refactor: use EnumText for ApiToolProvider.schema_type_str and Docume… (#33983)

api/tests/unit_tests/services/document_service_validation.py

@@ -111,6 +111,7 @@ from unittest.mock import Mock, patch
 import pytest
 
 from core.errors.error import LLMBadRequestError, ProviderTokenNotInitError
+from core.rag.index_processor.constant.index_type import IndexStructureType
 from dify_graph.model_runtime.entities.model_entities import ModelType
 from models.dataset import Dataset, DatasetProcessRule, Document
 from services.dataset_service import DatasetService, DocumentService
@@ -188,7 +189,7 @@ class DocumentValidationTestDataFactory:
     def create_knowledge_config_mock(
         data_source: DataSource | None = None,
         process_rule: ProcessRule | None = None,
-        doc_form: str = "text_model",
+        doc_form: str = IndexStructureType.PARAGRAPH_INDEX,
         indexing_technique: str = "high_quality",
         **kwargs,
     ) -> Mock:
@@ -326,8 +327,8 @@ class TestDatasetServiceCheckDocForm:
         - Validation logic works correctly
         """
         # Arrange
-        dataset = DocumentValidationTestDataFactory.create_dataset_mock(doc_form="text_model")
-        doc_form = "text_model"
+        dataset = DocumentValidationTestDataFactory.create_dataset_mock(doc_form=IndexStructureType.PARAGRAPH_INDEX)
+        doc_form = IndexStructureType.PARAGRAPH_INDEX
 
         # Act (should not raise)
         DatasetService.check_doc_form(dataset, doc_form)
@@ -349,7 +350,7 @@ class TestDatasetServiceCheckDocForm:
         """
         # Arrange
         dataset = DocumentValidationTestDataFactory.create_dataset_mock(doc_form=None)
-        doc_form = "text_model"
+        doc_form = IndexStructureType.PARAGRAPH_INDEX
 
         # Act (should not raise)
         DatasetService.check_doc_form(dataset, doc_form)
@@ -370,8 +371,8 @@ class TestDatasetServiceCheckDocForm:
         - Error type is correct
         """
         # Arrange
-        dataset = DocumentValidationTestDataFactory.create_dataset_mock(doc_form="text_model")
-        doc_form = "table_model"  # Different form
+        dataset = DocumentValidationTestDataFactory.create_dataset_mock(doc_form=IndexStructureType.PARAGRAPH_INDEX)
+        doc_form = IndexStructureType.PARENT_CHILD_INDEX  # Different form
 
         # Act & Assert
         with pytest.raises(ValueError, match="doc_form is different from the dataset doc_form"):
@@ -390,7 +391,7 @@ class TestDatasetServiceCheckDocForm:
         """
         # Arrange
         dataset = DocumentValidationTestDataFactory.create_dataset_mock(doc_form="knowledge_card")
-        doc_form = "text_model"  # Different form
+        doc_form = IndexStructureType.PARAGRAPH_INDEX  # Different form
 
         # Act & Assert
         with pytest.raises(ValueError, match="doc_form is different from the dataset doc_form"):

api/tests/unit_tests/services/segment_service.py

@@ -2,6 +2,7 @@ from unittest.mock import MagicMock, Mock, patch
 
 import pytest
 
+from core.rag.index_processor.constant.index_type import IndexStructureType
 from models.account import Account
 from models.dataset import ChildChunk, Dataset, Document, DocumentSegment
 from models.enums import SegmentType
@@ -91,7 +92,7 @@ class SegmentTestDataFactory:
         document_id: str = "doc-123",
         dataset_id: str = "dataset-123",
         tenant_id: str = "tenant-123",
-        doc_form: str = "text_model",
+        doc_form: str = IndexStructureType.PARAGRAPH_INDEX,
         word_count: int = 100,
         **kwargs,
     ) -> Mock:
@@ -210,7 +211,7 @@ class TestSegmentServiceCreateSegment:
     def test_create_segment_with_qa_model(self, mock_db_session, mock_current_user):
         """Test creation of segment with QA model (requires answer)."""
         # Arrange
-        document = SegmentTestDataFactory.create_document_mock(doc_form="qa_model", word_count=100)
+        document = SegmentTestDataFactory.create_document_mock(doc_form=IndexStructureType.QA_INDEX, word_count=100)
         dataset = SegmentTestDataFactory.create_dataset_mock(indexing_technique="economy")
         args = {"content": "What is AI?", "answer": "AI is Artificial Intelligence", "keywords": ["ai"]}
 
