feat(file-summary): 实现文件处理技能链路

review_agent/file_summary/services/inventory.py

@@ -0,0 +1,49 @@
+from __future__ import annotations
+
+from pathlib import Path
+
+from review_agent.models import FileSummaryBatch, FileSummaryItem
+
+
+SUPPORTED_EXTENSIONS = {"pdf", "doc", "docx", "xls", "xlsx", "ppt", "pptx"}
+
+
+def _directory_level(relative_path: Path) -> str:
+    if len(relative_path.parts) <= 1:
+        return ""
+    return "/".join(relative_path.parts[:-1])
+
+
+def scan_files_to_items(*, batch: FileSummaryBatch, roots: list[Path]) -> list[FileSummaryItem]:
+    files: list[tuple[Path, Path]] = []
+    for root in roots:
+        root = Path(root)
+        if root.is_file():
+            files.append((root.parent, root))
+            continue
+        for path in sorted(item for item in root.rglob("*") if item.is_file()):
+            if path.name.startswith(".") or path.stat().st_size == 0:
+                continue
+            files.append((root, path))
+
+    created: list[FileSummaryItem] = []
+    for index, (root, path) in enumerate(files, start=1):
+        relative = path.relative_to(root).as_posix()
+        file_type = path.suffix.lower().lstrip(".")
+        item = FileSummaryItem.objects.create(
+            batch=batch,
+            file_index=index,
+            directory_level=_directory_level(Path(relative)),
+            file_name=path.name,
+            file_type=file_type,
+            relative_path=relative,
+            storage_path=str(path),
+            statistics_status=FileSummaryItem.StatisticsStatus.SKIPPED,
+        )
+        created.append(item)
+
+    batch.total_files = len(created)
+    batch.supported_files = sum(1 for item in created if item.file_type in SUPPORTED_EXTENSIONS)
+    batch.unsupported_files = len(created) - batch.supported_files
+    batch.save(update_fields=["total_files", "supported_files", "unsupported_files"])
+    return created