/rapidata_order_builder.py

@@ -13,6 +13,9 @@ from rapidata.api_client.models.create_order_model_workflow import (
     CreateOrderModelWorkflow,
 )
 from rapidata.api_client.models.country_user_filter_model import CountryUserFilterModel
+from rapidata.api_client.models.language_user_filter_model import (
+    LanguageUserFilterModel,
+)
 from rapidata.rapidata_client.feature_flags import FeatureFlags
 from rapidata.rapidata_client.metadata.base_metadata import Metadata
 from rapidata.rapidata_client.dataset.rapidata_dataset import RapidataDataset
@@ -58,6 +61,7 @@ class RapidataOrderBuilder:
         self._validation_set_id: str | None = None
         self._feature_flags: FeatureFlags | None = None
         self._country_codes: list[str] | None = None
+        self._language_codes: list[str] | None = None
         self._selections: list[Selection] = []
         self._rapids_per_bag: int = 2
         self._priority: int = 50
@@ -80,22 +84,32 @@ class RapidataOrderBuilder:
         if self._referee is None:
             print("No referee provided, using default NaiveReferee.")
             self._referee = NaiveReferee()
-        if self._country_codes is None:
-            country_filter = None
-        else:
-            country_filter = CountryUserFilterModel(
-                _t="CountryFilter", countries=self._country_codes
+
+        user_filters = []
+
+        if self._country_codes is not None:
+            user_filters.append(
+                CreateOrderModelUserFiltersInner(
+                    CountryUserFilterModel(
+                        _t="CountryFilter", countries=self._country_codes
+                    )
+                )
+            )
+
+        if self._language_codes is not None:
+            user_filters.append(
+                CreateOrderModelUserFiltersInner(
+                    LanguageUserFilterModel(
+                        _t="LanguageFilter", languages=self._language_codes
+                    )
+                )
             )
 
         return CreateOrderModel(
             _t="CreateOrderModel",
             orderName=self._name,
             workflow=CreateOrderModelWorkflow(self._workflow.to_model()),
-            userFilters=(
-                [CreateOrderModelUserFiltersInner(country_filter)]
-                if country_filter
-                else []
-            ),
+            userFilters=user_filters,
             referee=CreateOrderModelReferee(self._referee.to_model()),
             validationSetId=self._validation_set_id,
             featureFlags=(
@@ -241,7 +255,7 @@ class RapidataOrderBuilder:
 
     def country_filter(self, country_codes: list[str]) -> "RapidataOrderBuilder":
         """
-        Set the target country codes for the order.
+        Set the target country codes for the order. E.g. `country_codes=["DE", "CH", "AT"]` for Germany, Switzerland, and Austria.
 
         Args:
             country_codes (list[str]): The country codes to be set.
@@ -252,6 +266,19 @@ class RapidataOrderBuilder:
         self._country_codes = country_codes
         return self
 
+    def language_filter(self, language_codes: list[str]) -> "RapidataOrderBuilder":
+        """
+        Set the target language codes for the order. E.g. `language_codes=["de", "fr", "it"]` for German, French, and Italian.
+
+        Args:
+            language_codes (list[str]): The language codes to be set.
+
+        Returns:
+            RapidataOrderBuilder: The updated RapidataOrderBuilder instance.
+        """
+        self._language_codes = language_codes
+        return self
+
     def aggregator(self, aggregator: AggregatorType) -> "RapidataOrderBuilder":
         """
         Set the aggregator for the order.

{rapidata-1.0.0.dist-info → rapidata-1.1.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: rapidata
-Version: 1.0.0
+Version: 1.1.0
 Summary: Rapidata package containing the Rapidata Python Client to interact with the Rapidata Web API in an easy way.
 License: Apache-2.0
 Author: Rapidata AG
@@ -11,6 +11,7 @@ Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Requires-Dist: pillow (>=10.4.0,<11.0.0)
 Requires-Dist: pydantic (>=2.8.2,<3.0.0)
 Requires-Dist: pyjwt (>=2.9.0,<3.0.0)

