[Therion] Discrepancy in Cave data (length)

[Olly Betts]

On Tue, Sep 02, 2025 at 12:58:00PM +0100, Andrew Atkinson wrote:
> For the data set
> 
> http://www.cave-registry.org.uk/svn/CheddarCatchment/LongwoodValleySink/
[Survex says:]
>  6> Total length of survey legs = 1004.00m (1003.23m adjusted)
[Therion says:]
> length: 1030.92m (surface 0.00m, duplicated 109.15m)
[...]
> 1030.92 to 1004.00

Does therion implement automatic averaging of consecutive repeated legs?

For example, these three legs would be average to one leg and have a total
length of just over 7m in Survex:

    30  31  7.082  159.22  70.98
    30  31 7.080  159.60  70.87
    30  31  7.104  159.97  70.58

Maybe therion treats this as 3 legs with a total length of just over 21m?

That's from TickleWest.th.  The other case I found which isn't flagged as
"duplicate" is also in that file:

    33  34  6.525  350.23  82.85    #  virtual stn
    33  34  6.295  351.22  82.56  
    33  34  6.279  349.95  82.24  

Just averaging the lengths, if my hypothesis is correct we'd expect to
get a length about 26.91m longer with therion, which is only 1cm away
from what you observed (which could just be from to subtracting two
answers rounded to the nearest cm).

Cheers,
    Olly