refactor: update TextAnalyzerConfig usage to new Configure class methods

Author: amourao

integration/test_collection_config.py

@@ -30,7 +30,6 @@
     PQEncoderType,
     PQEncoderDistribution,
     StopwordsPreset,
-    TextAnalyzerConfig,
     VectorDistances,
     VectorIndexType,
     Vectorizers,
@@ -2219,7 +2218,7 @@ def test_property_text_analyzer_ascii_fold_version_gate(
                     name="title",
                     data_type=DataType.TEXT,
                     tokenization=Tokenization.WORD,
-                    text_analyzer=TextAnalyzerConfig(ascii_fold=True),
+                    text_analyzer=Configure.TextAnalyzer.ascii_fold(),
                 ),
             ],
         )

test/collection/test_config.py

@@ -4,20 +4,20 @@
 from pydantic import ValidationError
 
 from weaviate.collections.classes.config import (
-    _AsyncReplicationConfig,
-    _ReplicationConfig,
-    _ReplicationConfigUpdate,
     Configure,
     DataType,
     Property,
     Reconfigure,
     ReferenceProperty,
-    TextAnalyzerConfig,
     Tokenization,
     Vectorizers,
+    _AsyncReplicationConfig,
     _CollectionConfigCreate,
     _GenerativeProvider,
+    _ReplicationConfig,
+    _ReplicationConfigUpdate,
     _RerankerProvider,
+    _TextAnalyzerConfigCreate,
     _VectorizerConfigCreate,
     _ReplicationConfigCreate,
     ReplicationDeletionStrategy,
@@ -3025,7 +3025,7 @@ def test_nested_property_with_id_name_is_allowed() -> None:
     assert prop.nestedProperties[0].name == "id"
 
 
-class TestTextAnalyzerConfig:
+class Test_TextAnalyzerConfigCreate:
     def test_property_without_text_analyzer_omits_key(self) -> None:
         prop = Property(name="title", data_type=DataType.TEXT)
         assert "textAnalyzer" not in prop._to_dict()
@@ -3034,7 +3034,7 @@ def test_property_with_ascii_fold_only(self) -> None:
         prop = Property(
             name="title",
             data_type=DataType.TEXT,
-            text_analyzer=TextAnalyzerConfig(ascii_fold=True),
+            text_analyzer=Configure.TextAnalyzer.ascii_fold(),
         )
         assert prop._to_dict()["textAnalyzer"] == {"asciiFold": True}
 
@@ -3043,7 +3043,7 @@ def test_property_with_ascii_fold_and_ignore(self) -> None:
             name="title",
             data_type=DataType.TEXT,
             tokenization=Tokenization.WORD,
-            text_analyzer=TextAnalyzerConfig(ascii_fold=True, ascii_fold_ignore=["é", "ñ"]),
+            text_analyzer=Configure.TextAnalyzer.ascii_fold(ignore=["é", "ñ"]),
         )
         out = prop._to_dict()
         assert out["textAnalyzer"] == {
@@ -3052,22 +3052,9 @@ def test_property_with_ascii_fold_and_ignore(self) -> None:
         }
         assert out["tokenization"] == "word"
 
-    def test_text_analyzer_default_omits_unset_fields(self) -> None:
-        prop = Property(
-            name="title",
-            data_type=DataType.TEXT,
-            text_analyzer=TextAnalyzerConfig(),
-        )
-        # exclude_none drops both unset fields, leaving an empty dict
-        assert prop._to_dict()["textAnalyzer"] == {}
-
-    def test_text_analyzer_only_ignore_list(self) -> None:
-        prop = Property(
-            name="title",
-            data_type=DataType.TEXT,
-            text_analyzer=TextAnalyzerConfig(ascii_fold_ignore=["é"]),
-        )
-        assert prop._to_dict()["textAnalyzer"] == {"asciiFoldIgnore": ["é"]}
+    def test_text_analyzer_rejects_ignore_without_ascii_fold(self) -> None:
+        with pytest.raises(ValidationError):
+            _TextAnalyzerConfigCreate(ascii_fold_ignore=["é"])
 
     def test_nested_property_with_text_analyzer(self) -> None:
         prop = Property(
@@ -3077,7 +3064,7 @@ def test_nested_property_with_text_analyzer(self) -> None:
                 Property(
                     name="title",
                     data_type=DataType.TEXT,
-                    text_analyzer=TextAnalyzerConfig(ascii_fold=True, ascii_fold_ignore=["ñ"]),
+                    text_analyzer=Configure.TextAnalyzer.ascii_fold(ignore=["ñ"]),
                 ),
             ],
         )
@@ -3087,13 +3074,8 @@ def test_nested_property_with_text_analyzer(self) -> None:
             "asciiFoldIgnore": ["ñ"],
         }
 
