A PGP armor message can contain any amount of headers. Up to Fedora 38
there was one:

  -----BEGIN PGP PUBLIC KEY BLOCK-----
  Version: rpm-4.18.0-beta1

  mQINBGIC2cYBEADJye1aE0AR17qwj6wsHWlCQlcihmqkL8s4gbOk1IevBbH4iXJx
  [...]
  =CHKS
  -----END PGP PUBLIC KEY BLOCK-----

Since Fedora 39 there is none:

  -----BEGIN PGP PUBLIC KEY BLOCK-----

  mQINBGLykg8BEADURjKtgQpQNoluifXia+U3FuqGCTQ1w7iTqx1UvNhLX6tb9Qjy
  l/vjl1iXxucrd2JBnrT/21BdtaABhu2hPy7bpcGEkG8MDinAMZBzcyzHcS/JiGHZ
  [...]
  =CHKS
  -----END PGP PUBLIC KEY BLOCK-----

RpmImportedKeys._query_db_for_gpg_keys() assumed exactly one header.
As a result if gpgkey_dns_verification configuration option was true,
DNF reported that Fedora 39 keys was revoked because the key
misextratracted from RPM database did not match a key in DNS:

    # dnf-3 upgrade
    DNSSEC extension: Testing already imported keys for their validity.
    DNSSEC extension: GPG Key [email protected] has been revoked and should be removed immediately

This patch implements skipping all armor headers.

https://bugzilla.redhat.com/show_bug.cgi?id=2249380

If an PGP key is stored in an RPM database without a "packager" RPM
header, or without an e-mail address there, DNS verification crashed
on converting the undefined address into a DNS domain. That was the
case of Fedora 13 key:

    # dnf-3 upgrade
    Traceback (most recent call last):
      File "/usr/bin/dnf-3", line 62, in <module>
        main.user_main(sys.argv[1:], exit_code=True)
      File "/usr/lib/python3.12/site-packages/dnf/cli/main.py", line 201, in user_main
        errcode = main(args)
                  ^^^^^^^^^^
      File "/usr/lib/python3.12/site-packages/dnf/cli/main.py", line 67, in main
        return _main(base, args, cli_class, option_parser_class)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
      File "/usr/lib/python3.12/site-packages/dnf/cli/main.py", line 106, in _main
        return cli_run(cli, base)
               ^^^^^^^^^^^^^^^^^^
      File "/usr/lib/python3.12/site-packages/dnf/cli/main.py", line 122, in cli_run
        cli.run()
      File "/usr/lib/python3.12/site-packages/dnf/cli/cli.py", line 1040, in run
        self._process_demands()
      File "/usr/lib/python3.12/site-packages/dnf/cli/cli.py", line 741, in _process_demands
        self.base.fill_sack(
      File "/usr/lib/python3.12/site-packages/dnf/base.py", line 403, in fill_sack
        dnf.dnssec.RpmImportedKeys.check_imported_keys_validity()
      File "/usr/lib/python3.12/site-packages/dnf/dnssec.py", line 286, in check_imported_keys_validity
        keys = RpmImportedKeys._query_db_for_gpg_keys()
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
      File "/usr/lib/python3.12/site-packages/dnf/dnssec.py", line 276, in _query_db_for_gpg_keys
        email = re.search('<(.*@.*)>', packager).group(1)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
      File "/usr/lib64/python3.12/re/__init__.py", line 177, in search
        return _compile(pattern, flags).search(string)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    TypeError: expected string or bytes-like object, got 'NoneType'

This patch defends the crash at two places: In
_query_db_for_gpg_keys() because here we know a NEVRA of the key
and can produce a meaningful message. And in _cache_miss() because
we can get there independenly and called email2location() would also
crash.

If a user had installed multiple keys for the same e-mail address in
an RPM database, and no records for the address existed in DNS, DNF
validated the first key correctly, but reported that the other key is
revoked:

    # rpm -q gpg-pubkey --qf '%{packager} %{nevra}\n' |grep nokey
    nokey1 <[email protected]> gpg-pubkey-7460757e-6553a6ab
    nokey2 <[email protected]> gpg-pubkey-c8d04ba8-6553a6b1
    # dnf-3 upgrade
    DNSSEC extension: Testing already imported keys for their validity.
    DNSSEC extension: GPG Key [email protected] has been revoked and should be removed immediately

The cause was a wrong test for a cached negative reponse. This patch
fixes it.

https://bugzilla.redhat.com/show_bug.cgi?id=2249380

Added option 'send_error_messages'

Fixed Option String List

Changed option to Boolean