I often have directory structures with many files of the same type. The first bytes check performs a very limited decimation of the candidates because the first bytes of files usually contain preambles, which are often the same for files of the same type.
In this case the last bytes check is much more efficient to decimate the candidate list.
So, if I could swap those two checks with a command line parameter (so that the last bytes check is done first, followed by the first bytes check) I could get a substantial improvement in scan times.
Somewhat related to #29.
I often have directory structures with many files of the same type. The first bytes check performs a very limited decimation of the candidates because the first bytes of files usually contain preambles, which are often the same for files of the same type. In this case the last bytes check is much more efficient to decimate the candidate list. So, if I could swap those two checks with a command line parameter (so that the last bytes check is done first, followed by the first bytes check) I could get a substantial improvement in scan times.