Make kadi events complete from start of mission

[taldcroft]

Description

This makes event processing robust to getting no applicable telemetry or events, thus allowing the start of the event database build process to be set to an early date like 1999:001.

The events3.db3 database was rebuilt from scratch on kady using the updated NOTES.build instructions.

• Confirmed that key events like Obsid and Manvr showed the expected new events coinciding with the beginning of available CXC L0 engineering telemetry on 1999:204.
• No errors and only expected processing warnings.

With the merging of this the new events file will be copied to flight.

Testing

• Passes unit tests on MacOS (using both current flight and the new events database)
• Functional testing

Functional testing

Used the validate/write_events_cmds.py script to write ECSV text files for each event type corresponding to two date ranges:

• 1999:240 - 2021:070 (expected to match)
• 1999:001 - 1999:240 (expected to have new entries)

ska3-kady$ diff -r -q flight new  # 1999:240 to 2021:070
Files flight/orbits.ecsv and new/orbits.ecsv differ   # OK: some few seconds diffs
Files flight/pass_plans.ecsv and new/pass_plans.ecsv differ  # OK, few changes in BOT/EOT like 0345 -> 345
Files flight/rad_zones.ecsv and new/rad_zones.ecsv differ # OK, few seconds diffs
Files flight/safe_suns.ecsv and new/safe_suns.ecsv differ  # OK, float format issues (e.g. 1.0000004 -> 0.9999996)
Files flight/scs107s.ecsv and new/scs107s.ecsv differ  # OK, float format issues

Inspection of the files for the 1999:001 - 1999:240 exports show the expected new values