puppet-filebucket-util/bucket-tool

#!/usr/bin/ruby

# BEGIN: Requires

require 'digest'

# END: Requires



# BEGIN: Helper Functions

def usage
  puts <<EOT
#{__FILE__} [ACTION <arg>] [<flags>]

Description:
  This utlity is meant to be used to interact with & manage the filebucket on P4 nodes due to the 
  utility for that `puppet filebucket -l <action>` being nonfunctional.

  This implements the same functionality (minus the puppet tie-in) and will allow the user to 
  search the filebucket and restore from it.

Actions:
  search <term>             : Search for bucket entries matching a portion of the filepath
  list                      : List all Bucket entries
  list-files                : List all files/paths that have been backed up to the bucket
  get <entry-hash>          : Get the content of a specific entry (by hash)
  restore <value>           : Restore previous state of file stored in bucket. Value can be hash or filename/filepath 
                            : Note: To restore to an alternate path you will need to provide the path via the -o flag
                            :
  backup <file>             : Backup the file to the bucket (will work relatively unless full path is provided)

Restore Flags:
  -o | --output-file <file> : Used to provide an alternate restoral path for the restore function

Global Flags:
  -d | --debug              : Set debug flag
  -h | --help               : This help message

Listing Filter Flags:
  --from-date <DATE>        : Filter listings from starting after this point (FORMAT: YYYY-MM-DD HH:MM:SS)
                            : EX:
                            : 1.  --from-date 2023-05-10
                            : 2.  --from-date "2023-05-10 13:23"
                            :
  --to-date <DATE>          : Filter listings from ending at this point (FORMAT: YYYY-MM-DD HH:MM:SS)
                            : EX: 
                            : 1.  --to-date 2024-05-10
                            : 2.  --to-date "2024-05-10 05:27"
                            :


Info Format Flags:
  -i | --inline             : Set the info format to inline (MTIME : HASH : FILENAME)
  -l | --long               : Set the info format to long
                            : Entry [HASH]:
                            :   Paths: path1,path2,...,pathn
                            :   MTIME: YYYY-MM-DD HH:MM:SS -####
                            :
  -c | --csv                : Set the info format to csv ( MTIME,HASH,FILENAME1[;FILENAMEn] )

Author:
  Name: Tristan Ancelet
  Email: tristan.ancelet@acumera.com
  Phone (Work) #: +1 (337) 965-1855

EOT
end

def log (message, level = 0)
  if $DEBUG == true or $CONFIG[:log_file] != ""
    if message == ""
      puts "log was called without providing a message"
      exit
    end

    case level
      when 0
        level="INFO"
      when 1
        level="WARN"
      when 2
        level="CRIT"
      else
        level="UNDF"
    end
    datestamp=Time.now
    log_message="#{datestamp} : #{$HOSTNAME} : #{level} : #{caller[0]} : #{message}"

    if $CONFIG[:log_file] != ""
      File.open($CONFIG[:log_file],'a') do |file|
        file.write("#{log_message}\n")
      end
    else
      puts log_message
    end
  end
end

def which(cmd)
  #https://stackoverflow.com/questions/2108727/which-in-ruby-checking-if-program-exists-in-path-from-ruby
  exts = ENV['PATHEXT'] ? ENV['PATHEXT'].split(';') : ['']
  ENV['PATH'].split(File::PATH_SEPARATOR).each do |path|
    exts.each do |ext|
      exe = File.join(path, "#{cmd}#{ext}")
      return exe if File.executable?(exe) && !File.directory?(exe)
    end
  end
  nil
end

def get_verification (prompt = "Do you want to continue?")
  while true
    puts "#{prompt} (y/n): "
    answer = STDIN.gets
    answer = answer.strip()
    case answer
      when 'y','yes'
        return true
      when 'n', 'no'
        return false
      else
        puts "#{answer} is not an acceptible answer"
    end
  end
end

