Non-portable %f scanning into float without width limit; potential precision/UB mismatch not enforced #16
Description
Where: DEFGSupport.cpp, fscanf
Why it’s wrong: %f is correct for float*, but the code later passes these to AddPoint(double,double,double) after negation; the precision truncation may not be intended for elevation data. If %lf was expected for double, this is an input precision bug.
Evidence: Variable types and format string.
GitHub
Fix: If inputs are double precision, change variables to double and use %lf; otherwise document the precision loss.