File Operations

What is this block?

The File Operations block manages files and folders inside your RAPFlow project storage. It supports:

• Create folders
• Move or rename items
• Delete by explicit names and/or glob patterns
• Search/list file/folder metadata
• Download external files into the project storage

Inputs and Configuration

Project base directory

The block automatically derives the storage root from the current RAPFlow project context.

• Do not provide msg.payload.path (the block overwrites it).
• Treat all paths you pass in msg.payload as relative to this storage root.

Depending on the operation, you’ll provide some of these keys in msg.payload:

• names (array of strings): items to create/delete/list/download
• sources (array of strings): items to move (for move)
• sources (string): item to rename (for rename)
• source (string, optional): rename workaround in some deployments
• destination (string): target directory/path (move/rename)
• pattern (string): glob pattern for delete/search
• destination_dir (string): download destination folder (relative)

Create Folder (create)

Required:

• msg.payload.names: array of folder paths to create (relative)

Move (move)

Required:

• msg.payload.sources: array of file/folder paths to move (relative)
• msg.payload.destination: destination directory/path (relative)

Rename (rename)

Required:

• msg.payload.sources: a single file/folder path to rename (string, not a list)
• msg.payload.destination: the new path/name (relative)

If rename fails in your environment, also try setting msg.payload.source to the same value as msg.payload.sources.

Delete (delete)

Provide at least one:

• msg.payload.names: array of items to delete
• msg.payload.pattern: glob pattern to match items

The block will delete files or folders based on the selected resource_type.

Search / List (search_list)

Provide either/both:

• msg.payload.names: array of explicit items to list
• msg.payload.pattern: glob pattern

If you provide neither, it defaults to listing everything (pattern="*") under the base directory.

Download (download)

Required:

• msg.payload.names: array of identifiers to download (URLs, object keys, file IDs, etc. depending on the selected Source)

Optional:

• msg.payload.destination_dir: relative destination folder

Configured in the block UI:

• source: s3, gdrive, onedrive, sharepoint, wget, or url
• file_access_type: public or private for Google Drive, OneDrive, and SharePoint

Additional runtime requirements depend on the selected source:

• s3: provide bucket_name in the block config or in msg.payload
• private gdrive / onedrive / sharepoint: provide an access token
• private onedrive / sharepoint: also provide drive_id
• public onedrive / sharepoint: names must contain share URLs

Outputs

On success, the block returns an object in msg.payload from the File Operations service. Common keys include status, and operation-specific fields like:

• input / paths / items

Typical output shapes:

• Create: { "status": "success", "names": [...], "type": "folder" }
• Move/Rename: { "status": "success", "input": { "sources": ..., "destination": ... } }
• Delete: { "status": "success", "input": { "names"?: [...], "pattern"?: "..." } }
• Search/List: { "status": "success", "items": [ { "name": "...", "type": "...", ... } ] }
• Download: { "status": "success" | "partial_success" | "failed", "paths": [...], ... }

Example

Create folders

{ "names": ["datasets/invoices", "datasets/2026"] }

Move files

{
  "sources": ["inbox/a.txt", "inbox/b.txt"],
  "destination": "archive/2026-02-06"
}

Delete by glob pattern

{ "pattern": "tmp/**/*.log" }

Download from URLs

{
  "destination_dir": "downloads",
  "names": ["https://example.com/file1.pdf", "https://example.com/file2.pdf"]
}

Limitations

• All paths/identifiers must be relative to the project storage root.
• Download enforces extension and MIME allowlists, optional virus scanning, and a maximum file size; disallowed files are rejected.
• Downloads can be partially successful (partial_success) if some files fail.

Common Mistakes

• Trying to set msg.payload.path (the block overwrites it).
• Using absolute paths in names/sources.
• For delete, providing neither msg.payload.names nor msg.payload.pattern.
• For rename, sending msg.payload.sources as a list instead of a single string.
• Assuming download always succeeds for every item.