def get_selections (reference_array, prompt = "Which of the following do you want to select? ", options = { :multiple => false, }, &procedure)
  ## Making clone of array since the selections were passed by reference
  selections = reference_array.clone

  def put_prompt (selections, prompt)
    puts prompt
    selections.each_with_index do |value,index|
      puts "#{index} : #{value}"
    end
  end

  if options[:multiple] == true
    output = Array.new
  else
    output = ""
  end

  put_prompt selections, prompt

  while true
    choice = Integer(STDIN.gets.strip())
    if choice.is_a? Integer
     if  choice >= 0 and choice < selections.count 
       if options[:multiple] == true
         output.push(selections[choice])
         selections.delete_at(choice)
         if get_verification "Are you done selecting?" 
           break
         end
         put_prompt selections, prompt
       else
         output = selections[choice]
         break
       end
     else
       puts "#{choice} is not between the values of 0 and #{selections.count}. Please try again."
     end
    else
      puts "#{choice} is not a valid option. Please try again."
    end
  end
  if procedure.respond_to? "call"
    output = procedure.call(output)
  end
  output
end


# END: Helper Functions



# BEGIN: Variables

if not (ARGV & ['-h', '--help']).empty?
  usage
  exit
end

if ENV["USER"] != 'root'
  puts "This script should only be run by root (permissions issues). Please rerun it as root or prepend \"sudo\""
  exit
end


$DEBUG=false

puppet_exe = which "puppet"

if puppet_exe == nil
  puts "The puppet utility was not found in $PATH. This utility will not be able to function"
  exit
end

$CONFIG = Hash.new
$CONFIG[:bucket_dir]=` #{puppet_exe} agent --configprint clientbucketdir `.strip()
$CONFIG[:puppet_version]=` #{puppet_exe} --version `
$CONFIG[:action]=""
$CONFIG[:search_term]=""
$CONFIG[:log_file]=""
$CONFIG[:alt_filepath]=""
$CONFIG[:info_format]="inline"
$CONFIG[:from_time]=Time.at(0) # Beginning of epoch time
$CONFIG[:to_time]=Time.now # Today
File.open('/etc/hostname') do |file|
  $HOSTNAME=file.read().strip()
end
FLAG_REGEX=/^\-+\S+/
DATE_REGEX=/^(?<year>[0-9]{4})-(?<month>[0-9]{1,2})-(?<day>[0-9]{1,2})[[:space:]]*((?<hour>[0-9]{1,2}):(?<minute>[0-9]{1,2}):(?<second>[0-9]{1,2}))?$/

# END: Variables



# BEGIN: Handle CLI Args


if ARGV.count == 0
  puts "No arguments were provided"
  usage
  exit
end

if not (ARGV & ['-d', '--debug']).empty?
  $DEBUG=true
end
i=0
case ARGV[i] 
  when 'search', 'get', 'restore', 'backup'
    $CONFIG[:action]=ARGV[i]
    log "$CONFIG[:action] was set to #{ARGV[i]}"
    log "user provided search action ARGV[i.next] == #{ARGV[i.next]}"
    if ARGV[i.next] != "" and not ARGV[i.next] =~ FLAG_REGEX
      $CONFIG[:search_term]=ARGV[i.next]
      log "search_term was set to #{ARGV[i.next]}"
      i+=2
    else
      puts "Flag[#{ARGV[i]}] : Argument[#{ARGV[i.next]}] : Either the argument was not provided or it was a flag"
      usage
      exit
    end

  when 'list', 'list-files'
    $CONFIG[:action] = ARGV[i]
    log "$CONFIG[:action] was set to #{ARGV[i]}"
    i+=1

  else
    puts "#{ARGV[i]} is not a valid action. Please make sure you use a valid action as the first argument of the script"
    usage
    exit
end

