Handle LiDAR data gaps explicitly instead of failing silently

[bloom256]

Description

Some datasets contain short gaps in LiDAR data because of hardware or firmware issues - for example, no points for about 100 ms.

Right now this can lead to odometry failure, but HDMapping does not report the real cause. From the user side, it just looks like odometry broke for no clear reason.

Expected behavior

If the input data contains a significant gap, HDMapping should detect it and report it explicitly - preferably as an error, or at least a strong warning.

Why

• Makes the root cause clear
• Saves debugging time
• Prevents confusing silent failures

Possible approach

• Check time gaps in input data
• Use a threshold for abnormal gaps
• If exceeded, report the gap size and time in the message
• Optionally stop or split session