PowerShell复杂备份恢复脚本求助：文件还原路径异常排查

Hey there! Let's work through optimizing your PowerShell backup/restore script to fix that frustrating drive-not-found error and clean up the logic.

First, let's break down the core issue you hit: when using pscustomobject to track file paths, passing the entire object directly to Copy-Item causes PowerShell to convert it to a string (like @{name=Device.txt; originpath=C:\...}), which the cmdlet misinterprets as a drive name. The temporary fix of assigning properties to variables works, but we can make the script cleaner and more robust.

Optimized Script: Backup + Restore with Structured Records #

This version uses a structured approach to track file details, supports both in-memory storage (ArrayList) and persistent CSV storage, and avoids the path misinterpretation bug.

Step 1: Backup Phase (Track Files & Save Records)

# Define your list of files to backup (can also import this from a CSV if needed)
$targetFiles = @(
    "C:\Example\Device.txt",
    "D:\Config\Settings.ini",
    "E:\Documents\CriticalFile.docx"
)

# Set your backup directory (create it if it doesn't exist)
$backupRoot = "C:\MySecureBackup"
if (-not (Test-Path $backupRoot)) {
    New-Item -ItemType Directory -Path $backupRoot | Out-Null
}

# Initialize a collection to store backup records
$backupLog = @()

foreach ($file in $targetFiles) {
    # Skip if the source file doesn't exist
    if (-not (Test-Path $file -PathType Leaf)) {
        Write-Warning "Skipping missing file: $file"
        continue
    }

    # Get file details and build backup path
    $fileName = Split-Path $file -Leaf
    $backupPath = Join-Path $backupRoot $fileName

    # Perform the backup
    Copy-Item -Path $file -Destination $backupPath -Force

    # Create a structured record to track original + backup paths
    $record = [PSCustomObject]@{
        OriginalFullPath = $file
        BackupFullPath   = $backupPath
        FileName         = $fileName
    }

    $backupLog += $record
}

# Optional: Save records to CSV for cross-session restore
$backupLog | Export-Csv -Path "$backupRoot\BackupManifest.csv" -NoTypeInformation
Write-Host "Backup completed. $($backupLog.Count) files backed up."

Step 2: Restore Phase (Use Records to Restore to Original Paths)

# Load backup records (use CSV if restoring in a new session, or $backupLog if same session)
$backupLog = Import-Csv -Path "$backupRoot\BackupManifest.csv"

foreach ($entry in $backupLog) {
    # Extract the actual path values (this fixes the drive-not-found error!)
    $originalPath = $entry.OriginalFullPath
    $backupSource = $entry.BackupFullPath

    # Verify backup file exists before restoring
    if (Test-Path $backupSource -PathType Leaf) {
        Copy-Item -Path $backupSource -Destination $originalPath -Force
        Write-Host "Successfully restored: $originalPath"
    } else {
        Write-Warning "Backup file missing, cannot restore: $backupSource"
    }
}

Key Improvements Over Your Initial Script #

• Clear Structured Tracking: The pscustomobject explicitly stores OriginalFullPath and BackupFullPath, eliminating logic confusion from vague array structures.
• Error Prevention: We validate file existence before backup/restore, so you get clear warnings instead of silent failures.
• No More Drive Errors: By explicitly extracting property values ($entry.OriginalFullPath instead of passing $entry), we avoid PowerShell converting the object to a garbage string path.
• Flexible Storage: Choose between in-memory $backupLog (for single-session operations) or CSV (for restoring days/weeks later without re-running the backup).

This approach stays true to your requirement of avoiding simple A→B→A copies, and keeps the script maintainable if you need to add more files or adjust logic later.

内容的提问来源于stack exchange，提问作者Kevin