while i < ARGV.count
  case ARGV[i]
    when '-c', '--csv'
      $CONFIG[:info_format]='csv'
      log "$CONFIG[:info_format] was set to #{$CONFIG[:info_format]}"
      i+=1
    when '-l', '--long'
      $CONFIG[:info_format]='long'
      log "$CONFIG[:info_format] was set to #{$CONFIG[:info_format]}"
      i+=1
    when '-i', '--inline'
      $CONFIG[:info_format]='inline'
      log "$CONFIG[:info_format] was set to #{$CONFIG[:info_format]}"
      i+=1
    when '-o', '--output-file'
      log "user provided ARGV[i.next] == #{ARGV[i.next]}"
      if ARGV[i.next] != "" and not ARGV[i.next] =~ FLAG_REGEX
        $CONFIG[:alt_filepath]=ARGV[i.next]
        log "search_term was set to #{ARGV[i.next]}"
        i+=2
      else
        puts "Flag[#{ARGV[i]}] : Argument[#{ARGV[i.next]}] : Either the argument was not provided or it was a flag"
        usage
        exit
      end
    when '--to-date', '--from-date'
      if ARGV[i.next] != "" and ARGV[i.next] =~ DATE_REGEX
        date_matches = DATE_REGEX.match(ARGV[i.next]).named_captures.each_value.select{|val| not val.nil?}.map(&:to_i)
        case ARGV[i]
          when '--to-date'
            $CONFIG[:to_time]=Time.new(*date_matches)
            log "$CONFIG[:to_time] was set to #{$CONFIG[:to_time]}}"
          when '--from-date'
            $CONFIG[:from_time]=Time.new(*date_matches)
            log "$CONFIG[:from_time] was set to #{$CONFIG[:from_time]}}"
        end
        i+=2
      else
        puts "Flag[#{ARGV[i]}] : Argument[#{ARGV[i.next]}] : Either the argument was not provided or it was a flag"
        usage
        exit
      end

    when FLAG_REGEX
      # Catch all to prevent user from specifying a non-accounted for flag
      puts "#{ARGV[i]} is not a valid flag."
      usage
      exit

    else
      i+=1
  end
end

## BEGIN: Checks

if $CONFIG[:action] == ""
  puts "Action was not provided"
end

case $CONFIG[:action]
  when 'search', 'get'
    if $CONFIG[:search_term] == ""
      puts "Search Term was not provided"
      usage
      exit
    end

  when 'backup'
    if not File.exist? $CONFIG[:search_term]
      puts "File #{$CONFIG[:search_term]} does not exist. Please check to make sure that there are not typos and attempt the run again."
      exit
    end
end

## END: Checks

# END: Handle CLI Args


# BEGIN: Classes

class BucketEntry
  attr_reader :hash, :filepaths, :mtime
  def initialize (entry_dir)
    @entry_dir = entry_dir
    @hash = File.basename(entry_dir)
    File.open("#{entry_dir}/paths") do |file|
      @filepaths = file.read.split(/\n/)
    end
    @mtime = File.mtime(entry_dir)
    log "BucketEntry was created from #{entry_dir}"
  end

  def path_include? (path_string)
    log "BucketEntry[#{hash}] was called with #{path_string}"
    @filepaths.any?{|path| path.include? path_string}
  end

  def long_info
    "Entry [#{@hash}]:
      Paths: #{@filepaths.join(',')}
      MTIME: #{@mtime}

"
  end

  def csv_info
    [@mtime,@hash,@filepaths.join(';')].join(',')
  end

  def inline_info
    "#{@mtime} : #{@hash} : #{@filepaths.join(',')}"
  end

  def info
    case $CONFIG[:info_format]
      when 'long'
        long_info
      when 'inline'
        inline_info
      when 'csv'
        csv_info
    end
  end

  def content 
    log "BucketEntry[#{@hash}] getting contents"
    File.open("#{@entry_dir}/contents",'r') do |file|
      file.read()
    end
  end

end

class Bucket
  attr_reader :bucketdir, :entries
  def initialize (clientbucketdir)
    log "Bucket is being created from #{clientbucketdir}"
    @bucketdir = clientbucketdir
    @entries = Hash.new
    load_bucket
  end

