fix(storage): disambiguate storage entries with duplicate paths

[Light2Dark]

📝 Summary

Split out from #9708.

Some backends (e.g. Google Drive) allow multiple files to share the same path, which made entry rows collide when keyed by path alone. This surfaces the backend's stable id through the fsspec backend and prefers it when keying entry rows in the storage inspector, falling back to path + index when no id is available.

📋 Pre-Review Checklist

• For large changes, or changes that affect the public API: this change was discussed or approved through an issue, on Discord, or the community discussions (Please provide a link if applicable).
• Any AI generated code has been reviewed line-by-line by the human PR author, who stands by it.
• Video or media evidence is provided for any visual changes (optional).

✅ Merge Checklist

• I have read the contributor guidelines.
• Documentation has been updated where applicable, including docstrings for API changes.
• Tests have been added for the changes made.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
marimo-docs	Ready	Preview, Comment	Jun 9, 2026 10:11am

[Copilot]

Pull request overview

Disambiguates remote storage entries that can legitimately share the same path (e.g. Google Drive) by surfacing a stable backend id to the frontend and using it for row identity in the storage inspector.

Changes:

• Propagate backend-provided id into StorageEntry.metadata for fsspec-based storage entries.
• Update the storage inspector to key/render rows using metadata.id when present, falling back to a derived key when absent.
• Add backend + frontend tests to cover the new id plumbing and keying behavior.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
tests/_data/_external_storage/test_storage_models.py	Adds coverage to ensure backend id is preserved in StorageEntry.metadata.
marimo/_data/_external_storage/storage.py	Includes id in the fsspec entry metadata created from ls(..., detail=True) results.
frontend/src/components/storage/storage-inspector.tsx	Uses a stable row key based on backend id (with fallback) and threads it into row rendering.
frontend/src/components/storage/tests/storage-inspector.test.ts	Adds unit tests for storageEntryKey behavior with/without backend ids.

[cubic-dev-ai]

Review completed

Architecture diagram

sequenceDiagram
    participant UI as Storage Inspector UI
    participant Backend as Backend API (fsspec)
    participant DB as Google Drive / Backend
    participant State as React Component State

    Note over UI,State: Entry listing from backend

    UI->>Backend: listFiles(namespace, path)
    Backend->>DB: Fetch file metadata
    DB-->>Backend: File list (name, id, ...)
    Backend-->>UI: StorageEntry[] (metadata.id populated)

    Note over UI,State: Render each entry with a stable key

    UI->>UI: For each entry + index → storageEntryKey()

    alt metadata.id is a non-empty string (e.g. Google Drive)
        UI->>UI: Use id as key (no index needed)
    else metadata.id missing / empty / non-string
        UI->>UI: Fallback to path::index
    end

    UI->>State: set entries with unique keys
    State-->>UI: Re-render rows

    Note over UI,State: Row selection / navigation

    UI->>UI: User clicks or selects row
    UI->>UI: Build value = namespace:rowKey
    UI->>UI: Handle click (expand dirs, open files)

    alt Directory click
        UI->>Backend: Recurse into directory (re-requests children)
        Backend-->>UI: New StorageEntry[] for sub-path
    else File click
        UI->>UI: onOpenFile(rowKey)
    end

Loading

_{Re-trigger cubic}

[kirangadhave]

🚀 minor nit re exports

[kirangadhave]

why not just export it normally?

or add barrel exports from an index.ts where we skip storageEntryKey, then the tests can export directly from the file.