delete items

delete-by-query.ps1

# Delete items using a query
#
# -------------
$eshost = "http://LONAPP101.uklive.exabre.co.uk:50402"
$index = "live_20_34_item"
$type = "statiogn"

$query = '
{
  "fields": [],
  "size": 100,
  "query": {
    "filtered": {
      "filter": {
        "term": {
          "Locales.Invariant.Metadata.@@station.@publisher": "abc"
        }
      }
    }
  }
}
'

# -------------

# proxy for monitoring progress, set to a url, or to $null to disable
#$proxy = $null
$proxy = 'http://127.0.0.1:8858'

# -------------

echo "This script will delete all items matching:"
echo $eshost/$index/$type
echo $query

# TODO confirmation

$choice = ""
while ($choice -ne "ok")
{
    $choice = read-host "Type 'ok' to continue or Ctrl-C to exit"
}

if ($choice -ne "ok") { return }

$totalProcessed = 0
while (1) {
    $processedThisBatch = 0
    $ids = @()
    
    # the below query selects what will be deleted 
    @($(irm "$eshost/$index/$type/_search" -Proxy $proxy -Method POST -Body $query).hits.hits `
    | foreach -process { 
        $processedThisBatch += 1
        $ids += $_._id
    })

    if ($ids.Count -gt 0)
    {
        # build up the bulk json to delete all the duplicate items for this url
        $bulks = @()
        foreach ($id in $ids) {
            $bulk = @"
    { "delete": { "_index": "$index", "_type": "$type", "_id": "$id" } }
"@

            $bulks += $bulk
        }

        $bulks += [Environment]::NewLine
        $bulkBody = $bulks -join [Environment]::NewLine | Out-String
        
        # delete the items...
        $(irm "$eshost/_bulk" -Method Post -Body $bulkBody -Proxy $proxy) | Out-Null
        echo "Deleted $processedThisBatch"
    
        # echo $ids
    }

    $totalProcessed += $totalProcessed
    if ($processedThisBatch -lt 1) { break }
}

echo "Deleted $totalProcessed for query $query"