  def select(&proc)
    @entries.each_value.select &proc
  end

  def any?(&proc)
    @entries.each_value.any? &proc
  end
    

  def load_bucket
    log "Bucket[#{@bucketdir}] is loading entries"
    Dir["#{@bucketdir}/**/paths"].each.map{|path| File.dirname(path)}.each do |directory|
      log "\"#{directory}\" was grabbed from bucket directory. Making new BucketEntry"
      entry = BucketEntry.new(directory)
      if entry.mtime <= $CONFIG[:to_time] and entry.mtime >= $CONFIG[:from_time]
        @entries[entry.hash]=entry
        log "BucketEntry[#{entry.hash}] was added to @entries Size=#{@entries.count()}"
      else
        log "Entry[#{entry.hash}] was filtered out by the user provided time constraints"
      end
    end
    log "Bucket[#{@bucketdir}] was loaded"
  end

  def filenames
    filenames = Array.new
    @entries.each_value do |entry|
      entry.filepaths.each do |path|
        if not filenames.include? path
          filenames.push(path)
        end
      end
    end
    filenames
  end
end

# END: Classes



# BEGIN: Work Functions

def search_entries_paths (bucket)
  puts bucket.select{|entry| entry.path_include? $CONFIG[:search_term]}.sort_by(&:mtime).map(&:info)
end

def get_content_of_entry_hash (bucket)
  if bucket.entries.has_key? $CONFIG[:search_term]
    puts bucket.entries[$CONFIG[:search_term]].content
  else
    puts "There were no entries corresponding to #{$CONFIG[:search_term]}"
    exit
  end
end

def list_all_entries (bucket)
  puts bucket.filenames.map{|filename| bucket.select{|entry| entry.path_include? filename}.sort_by(&:mtime).map(&:info)}
end

def list_entry_files (bucket)
  puts bucket.filenames.sort.join("\n")
end

def get_entry_by_file (bucket, filenames)
  entry = nil
  if filenames.count == 1
    filename = filenames[0]
  else
    filename = get_selections filenames, "Your filename matched multiple files. Please select one to restore"
  end

  entries = bucket.select{|entry| entry.path_include? filename}
  if entries.count == 1
    entry = entries.first
  else
    while true 
      mtimes = entries.map{|entry| entry.mtime}
      entry_mtime = get_selections(mtimes , "Which timestamp to you want to revert the file to?") 
      entry = entries.lazy.select{|entry| entry.mtime == entry_mtime}.first

      if get_verification "Do you want to see the contents of #{filename} at this time?"
        puts entry.content
        if get_verification "Is this the entry you want to overwrite #{filename} with?"
          break
        end
      else
        break
      end
    end
  end

  if filename[0] != '/'
    filename = "/#{filename}"
  end

  return entry, filename
end

def get_entry_by_hash (bucket)
  if bucket.entries.has_key? $CONFIG[:search_term]
    entry = bucket.entries[$CONFIG[:search_term]]
    filepath = ""
    if entry.filepaths.count == 1
      filepath = entry.filepaths[0]
    else 
      filepath = get_selections entry.filepaths, "What filepath do you wish to restore to?"
    end

    if filepath[0] != '/'
      filepath = "/#{filepath}"
    end

    return entry, filepath

  else
    puts "There were no entries corresponding to #{$CONFIG[:search_term]}"
  end
end

def restore_entry (bucket)
  entry = nil
  if bucket.any?{ |entry| entry.path_include? $CONFIG[:search_term] }
    filenames = bucket.filenames.select {|filename| filename.include? $CONFIG[:search_term]}
    entry, filepath = get_entry_by_file bucket, filenames
  else
    entry, filepath = get_entry_by_hash bucket
  end

  if $CONFIG[:alt_filepath] != ""
    filepath=$CONFIG[:alt_filepath]
  end

  if get_verification "Are you sure you want to overwrite #{filepath}?" 
    File.open(filepath,'w') do |file|
      file.write(entry.content)
    end
  else
    puts "Ok not overwriting."
  end
end

