Fix #1325 by handling the exception

[reidsunderland]

Current behaviour stays the same:

• if putNewMessage returns False, failed message and all remaining messages* in the worklist are put into worklist.failed

New behaviour:

• if putNewMessage crashes the exception is handled. The current message and all remaining messages* are put into worklist.failed

* the remaining messages will not attempt to be posted

Any messages that were successfully posted will be put into worklist.ok

[github-actions]

Test Results

238 tests   235 ✅  1m 31s ⏱️
  1 suites    1 💤
  1 files      2 ❌

For more details on these failures, see this check.

Results for commit b6011b5.

[reidsunderland]

If I add an obvious error to putNewMessage to force it to crash, this is the result:

2024-12-09 16:37:59,835 [ERROR] sarracenia.flowcb.post.message post failed: name 'y' is not defined
2024-12-09 16:37:59,835 [DEBUG] sarracenia.flowcb.post.message post Exception details:
Traceback (most recent call last):
  File "/local/home/sarra/sr3/sarracenia/flowcb/post/message.py", line 41, in post
    if all_good and hasattr(self.poster,'putNewMessage') and self.poster.putNewMessage(m):
  File "/local/home/sarra/sr3/sarracenia/moth/amqp.py", line 641, in putNewMessage
    x = y
NameError: name 'y' is not defined

The log plugin does not incorrectly state that the message was posted.

And the messages are going into the post retry disk queue:

2024-12-09 16:37:59,996 [DEBUG] sarracenia.flowcb.retry after_work loading from post_retry_014: qty=12 ... got: 0
...
2024-12-09 16:38:59,502 [DEBUG] sarracenia.flowcb.retry on_housekeeping on_housekeeping
2024-12-09 16:38:59,600 [DEBUG] sarracenia.flowcb.retry after_work loading from post_retry_014: qty=12 ... got: 12
2024-12-09 16:38:59,604 [DEBUG] sarracenia.flowcb.retry on_housekeeping on_housekeeping

** I changed "failed:" to "crashed:" after this test, because we usually use the word crashed when a method threw an exception. Failed makes more sense in the case where putNewMessage returns False.