Feature Request: Improve usability of `proxmox_files` for conditional file usage

[dgomesb]

Description

The proxmox_files data source is very useful for listing files in a Proxmox datastore. However, using it in combination with proxmox_virtual_environment_download_file to implement a very common workflow is currently more complex than expected.

This is especially noticeable when working with cloud images (content_type = "import") used as base disks for virtual machines.

---

Problem

A frequent use case is:

"Use an existing base image if it exists in the datastore; otherwise, download it and use it."

While the provider already offers all the necessary building blocks (proxmox_files and proxmox_virtual_environment_download_file), implementing this behavior requires a significant amount of boilerplate and manual logic.

---

Current Workaround (Full Example)

locals {
  target_file_name = "noble-server-cloudimg-amd64.img"

  matching_files = [
    for f in data.proxmox_files.images.files : f if f.name == local.target_file_name
  ]

  image_file_id = length(local.matching_files) > 0 ? local.matching_files[0].id : proxmox_virtual_environment_download_file.ubuntu_image[0].id
}

data "proxmox_files" "images" {
  node_name    = "pve"
  datastore_id = "local"
  content_type = "import"
}

resource "proxmox_virtual_environment_download_file" "ubuntu_image" {
  count = length(local.matching_files) == 0 ? 1 : 0

  node_name    = "pve"
  datastore_id   = "local"
  content_type  = "import"

  url              = "https://cloud-images.ubuntu.com/noble/current/noble-server-cloudimg-amd64.img"
  file_name = "noble-server-cloudimg-amd64.img"
}

resource "proxmox_virtual_environment_vm" "vm" {
  name      = "example-vm"
  node_name = "pve"

  disk {
    datastore_id = "local"
    file_id      = local.image_file_id
  }
}

---

Why this is problematic

For such a common workflow, the current approach requires:

• Manual iteration and filtering over files
• Multiple locals variables
• Conditional resource creation using count
• Conditional selection of IDs
• Handling Terraform indexing ([0])

This increases complexity and reduces readability.

---

Proposed Improvement

Improve the usability of proxmox_files by allowing direct lookup of a file by name.

Example (hypothetical)

data "proxmox_files" "base_image" {
  node_name    = var.node_name
  datastore_id = var.datastore_id
  content_type = "import"

  file_name = "noble-server-cloudimg-amd64.qcow2"
}

Expected behavior

• If the file exists → return it (e.g., expose id)
• If the file does not exist → return null (or empty result)

---

Simplified Workflow with This Improvement

resource "proxmox_virtual_environment_download_file" "base_image" {
  count = data.proxmox_files.base_image.id == null ? 1 : 0

  node_name    = var.node_name
  datastore_id = var.datastore_id
  content_type = "import"

  url  = "https://cloud-images.ubuntu.com/noble/current/noble-server-cloudimg-amd64.img"
  file_name = "noble-server-cloudimg-amd64.qcow2"
}

resource "proxmox_virtual_environment_vm" "vm" {
  name      = "example-vm"
  node_name = var.node_name

  disk {
    datastore_id = var.datastore_id
    import_from      = data.proxmox_files.base_image.id != null
    ? data.proxmox_files.base_image.id
    : proxmox_virtual_environment_download_file.base_image[0].id
  }
}

---

Benefits

• Eliminates manual filtering logic
• Reduces boilerplate significantly
• Keeps download responsibility in proxmox_virtual_environment_download_file
• Makes the workflow much more intuitive and readable
• Better aligns with common Terraform patterns

---

Summary

While the provider already supports all required operations, the current user experience for "use existing or download if missing" is unnecessarily complex.

Allowing proxmox_files to query directly by file name would greatly simplify this workflow without changing existing responsibilities between data sources and resources.

Thank you for your work on this provider!