Comments by CDRxxx (Top 4 by date)

Well, the question really is:
On a daily basis a textfile is generated with about a million strings (C:\Temp\NGDDIVlist.txt). Each string is in fact a record with two fields. The project needs to search this textfile for matches on field 0 of each record and has to do that for more or less 400.000 times. For the time being this is done using Instr, but I am looking for faster ways to do this. If I can split up the initial textfile in less longer subfiles, the Instr will take less time. Therefore these subfiles need to be named with the 'key' used for splitting. For this key I want to use the last two characters of field 0 of each record in the initial textfile. To build those subfiles fast, I want to use stringbuilders and since the 'key' = 2 characters, there are 36 x 36 = cases...