{rapidata-1.0.0.dist-info → rapidata-1.1.0.dist-info}/RECORD

@@ -287,7 +287,7 @@ rapidata/rapidata_client/dataset/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRk
 rapidata/rapidata_client/dataset/rapidata_dataset.py,sha256=Rwt7sM5T74HT7gbtfjmdF9tDaPqHfpMLggIWOsYCgYU,3205
 rapidata/rapidata_client/dataset/rapidata_validation_set.py,sha256=Y7iaThkLxDyHJuKA34rgPkIQ3_8l-hJyZwfomzOyUqk,10096
 rapidata/rapidata_client/dataset/validation_rapid_parts.py,sha256=uzpOZFqQu8bG6vmjcWWUNJPZsRe28OmnyalRE6ry8tk,2317
-rapidata/rapidata_client/dataset/validation_set_builder.py,sha256=VOb0oGBljIM7CK0Xau-Ha5thHXS38d3gMUFieQMfJsA,8134
+rapidata/rapidata_client/dataset/validation_set_builder.py,sha256=MhIG6to34Sjt2q7hkWJOFVZzmz0_aI2jRrESlmsljyg,7957
 rapidata/rapidata_client/feature_flags/__init__.py,sha256=IYkcK_bZCl5RfyQFiWjjUdz4y0jipiW9qfeopq4EjQQ,40
 rapidata/rapidata_client/feature_flags/feature_flags.py,sha256=hcS9YRzpsPWpZfw-3QwSuf2TaVg-MOHBxY788oNqIW4,3957
 rapidata/rapidata_client/metadata/__init__.py,sha256=qMmo4wqScUCAJ6YXRWxvJLmbFA5YRbK39p9_exV1d50,246
@@ -298,7 +298,7 @@ rapidata/rapidata_client/metadata/public_text_metadata.py,sha256=LTiBQHs6izxQ6-C
 rapidata/rapidata_client/metadata/transcription_metadata.py,sha256=THtDEVCON4UlcXHmXrjilaOLHys4TrktUOPGWnXaCcc,631
 rapidata/rapidata_client/order/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 rapidata/rapidata_client/order/rapidata_order.py,sha256=DCfw_xH7Ja8KURlJYaxVuX6dM81sUF7HLyX4iTfqqZc,3098
-rapidata/rapidata_client/order/rapidata_order_builder.py,sha256=R_SpEImqwz1U96Ucg3sNzl3KyKpBYtOc98ujozxoc3Y,10968
+rapidata/rapidata_client/order/rapidata_order_builder.py,sha256=E879fYGbdkTfELfuWxy87tmRtrLSmIndevKujRlH8gE,11964
 rapidata/rapidata_client/rapidata_client.py,sha256=s3TlhYlvMuxcyEDIFDqpwRsP0SIi9b158uIHLUyV1gY,4699
 rapidata/rapidata_client/referee/__init__.py,sha256=Ow9MQsONhF4sX2wFK9jbvSBrpcJgtq3OglIQMkBUdIY,167
 rapidata/rapidata_client/referee/base_referee.py,sha256=bMy7cw0a-pGNbFu6u_1_Jplu0A483Ubj4oDQzh8vu8k,493
@@ -325,7 +325,7 @@ rapidata/service/local_file_service.py,sha256=pgorvlWcx52Uh3cEG6VrdMK_t__7dacQ_5
 rapidata/service/openapi_service.py,sha256=-vrM2jEzQxr9KAerOYkVhpvMEeHwjzRwm9L_VFyzOT0,1537
 rapidata/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 rapidata/utils/image_utils.py,sha256=TldO3eJWG8IhfJjm5MfNGO0mEDm1mQTsRoA0HLU1Uxs,404
-rapidata-1.0.0.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-rapidata-1.0.0.dist-info/METADATA,sha256=KOEI7yC4VdEa7msbQJQ_7ypW73a_54rpwKWgFLOJ8Og,961
-rapidata-1.0.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-rapidata-1.0.0.dist-info/RECORD,,
+rapidata-1.1.0.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+rapidata-1.1.0.dist-info/METADATA,sha256=OEWHxhQ0JHX20wGLRT9ISppda18KuLHBLLicZfwKxeE,1012
+rapidata-1.1.0.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
+rapidata-1.1.0.dist-info/RECORD,,

{rapidata-1.0.0.dist-info → rapidata-1.1.0.dist-info}/WHEEL

@@ -1,4 +1,4 @@
 Wheel-Version: 1.0
-Generator: poetry-core 1.9.0
+Generator: poetry-core 1.9.1
 Root-Is-Purelib: true
 Tag: py3-none-any