@@ -429,7 +430,7 @@ class TestSegmentServiceUpdateSegment:
         """Test update segment with QA model (includes answer)."""
         # Arrange
         segment = SegmentTestDataFactory.create_segment_mock(enabled=True, word_count=10)
-        document = SegmentTestDataFactory.create_document_mock(doc_form="qa_model", word_count=100)
+        document = SegmentTestDataFactory.create_document_mock(doc_form=IndexStructureType.QA_INDEX, word_count=100)
         dataset = SegmentTestDataFactory.create_dataset_mock(indexing_technique="economy")
         args = SegmentUpdateArgs(content="Updated question", answer="Updated answer", keywords=["qa"])
 

api/tests/unit_tests/services/test_dataset_service_lock_not_owned.py

@@ -4,6 +4,7 @@ from unittest.mock import Mock, create_autospec
 import pytest
 from redis.exceptions import LockNotOwnedError
 
+from core.rag.index_processor.constant.index_type import IndexStructureType
 from models.account import Account
 from models.dataset import Dataset, Document
 from services.dataset_service import DocumentService, SegmentService
@@ -76,7 +77,7 @@ def test_save_document_with_dataset_id_ignores_lock_not_owned(
     info_list = types.SimpleNamespace(data_source_type="upload_file")
     data_source = types.SimpleNamespace(info_list=info_list)
     knowledge_config = types.SimpleNamespace(
-        doc_form="qa_model",
+        doc_form=IndexStructureType.QA_INDEX,
         original_document_id=None,  # go into "new document" branch
         data_source=data_source,
         indexing_technique="high_quality",
@@ -131,7 +132,7 @@ def test_add_segment_ignores_lock_not_owned(
     document.id = "doc-1"
     document.dataset_id = dataset.id
     document.word_count = 0
-    document.doc_form = "qa_model"
+    document.doc_form = IndexStructureType.QA_INDEX
 
     # Minimal args required by add_segment
     args = {
@@ -174,4 +175,4 @@ def test_multi_create_segment_ignores_lock_not_owned(
     document.id = "doc-1"
     document.dataset_id = dataset.id
     document.word_count = 0
-    document.doc_form = "qa_model"
+    document.doc_form = IndexStructureType.QA_INDEX

api/tests/unit_tests/services/test_summary_index_service.py

@@ -11,6 +11,7 @@ from unittest.mock import MagicMock
 import pytest
 
 import services.summary_index_service as summary_module
+from core.rag.index_processor.constant.index_type import IndexStructureType
 from models.enums import SegmentStatus, SummaryStatus
 from services.summary_index_service import SummaryIndexService
 
@@ -48,7 +49,7 @@ def _segment(*, has_document: bool = True) -> MagicMock:
     if has_document:
         doc = MagicMock(name="document")
         doc.doc_language = "en"
-        doc.doc_form = "text_model"
+        doc.doc_form = IndexStructureType.PARAGRAPH_INDEX
         segment.document = doc
     else:
         segment.document = None
@@ -623,13 +624,13 @@ def test_generate_summaries_for_document_skip_conditions(monkeypatch: pytest.Mon
     dataset = _dataset(indexing_technique="economy")
     document = MagicMock(spec=summary_module.DatasetDocument)
     document.id = "doc-1"
-    document.doc_form = "text_model"
+    document.doc_form = IndexStructureType.PARAGRAPH_INDEX
     assert SummaryIndexService.generate_summaries_for_document(dataset, document, {"enable": True}) == []
 
     dataset = _dataset()
     assert SummaryIndexService.generate_summaries_for_document(dataset, document, {"enable": False}) == []
 
