Linux lhjmq-records 5.15.0-118-generic #128-Ubuntu SMP Fri Jul 5 09:28:59 UTC 2024 x86_64
Your IP : 3.15.195.84
"""Helpers for reliable persistent message queues."""
ANCIENT = 1
def got_next_expected(store, next_expected):
"""Our peer has told us what it expects our next message's sequence to be.
Call this with the message store and sequence number that the peer
wants next; this will do various things based on what *this* side
has in its outbound queue store.
1. The peer expects a sequence greater than what we last
sent. This is the common case and generally it should be
expecting last_sent_sequence+len(messages_sent)+1.
2. The peer expects a sequence number our side has already sent,
and we no longer have that message. In this case, just send
*all* messages we have, including the previous generation,
starting at the sequence number the peer expects (meaning that
messages have probably been lost).
3. The peer expects a sequence number we already sent, and we
still have that message cached. In this case, we send starting
from that message.
If the next expected sequence from the server refers to a message
older than we have, then L{ANCIENT} will be returned.
"""
ret = None
old_sequence = store.get_sequence()
if next_expected > old_sequence:
store.delete_old_messages()
pending_offset = next_expected - old_sequence
elif next_expected < (old_sequence - store.get_pending_offset()):
# "Ancient": The other side wants messages we don't have,
# so let's just reset our counter to what it expects.
pending_offset = 0
ret = ANCIENT
else:
# No messages transferred, or
# "Old": We'll try to send these old messages that the
# other side still wants.
pending_offset = (store.get_pending_offset() +
next_expected - old_sequence)
store.set_pending_offset(pending_offset)
store.set_sequence(next_expected)
return ret
|