Google presentation open redirect phishing

type.inbound
and not strings.icontains(body.current_thread.text, 'invited you to edit')
and any(body.links,
        // body link is to a google doc presentation
        .href_url.domain.domain == "docs.google.com"
        and strings.istarts_with(.href_url.path, '/presentation/')

        // prefilter some to avoid clicking on _every_ google presentation link
        and ( 
          ( 
            // make sure the display text is in the current thread and not a previous one.
            strings.icontains(body.current_thread.text, .display_text)
            // the display_text ends with a word that is 4-10 long
            and regex.icontains(.display_text, '[[:punct:]\s][a-z0-9]{5,9}$')
            // that word has to include a letter AND a number
            and regex.icontains(.display_text,
                                '[[:punct:]\s](?:[a-z0-9]*[a-z][0-9][a-z0-9]*|[a-z0-9]*[0-9][a-z][a-z0-9]*)$'
            )
            and strings.iends_with(.href_url.path, '/pub')
          )
          or 
          // finally send the link to link analysis that presentation...
          (
            // contains a single link
            length(ml.link_analysis(., mode="aggressive").final_dom.links) == 1

            // cannot be edited via link provided
            and strings.contains(ml.link_analysis(., mode="aggressive").final_dom.raw,
                                 'canEdit:  false'
            )

            // and a single page
            and strings.contains(ml.link_analysis(., mode="aggressive").final_dom.raw,
                                 'slidePageCount:  1.0'
            )

            // where we have links which have been written via a google open redirect
            and any(ml.link_analysis(., mode="aggressive").final_dom.links,
                    // links are not in thhe org_domains
                    .href_url.domain.domain not in $org_domains
                    and (
                      (
                        // don't include high rep domains
                        .href_url.domain.domain not in $tranco_1m
                        and .href_url.domain.domain not in $umbrella_1m
                      )
                      // if it's in Tranco or Umbrella, still include it if it's one of these
                      or .href_url.domain.domain in $free_file_hosts
                      or .href_url.domain.root_domain in $free_file_hosts
                      or .href_url.domain.root_domain in $free_subdomain_hosts
                      // or it's a url shortner
                      or .href_url.domain.root_domain in $url_shorteners
                      or .href_url.domain.root_domain in $social_landing_hosts
                    )
                    // which have been "unrolled" by the google_open_redirect rule
                    and any(.href_url.rewrite.encoders,
                            . == "google_open_redirect"
                    )
            )
          )
          // or the presentation has been removed for violation of terms of service
          or strings.icontains(ml.link_analysis(., mode="aggressive").final_dom.display_text,
                               "We're sorry. You can't access this item because it is in violation of our Terms of Service."
          )
        )
)
// when the sender is not google, the sender should not be a common prevalence
and (
  ( // the message is not from google actual
    sender.email.email not in (
      'comments-noreply@docs.google.com',
      'drive-shares-dm-noreply@google.com',
      'drive-shares-noreply@google.com',
      'calendar-notification@google.com'
    )
    // ensure the sender prevalence is not common
    and profile.by_sender().prevalence != "common"
  )
  // or the message is from google actual
  or (
    sender.email.email in (
      'comments-noreply@docs.google.com',
      'drive-shares-dm-noreply@google.com',
      'drive-shares-noreply@google.com',
      'calendar-notification@google.com'
    )
  )
)

// not where the sender display name of the message is within org_display_names
and not (
  // the message is from google actual
  sender.email.email in (
    'comments-noreply@docs.google.com',
    'drive-shares-dm-noreply@google.com',
    'drive-shares-noreply@google.com',
    'calendar-notification@google.com'
  )
  and headers.auth_summary.dmarc.pass

  // but the sender display name is within org_display_names
  and any($org_display_names,
          strings.istarts_with(sender.display_name,
                               strings.concat(., " (via Google ")
          )
          or strings.istarts_with(sender.display_name,
                                  strings.concat(., " (Google ")
          )
  )
)
// negate highly trusted sender domains unless they fail DMARC authentication
// but ignore high_trust if the sender is one of the google actual senders
and (
  (
    (
      sender.email.domain.root_domain in $high_trust_sender_root_domains
      and not headers.auth_summary.dmarc.pass
    )
    or (
      sender.email.domain.root_domain in $high_trust_sender_root_domains
      and sender.email.email in (
        'comments-noreply@docs.google.com',
        'drive-shares-dm-noreply@google.com',
        'drive-shares-noreply@google.com',
        'calendar-notification@google.com'
      )
    )
    or sender.email.domain.root_domain not in $high_trust_sender_root_domains
  )
)