-    document.doc_form = "qa_model"
+    document.doc_form = IndexStructureType.QA_INDEX
     assert SummaryIndexService.generate_summaries_for_document(dataset, document, {"enable": True}) == []
 
 
@@ -637,7 +638,7 @@ def test_generate_summaries_for_document_runs_and_handles_errors(monkeypatch: py
     dataset = _dataset()
     document = MagicMock(spec=summary_module.DatasetDocument)
     document.id = "doc-1"
-    document.doc_form = "text_model"
+    document.doc_form = IndexStructureType.PARAGRAPH_INDEX
 
     seg1 = _segment()
     seg2 = _segment()
@@ -673,7 +674,7 @@ def test_generate_summaries_for_document_no_segments_returns_empty(monkeypatch:
     dataset = _dataset()
     document = MagicMock(spec=summary_module.DatasetDocument)
     document.id = "doc-1"
-    document.doc_form = "text_model"
+    document.doc_form = IndexStructureType.PARAGRAPH_INDEX
 
     session = MagicMock()
     query = MagicMock()
@@ -696,7 +697,7 @@ def test_generate_summaries_for_document_applies_segment_ids_and_only_parent_chu
     dataset = _dataset()
     document = MagicMock(spec=summary_module.DatasetDocument)
     document.id = "doc-1"
-    document.doc_form = "text_model"
+    document.doc_form = IndexStructureType.PARAGRAPH_INDEX
     seg = _segment()
 
     session = MagicMock()
@@ -935,7 +936,7 @@ def test_update_summary_for_segment_skip_conditions() -> None:
         SummaryIndexService.update_summary_for_segment(_segment(), _dataset(indexing_technique="economy"), "x") is None
     )
     seg = _segment(has_document=True)
-    seg.document.doc_form = "qa_model"
+    seg.document.doc_form = IndexStructureType.QA_INDEX
     assert SummaryIndexService.update_summary_for_segment(seg, _dataset(), "x") is None
 
 

api/tests/unit_tests/services/test_vector_service.py

@@ -9,6 +9,7 @@ from unittest.mock import MagicMock
 import pytest
 
 import services.vector_service as vector_service_module
+from core.rag.index_processor.constant.index_type import IndexStructureType
 from services.vector_service import VectorService
 
 
@@ -32,7 +33,7 @@ class _ParentDocStub:
 def _make_dataset(
     *,
     indexing_technique: str = "high_quality",
-    doc_form: str = "text_model",
+    doc_form: str = IndexStructureType.PARAGRAPH_INDEX,
     tenant_id: str = "tenant-1",
     dataset_id: str = "dataset-1",
     is_multimodal: bool = False,
@@ -106,7 +107,7 @@ def test_create_segments_vector_regular_indexing_loads_documents_and_keywords(mo
     factory_instance.init_index_processor.return_value = index_processor
     monkeypatch.setattr(vector_service_module, "IndexProcessorFactory", MagicMock(return_value=factory_instance))
 
-    VectorService.create_segments_vector([["k1"]], [segment], dataset, "text_model")
+    VectorService.create_segments_vector([["k1"]], [segment], dataset, IndexStructureType.PARAGRAPH_INDEX)
 
     index_processor.load.assert_called_once()
     args, kwargs = index_processor.load.call_args
@@ -131,7 +132,7 @@ def test_create_segments_vector_regular_indexing_loads_multimodal_documents(monk
     factory_instance.init_index_processor.return_value = index_processor
     monkeypatch.setattr(vector_service_module, "IndexProcessorFactory", MagicMock(return_value=factory_instance))
 
-    VectorService.create_segments_vector([["k1"]], [segment], dataset, "text_model")
+    VectorService.create_segments_vector([["k1"]], [segment], dataset, IndexStructureType.PARAGRAPH_INDEX)
 
     assert index_processor.load.call_count == 2
     first_args, first_kwargs = index_processor.load.call_args_list[0]
@@ -153,7 +154,7 @@ def test_create_segments_vector_with_no_segments_does_not_load(monkeypatch: pyte
     factory_instance.init_index_processor.return_value = index_processor
     monkeypatch.setattr(vector_service_module, "IndexProcessorFactory", MagicMock(return_value=factory_instance))
 
