Add another check for "no price/size" empty values

Also adds some historical debug details for trying to figure out why the
IBKR APIs return bad SPX values after hours (it always does this and
even their official apps/sites report the bad values).

As far as I can tell, there's a bug in the IBKR data APIs where, after
hours, it returns the _correct_ value followed immediately by some
incorrect value (maybe it's adjusted for something?) but we can't seem
to capture the "good value" before the bad value replaces it because
there's no indicator which value is which (other than the order they
arrive in immediately when connecting).

Author: mattsta

ib_async/wrapper.py

@@ -925,14 +925,29 @@ def priceSizeTick(self, reqId: int, tickType: int, price: float, size: float):
             ticker.askSize = size
         elif tickType in {4, 68}:
             # for 'last' values, price can be valid with size=0 for updates like 'last SPX price' since SPX doesn't trade
-            if price == -1 and size == 0:
+            # Workaround: for TICK-NYSE, it is valid to have price=-1, size=0 because it can float between -10,000 and 10,000
+            #             and it also never reports a size. As a workaround, check if ticker.close exists as a proxy for "not TICK-NYSE"
+            #             because TICK-NYSE never has open/close values populated.
+            if price == -1 and size == 0 and ticker.close > 0:
                 price = self.defaultEmptyPrice
                 size = self.defaultEmptySize
 
+            # BUG? IBKR is sometimes sending a GOOD VALUE followed by a PREVIOUS value all under tickType=4?
+            # e.g. I get the SPX close delivered first with size=0, then I get another data point with size=1 priced one point lower,
+            # but since the older price is delivered second, it replaces the "last" value with a wrong value? Not sure if it's
+            # an IBKR data problem or a logic problem somewhere here?
+            # More research: IBKR also shows the bad value in their own app, so there is a data bug in their own server logic somewhere.
+
+            # self._logger.error(f"[{tickType=}] updating last price size: {price=} {size=} :: BEFORE {ticker=}")
+            # self._logger.error(f"[{tickType=}] SETTING {ticker.prevLast=} = {ticker.last=}; {ticker.prevLastSize=} = {ticker.lastSize=}")
+
             ticker.prevLast = ticker.last
             ticker.prevLastSize = ticker.lastSize
             ticker.last = price
             ticker.lastSize = size
+
+            # self._logger.error(f"[{tickType=}] SET {ticker.prevLast=} = {ticker.last=}; {ticker.prevLastSize=} = {ticker.lastSize=}")
+            # self._logger.error(f"[{tickType=}] updating last price size: {price=} {size=} :: AFTER {ticker=}")
         else:
             assert (
                 tickType in PRICE_TICK_MAP