Improve the getECSV algorithm

[markmac99]

The current method looks for FF files with matching filenames to the camera id and datestamp eg UK0006, 2024-05-01T20:14:20 would match FF_UK0006_20240501201420*, then obtains the first available matching event from the ftpdetect file. This doesn't work properly for FF files containing multiple detections with different timestamps.

For example

matches/RMSCorrelate/UK005U/UK005U_20240809_201258_058797/FTPdetectinfo_UK005U_20240809_201258_058797.txt

contains three events in FF_UK005U_20240810_004140_343_0401920.fits. The API will only return the first of these. I need to check more carefully, without an unacceptable performance or cost impact.

[markmac99]

fetchECSV(camid, datestr, metcount=1):

• add optional parameter &metcount=n to allow selection of the nth meteor in an FF
• also parse exact date in milliseconds, if needed

[markmac99]

Looks to be done