-    VectorService.create_segments_vector(None, [], dataset, "text_model")
+    VectorService.create_segments_vector(None, [], dataset, IndexStructureType.PARAGRAPH_INDEX)
     index_processor.load.assert_not_called()
 
 
@@ -392,7 +393,7 @@ def test_update_segment_vector_economy_uses_keyword_without_keywords_list(monkey
 
 
 def test_generate_child_chunks_regenerate_cleans_then_saves_children(monkeypatch: pytest.MonkeyPatch) -> None:
-    dataset = _make_dataset(doc_form="text_model", tenant_id="tenant-1", dataset_id="dataset-1")
+    dataset = _make_dataset(doc_form=IndexStructureType.PARAGRAPH_INDEX, tenant_id="tenant-1", dataset_id="dataset-1")
     segment = _make_segment(segment_id="seg-1")
 
     dataset_document = MagicMock()
@@ -439,7 +440,7 @@ def test_generate_child_chunks_regenerate_cleans_then_saves_children(monkeypatch
 
 
 def test_generate_child_chunks_commits_even_when_no_children(monkeypatch: pytest.MonkeyPatch) -> None:
-    dataset = _make_dataset(doc_form="text_model")
+    dataset = _make_dataset(doc_form=IndexStructureType.PARAGRAPH_INDEX)
     segment = _make_segment()
     dataset_document = MagicMock()
     dataset_document.doc_language = "en"

api/tests/unit_tests/services/vector_service.py

@@ -121,6 +121,7 @@ import pytest
 from core.rag.datasource.vdb.vector_base import BaseVector
 from core.rag.datasource.vdb.vector_factory import Vector
 from core.rag.datasource.vdb.vector_type import VectorType
+from core.rag.index_processor.constant.index_type import IndexStructureType
 from core.rag.models.document import Document
 from models.dataset import ChildChunk, Dataset, DatasetDocument, DatasetProcessRule, DocumentSegment
 from services.vector_service import VectorService
@@ -151,7 +152,7 @@ class VectorServiceTestDataFactory:
     def create_dataset_mock(
         dataset_id: str = "dataset-123",
         tenant_id: str = "tenant-123",
-        doc_form: str = "text_model",
+        doc_form: str = IndexStructureType.PARAGRAPH_INDEX,
         indexing_technique: str = "high_quality",
         embedding_model_provider: str = "openai",
         embedding_model: str = "text-embedding-ada-002",
@@ -493,7 +494,7 @@ class TestVectorService:
         """
         # Arrange
         dataset = VectorServiceTestDataFactory.create_dataset_mock(
-            doc_form="text_model", indexing_technique="high_quality"
+            doc_form=IndexStructureType.PARAGRAPH_INDEX, indexing_technique="high_quality"
         )
 
         segment = VectorServiceTestDataFactory.create_document_segment_mock()
@@ -505,7 +506,7 @@ class TestVectorService:
         mock_index_processor_factory.return_value.init_index_processor.return_value = mock_index_processor
 
         # Act
-        VectorService.create_segments_vector(keywords_list, [segment], dataset, "text_model")
+        VectorService.create_segments_vector(keywords_list, [segment], dataset, IndexStructureType.PARAGRAPH_INDEX)
 
         # Assert
         mock_index_processor.load.assert_called_once()
@@ -649,7 +650,7 @@ class TestVectorService:
         mock_index_processor_factory.return_value.init_index_processor.return_value = mock_index_processor
 
         # Act
-        VectorService.create_segments_vector(None, [], dataset, "text_model")
+        VectorService.create_segments_vector(None, [], dataset, IndexStructureType.PARAGRAPH_INDEX)
 
         # Assert
         mock_index_processor.load.assert_not_called()