contest: reset the qso when the form becomes empty Users assume that backspacing the form will reset the start time. Let's not surprise them. Note that if the form has whitespace, then it is not considered empty. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
xlua: grid helper script shouldn't use bad QSOs for points & mults They should, however, be included in all QSOs list. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
watch: move the stdout locking to the top level This reduces the concurrency slightly, but it guarantees that all output for a single packet is contiguous. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
watch: dump info announcement CBOR payload Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
watch: improve unknown announcement warning text Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
watch: move hexdumping into a separate function Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
watch: print a warning about received address being longer than xsockaddr If we get one of these addresses, we print a warning and then NULL-out the address pointer. The printing function displays a "???" instead of trying to use a NULL pointer. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
watch: print source address for each announcement Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
watch: print timestamp for non-announcement packets Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
watch: move timestamp printing to a helper function This will make it easier to also print the source address. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
rpc: make sure 'all' notification callback is always invoked Deduplication of 'qso' callbacks shouldn't affect the 'all' notification callbacks. We want 'all' to be called for every single announcement packet. The type-specific callbacks can have their own semantics. Currently that means: - qso: deduplicated - loc, rig: called for each - info: no callback Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
rpc: constify announcement message args in tracking code Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
common: add time_set to set a struct time to an arbitrary timestamp Additionally, this makes time_set_now a trivial wrapper. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
xlua: switch CQ-WW-VHF to grid helper Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
xlua: switch ARRL-VHF-{JAN,JUN,SEP} to grid helper
Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
xlua: grid-based contest helper Contest that use Maidenhead grid squares as the exchange are largely the same. They (almost) all have the concept of QSO points and multipliers. Implementing those rules over and over is tedious and error prone. This helper object implements a generic contest with easy-to-use hooks to tweak behavior based on the specific contest's rules. Currently unused. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
contest: try to load all QSOs on start up This invokes the qso-load event for each QSO in the log. Currently, nothing implements the load-qso event, so this turns into a big no-op. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
contest: add qso-load event This event is used to notify the script about a qso entry that was added outside its knowledge. This can be used in the future for loading existing entries on startup or when a new entry appears via wsjtx or announcements. The big difference between qso-load and qso-done events is that qso-load gets a read-only qso structure while qso-done gets a read-write qso struct. Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
contest: split startup event into startup and fill-template The startup event is executed always. The fill-template event is only executed when there is a template to fill (i.e., it is not an export). Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
rpc: print a summary of hamlib quirks Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>