wrong regexp for pubsec authfile

See forum : https://forum.duniter.org/t/duniter-silkaj-duniterpy-scuttlebot-comment-partager-les-clefs-de-chiffrage-des-crypto-applications/6775/36

While Silkaj's regexp for pubsec authfile format is :

    regex_gannonce = compile(
        "^pub: [1-9A-HJ-NP-Za-km-z]{43,44}\nsec: [1-9A-HJ-NP-Za-km-z]{88,90}.*$"
    )

Duniterpy's one is :

        regex_pubkey = re.compile("pub: ([1-9A-HJ-NP-Za-km-z]{43,44})", re.MULTILINE)
        regex_signkey = re.compile("sec: ([1-9A-HJ-NP-Za-km-z]{87,90})", re.MULTILINE)

We can see that :

• for Silkaj, length is {88, 90} and it should be {87, 90}. So, Silkaj refuses valid privkeys.
• Silkaj only matches pubsec files that begin with pub : XXX and are not compatible with the pubsec format generated by Duniterpy.

Proposal :

either keep the newline between pub and sec lines (no ^ at the beginning and length {87, 90}:

    regex_gannonce = compile(
        "pub: [1-9A-HJ-NP-Za-km-z]{43,44}\nsec: [1-9A-HJ-NP-Za-km-z]{87,90}.*$"
    )

or use Duniterpy's regexp as shown above, and change following "if" statement.

Consequences :

Silkaj will accept PubSec format compatible with Cesium and Duniterpy. However, it is not sure that this format is compatible with Duniter's one.

Duniter's Pubsec format :

pub : "<pubkey>"
sec : "<privkey>"

Pubsec V1 :

Type: PubSec
Version: 1
pub: <pubkey>
sec: <privkey>