feat(file-summary): 实现文件处理技能链路

tests/test_file_summary_page_count.py

@@ -0,0 +1,66 @@
+import pytest
+from docx import Document
+from openpyxl import Workbook
+from pptx import Presentation
+
+from review_agent.file_summary.services.page_count import count_document_pages
+from review_agent.file_summary.skills.document_page_count import DocumentPageCountSkill
+from review_agent.file_summary.skills.base import WorkflowContext
+from review_agent.models import Conversation, FileSummaryBatch, FileSummaryItem
+
+
+pytestmark = pytest.mark.django_db
+
+
+def test_count_document_pages_for_office_formats(tmp_path):
+    docx_path = tmp_path / "a.docx"
+    Document().save(docx_path)
+
+    xlsx_path = tmp_path / "a.xlsx"
+    workbook = Workbook()
+    workbook.create_sheet("第二页")
+    workbook.save(xlsx_path)
+
+    pptx_path = tmp_path / "a.pptx"
+    presentation = Presentation()
+    presentation.slides.add_slide(presentation.slide_layouts[6])
+    presentation.save(pptx_path)
+
+    assert count_document_pages(docx_path).status in {"success", "uncertain"}
+    assert count_document_pages(xlsx_path).page_count == 2
+    assert count_document_pages(pptx_path).page_count == 1
+
+
+def test_document_page_count_skill_marks_unsupported_and_success(tmp_path, django_user_model):
+    xlsx_path = tmp_path / "a.xlsx"
+    workbook = Workbook()
+    workbook.save(xlsx_path)
+    txt_path = tmp_path / "a.txt"
+    txt_path.write_text("x", encoding="utf-8")
+    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-P")
+    xlsx_item = FileSummaryItem.objects.create(
+        batch=batch,
+        file_index=1,
+        file_name="a.xlsx",
+        file_type="xlsx",
+        relative_path="a.xlsx",
+        storage_path=str(xlsx_path),
+    )
+    txt_item = FileSummaryItem.objects.create(
+        batch=batch,
+        file_index=2,
+        file_name="a.txt",
+        file_type="txt",
+        relative_path="a.txt",
+        storage_path=str(txt_path),
+    )
+
+    result = DocumentPageCountSkill().run(WorkflowContext(batch=batch))
+
+    xlsx_item.refresh_from_db()
+    txt_item.refresh_from_db()
+    assert result.success is True
+    assert xlsx_item.statistics_status == FileSummaryItem.StatisticsStatus.SUCCESS
+    assert txt_item.statistics_status == FileSummaryItem.StatisticsStatus.UNSUPPORTED