Duplicate diaObjects from different tracts/patches

[rearmstr]

Bruno noticed that diaObjects from different tracts/patches had essentially the same set of diaSrc lists. Here is one example:

diaObjectId         diSrcIds      
19492442784924347  10702592027918491
19492442784924347  10703469946077392
19492442784924347  27984868264640743
19492442784924347  28172647456047616
19492442784924347  30085989217075588
19492442784924347  30299931201765613
19492442784924347  30372874577510632
19492442784924347  31291195695038695
19492442784924347  31295218334564663
19492442784924347  31298376209268933
19492442784924347  32164143502983335
19492442784924347  26869728593052030
19492442784924347  30302405035819237
19492442784924347  30370993314726164
20407623006290383  10702592027918491
20407623006290383  10703469946077392
20407623006290383  26869728593052030
20407623006290383  27984868264640743
20407623006290383  30085989217075588
20407623006290383  30299931201765613
20407623006290383  30302405035819237
20407623006290383  30370993314726164
20407623006290383  30372874577510632
20407623006290383  31291195695038695
20407623006290383  31295218334564663
20407623006290383  31298376209268933
20407623006290383  32164143502983335
20407623006290383  28172647456047616

[rearmstr]

The current code was only filtering diaSrc objects for association by having them being located in a inner patch region. However, there are overlaps between inner patch regions between different tracts. I have added another filter to only associate diaSrc objects in the inner tract region as well.

This fixes the problems that Bruno reported above.