* As we're now not sending *all* journal events on the journal schema
this refactors only the actual sending in export_journal_entry(), with
the other functionality in export_journal_generic(), so that the new
export_journal_fssdiscoveryscan() can use export_journal_entry() as
well.
In the future EDDN will move to a schema per event type, so we'll lose
_generic(). Factoring out some things like augmentations into their own
functions will come next.
This also entailed slightly reworking the way the EDDN code uses this
URL. It was very generalised, so as to allow for the debug "just send
and log locally" code, but as the only URL is the 'upload' one much of
that seemed un-necessary.
So that code has been simplified.
This should have been set from the latest Journal file, and the
key/value is in LoadGame in Odyssey. It's only documented as:
"Whether the sending Cmdr has an Odyssey expansion."
in the current EDDN Journal schema. Thus sending this as False if the
Cmdr last logged into Horizons/base game should be OK.
* Ensure we have the discarded_events list *before* starting to consume
events. There's the *same* risk as before with EDSM just being down
and us queueing up more and more, just now it will be in the worker
queue rather than pending.
* We need to test for some events in should_send() that are in the
discard list, and thus will never be in the `entries` (pending) passwd
in, so pass in the name of the latest event as well.
* trace-on 'edsm-cmdr-events' for how Commanders, their API keys and
using them are handled. See #1134
* trace-on 'edsm-locations' for what was commented-out logging to do
with ensuring code reacted correctly to any change of system.
* mypy/types-* seem to have had changes causing extra carping, so
addressing that.
A simple HTTP handler has been added that will dump any and all POST
data it gets to EDMC's log.
Additionally, a new command line argument to switch EDMC to using this
as its EDDN target has been added.
Athanasius
A_D
norohind