-    def test_text_analyzer_accepts_snake_case_alias(self) -> None:
-        ta = TextAnalyzerConfig(ascii_fold=True, ascii_fold_ignore=["é"])
-        assert ta.asciiFold is True
-        assert ta.asciiFoldIgnore == ["é"]
-
     def test_text_analyzer_rejects_wrong_types(self) -> None:
         with pytest.raises(ValidationError):
-            TextAnalyzerConfig(ascii_fold="yes")  # type: ignore[arg-type]
+            _TextAnalyzerConfigCreate(ascii_fold="yes")  # type: ignore[arg-type]
         with pytest.raises(ValidationError):
-            TextAnalyzerConfig(ascii_fold_ignore="é")  # type: ignore[arg-type]
+            _TextAnalyzerConfigCreate(ascii_fold_ignore="é")  # type: ignore[arg-type]

weaviate/collections/classes/config.py

@@ -15,7 +15,14 @@
 )
 
 from deprecation import deprecated as docstring_deprecated
-from pydantic import AnyHttpUrl, Field, TypeAdapter, ValidationInfo, field_validator, model_validator
+from pydantic import (
+    AnyHttpUrl,
+    Field,
+    TypeAdapter,
+    ValidationInfo,
+    field_validator,
+    model_validator,
+)
 from typing_extensions import TypeAlias
 from typing_extensions import deprecated as typing_deprecated
 
@@ -1677,6 +1684,9 @@ class _TextAnalyzerConfig(_ConfigBase):
     ascii_fold_ignore: Optional[List[str]]
 
 
+TextAnalyzerConfig = _TextAnalyzerConfig
+
+
 @dataclass
 class _NestedProperty(_ConfigBase):
     data_type: DataType
@@ -2171,7 +2181,7 @@ class _ShardStatus:
 ShardStatus = _ShardStatus
 
 
-class TextAnalyzerConfig(_ConfigCreateModel):
+class _TextAnalyzerConfigCreate(_ConfigCreateModel):
     """Text analysis options for a property.
 
     Configures ASCII folding behavior for `text` and `text[]` properties that use an
@@ -2194,7 +2204,7 @@ class TextAnalyzerConfig(_ConfigCreateModel):
     asciiFoldIgnore: Optional[List[str]] = Field(default=None, alias="ascii_fold_ignore")
 
     @model_validator(mode="after")
-    def _validate_ascii_fold_ignore(self) -> "TextAnalyzerConfig":
+    def _validate_ascii_fold_ignore(self) -> "_TextAnalyzerConfigCreate":
         if self.asciiFold is not True and self.asciiFoldIgnore is not None:
             raise ValueError("asciiFoldIgnore cannot be set when asciiFold is not enabled")
         return self
@@ -2209,14 +2219,14 @@ class _TextAnalyzer:
     @staticmethod
     def ascii_fold(
         ignore: Optional[List[str]] = None,
-    ) -> _TextAnalyzerConfig:
+    ) -> _TextAnalyzerConfigCreate:
         """Create a text analyzer config with ASCII folding enabled.
 
         Args:
             ignore: Optional list of characters that should be excluded from
                 ASCII folding (e.g. ``['é']`` keeps 'é' from being folded to 'e').
         """
-        return _TextAnalyzerConfig(asciiFold=True, asciiFoldIgnore=ignore)
+        return _TextAnalyzerConfigCreate(ascii_fold=True, ascii_fold_ignore=ignore)
 
 
 class Property(_ConfigCreateModel):
@@ -2248,7 +2258,7 @@ class Property(_ConfigCreateModel):
         default=None, alias="nested_properties"
     )
     skip_vectorization: bool = Field(default=False)
-    textAnalyzer: Optional[TextAnalyzerConfig] = Field(default=None, alias="text_analyzer")
+    textAnalyzer: Optional[_TextAnalyzerConfigCreate] = Field(default=None, alias="text_analyzer")
     tokenization: Optional[Tokenization] = Field(default=None)
     vectorize_property_name: bool = Field(default=True)