How should blank values be handled?

Ypar

This is a general question with a specific case as example.

Should blank values in fields that normally should not be blank, be considered an error? Should we just ignore the field? Should it be 0 (or equivalent for the data type)?

In my particular case, I have an unspecified event that would normally represent a trade, except for the presence of a quote price & qty field that are blank. The easy solution would be to ignore fields with blank values in this case, but how general is this solution? Should it be a solution only for unspecified events? Any event? Any field? Any data type?

Stream RBU7-X7 --- RSSL_MC_UPDATE/RDM_UPD_EVENT_TYPE_UNSPECIFIED field values
   TIMACT (5)                18:18:00:000:000:000
   TRDTIM_1 (18)             18:18:00:000:000:000
   ACT_TP_1 (270)             Þ (2)
   SC_ACT_TP1 (280)           + (1)
   TRDPRC_1 (6)              0.1323
   PRIMACT_1 (393)           0.1323
   NETCHNG_1 (11)            0.0062
   SEC_ACT_1 (275)           0.0062
   PRCTCK_1 (14)             Þ (1)
   ACVOL_1 (32)              1935
   PCTCHNG (56)              4.92
   NUM_MOVES (77)            1234
   ASK (25)                  <<blank>>
   ASKSIZE (31)              <<blank>>
   TRDVOL_1 (178)            38
   SALTIM (379)              18:18:32:000:000:000
   SEQNUM (1021)             63235757

Ypar

So what I figured is that blank values can sometimes mean something. I still have not figured out what they would mean in my example above, but in other cases, they tell you indirect information about the data. E.g. For futures, a blank "number of orders" indicates an implied order.

So my conclusion is that blank values should be ignored (as if the tag was not part of the data set) unless otherwise stated and the information they provide is needed.