RE: [PBML] Re: Hashes and Filtering (fwd)
- Just how big is the data file now, and how big *may* it get?
For any given 9 digit code, will it only have two possible records (itself
and the 18 character code), or could there be three or more records?
IF the table can be kept in memory, AND there are at most two records for a
key (itself and/or the 18 character code), I believe the attached will do
what you want, or at least give you a good starting point. (As I was
examining it, I reformatted it to reduce the number of screen lines.)
From: Elliot L. Tobin [mailto:elliot@...]
Sent: Tuesday, February 27, 2001 1:38 PM
Subject: Re: [PBML] Re: Hashes and Filtering (fwd)
In case anyone else can help that'd be great too..
Thanks for all the help. I am only stuck with one last problem. Say my
data file looks like this:
999009907 -> 999009907 #correct
999009910 -> 999009910 # NOT correct
99009904 -> 60305200801541567 # correct
To remind you, if there is a duplicate on the right side, I want to keep
the left side which is 18 characters, not 9. But here, since the 18
character code comes BEFORE the 9 digit code, it keeps the 9 digit code..
If I reverse the order of the lines containing 999009910, it does work.
How should I deal with this if I cannot modify the input file? Thanks