feat(regulatory): 增加法规核查基础服务

tests/test_regulatory_completeness.py

@@ -0,0 +1,44 @@
+import pytest
+
+from review_agent.models import Conversation, FileSummaryBatch, FileSummaryItem
+from review_agent.regulatory_review.services.completeness_check import run_completeness_check
+from review_agent.regulatory_review.services.rule_loader import load_rule_file
+
+
+pytestmark = pytest.mark.django_db
+
+
+def test_completeness_check_matches_existing_files_and_reports_missing(django_user_model):
+    user = django_user_model.objects.create_user(username="owner", password="pass")
+    conversation = Conversation.objects.create(user=user, title="会话")
+    batch = FileSummaryBatch.objects.create(
+        conversation=conversation,
+        user=user,
+        batch_no="FS-CHECK",
+        status=FileSummaryBatch.Status.SUCCESS,
+    )
+    FileSummaryItem.objects.create(
+        batch=batch,
+        file_index=1,
+        file_name="产品技术要求.docx",
+        file_type="docx",
+        relative_path="产品技术要求.docx",
+        storage_path="x/product.docx",
+    )
+    FileSummaryItem.objects.create(
+        batch=batch,
+        file_index=2,
+        file_name="说明书.docx",
+        file_type="docx",
+        relative_path="说明书.docx",
+        storage_path="x/ifu.docx",
+    )
+
+    findings = run_completeness_check(batch, load_rule_file())
+
+    titles = [finding.title for finding in findings]
+    assert "缺少注册检验报告" in titles
+    assert "缺少产品技术要求" not in titles
+    missing = next(finding for finding in findings if finding.rule_code == "registration_test_report")
+    assert missing.severity == "blocking"
+    assert missing.category == "completeness"

tests/test_regulatory_consistency.py

@@ -0,0 +1,14 @@
+from review_agent.regulatory_review.services.consistency_check import run_consistency_check
+
+
+def test_consistency_check_reports_product_name_mismatch():
+    document_texts = {
+        "说明书.docx": "产品名称：甲胎蛋白检测试剂盒\n型号规格：20人份/盒\n预期用途：定量检测AFP",
+        "技术要求.docx": "产品名称：乙肝表面抗原检测试剂盒\n型号规格：20人份/盒\n预期用途：定量检测AFP",
+    }
+
+    findings = run_consistency_check(document_texts)
+
+    assert len(findings) == 1
+    assert findings[0].category == "consistency"
+    assert "产品名称" in findings[0].title

tests/test_regulatory_storage.py

@@ -0,0 +1,26 @@
+import pytest
+
+from review_agent.models import Conversation, FileSummaryBatch, RegulatoryReviewBatch
+from review_agent.regulatory_review.storage import save_artifact
+
+
+pytestmark = pytest.mark.django_db
+
+
+def test_save_artifact_writes_file_and_records_hash(settings, tmp_path, django_user_model):
+    settings.MEDIA_ROOT = tmp_path
+    user = django_user_model.objects.create_user(username="owner", password="pass")
+    conversation = Conversation.objects.create(user=user, title="会话")
+    summary = FileSummaryBatch.objects.create(conversation=conversation, user=user, batch_no="FS-OK")
+    batch = RegulatoryReviewBatch.objects.create(
+        conversation=conversation,
+        user=user,
+        source_summary_batch=summary,
+        batch_no="RR-ART",
+    )
+
+    artifact = save_artifact(batch, name="raw.json", content='{"ok": true}', artifact_type="json")
+
+    assert artifact.content_hash
+    assert artifact.storage_path.endswith("raw.json")
+    assert (tmp_path / "regulatory_review" / "work" / "RR-ART" / "raw.json").exists()

tests/test_regulatory_structure.py

@@ -0,0 +1,13 @@
+from review_agent.regulatory_review.services.rule_loader import load_rule_file
+from review_agent.regulatory_review.services.structure_check import run_structure_check
+
+
+def test_structure_check_reports_missing_instruction_sections():
+    document_texts = {
+        "说明书.docx": "产品名称：甲胎蛋白检测试剂盒\n样本要求：血清样本\n有效期：12个月"
+    }
+
+    findings = run_structure_check(document_texts, load_rule_file())
+
+    assert any(finding.rule_code == "instructions_for_use:储存条件" for finding in findings)
+    assert all("样本要求" not in finding.title for finding in findings)

tests/test_regulatory_text_extract.py

@@ -0,0 +1,24 @@
+from pathlib import Path
+
+from review_agent.regulatory_review.services.text_extract import extract_text
+
+
+def test_extract_text_reads_plain_text(tmp_path):
+    path = tmp_path / "说明书.txt"
+    path.write_text("产品名称：甲胎蛋白检测试剂盒\n储存条件：2-8℃", encoding="utf-8")
+
+    result = extract_text(path)
+
+    assert "甲胎蛋白" in result.text
+    assert result.status == "success"
+    assert result.content_hash
+
+
+def test_extract_text_reports_unsupported_file(tmp_path):
+    path = tmp_path / "image.png"
+    path.write_bytes(b"png")
+
+    result = extract_text(path)
+
+    assert result.status == "unsupported"
+    assert result.text == ""