def backup_file (bucket)
  hash=""
  if  Gem::Version.new($CONFIG[:puppet_version]) >= Gem::Version.new("7.0.0")
      log "Puppet Version is +7.* so hashing algo will be SHA256"
      File.open($CONFIG[:search_term],'r') do |file|
        hash = Digest::SHA2.hexdigest file.read()
      end

  else
      log "Puppet Version is not 7.* so hashing algo will be MD5"
      File.open($CONFIG[:search_term],'r') do |file|
        hash = Digest::MD5.hexdigest file.read()
      end
  end

  log "Hash for #{$CONFIG[:search_term]} was generated to be #{hash}"

  if bucket.entries.has_key? hash
    log "Hash was found to already be backed up to bucket"
    puts "This file (hash: #{hash} has already been backed up in the bucket"
    puts bucket.entries[hash].info
    exit
  end

  puppet_uid=0
  puppet_gid=0
  File.open('/etc/passwd','r') do |file|
    log "Getting puppet UID & GID from /etc/passwd"
    passd = file.read.split(/\n/).select{|line| line =~ /puppet.+/}.first
    puppet_uid=Integer(passd.split(':')[2])
    log "Retrieved Puppet UID as #{puppet_uid}"
    puppet_gid=Integer(passd.split(':')[3])
    log "Retrieved Puppet GID as #{puppet_gid}"
  end

  log "Created preceeding directorys to hash directory"
  dir=$CONFIG[:bucket_dir]
  hash.chars[0,8].each{|char|
    dir+="/#{char}"
    log "Checking to make sure that #{dir} doesn't exist"
    if not Dir.exist? dir
      log "#{dir} didn't exist so creating the directory & changing the UID to #{puppet_uid} & GID #{puppet_gid}"
      Dir.mkdir dir
      File.chown(puppet_gid, puppet_uid, dir)
    end
  }

  entry_dir="#{dir}/#{hash}"
  if not Dir.exist? entry_dir
    log "#{entry_dir} didn't exist so creating the directory & changing the UID to #{puppet_uid} & GID #{puppet_gid}"
    Dir.mkdir  entry_dir
    File.chown(puppet_gid, puppet_uid, entry_dir)
  end

  contents_path="#{entry_dir}/contents"
  log "#{contents_path} will be created"
  paths_path="#{entry_dir}/paths"
  log "#{paths_path} will be created"

  log "Creating #{contents_path}"
  File.open(contents_path, 'w') do |contents_file|
    log "Opened #{contents_path} to be written to"
    File.open($CONFIG[:search_term], 'r') do |source_file|
      log "Opened #{$CONFIG[:search_term]} to be read from"
      contents_file.write(source_file.read)
      contents_file.chown(puppet_gid, puppet_uid)
      log "#{contents_path} was created & was chowned to UID to #{puppet_uid} & GID #{puppet_gid}"
    end
  end

  log "Creating #{paths_path}"
  File.open(paths_path, 'w') do |paths_file|
    log "Opened #{paths_path} to be written to"
    paths_file.write($CONFIG[:search_term])
    log "Just wrote #{$CONFIG[:search_term]} to #{paths_path}"
    paths_file.chown(puppet_gid, puppet_uid)
    log "#{paths_path} was created & was chowned to UID to #{puppet_uid} & GID #{puppet_gid}"
  end

  puts "File #{$CONFIG[:search_term]} was backed up to #{entry_dir}"
end

# END: Work Functions


# BEGIN: Work

if __FILE__ == $0
  bucket = Bucket.new($CONFIG[:bucket_dir])

  case $CONFIG[:action]
    when 'search'
      search_entries_paths bucket

    when 'get'
      get_content_of_entry_hash bucket

    when 'list'
      list_all_entries bucket

    when 'list-files'
      list_entry_files bucket

    when 'restore'
      restore_entry bucket

    when 'backup'
      backup_file bucket
  end

end
# END: Work