Bug fix of indexing for the TGC Big-Wheel coincidence LUT of ascii DB files

Fixed a bug of indexing for the TGC Big-Wheel coincidence Look-Up-Tables realized by large multi-dimensional maps.

The start- and end-loop indexes are wrong in some cases, the referred pointer was also not correct. The bug was not identified by running the code, due to several unlucky cases. For example, an undefined index just dropped the candidate, or change to other thresholds. The LUT-size optimization, as same as the Run-2 Coincidence Window has been also implemented at this time as was planned on my to-do list.

This fix changes the trigger counts from the L1Muon endcap, not only MU3, but also other thresholds. The LUT optimization of contents itself will be implemented in the future. Anyway, new validation is needed using enough statistics samples.

The private test has been done for both MC and data. test_trig_data_v1Dev_L1SimOnly_build.py has been also passed and the counts in L1AV.txt of L1_MU3VF is reasonable.

Tagging the Trigger Menu coordinators (@dzanzi and @khoo) and other L1Muon endcap developers (@masato and @tnobe)