Me neither... but it sounds like you have many rows of data that are (sometimes) referencing the same primary key. Is this correct:
Item A
Item B
Item C
Item B
Item D
Item A
Item A
... after the importation of which you'd want a table that had:
Item A: 3
Item B: 2
Item C: 1
Item D: 1
... is that correct? If so, then yes, you want a single table with the primary key being the item you require to be unique and a "tally" column that keeps the number of occurrences.
But it sounds to me like if you could post an example of some data we'd all do better for you here...
Perk, you hit the nail on its head. A good example, in theory, would be WordTracker. If you search
WT for "fart fetish", WT will tell you how many times "fart fetish" occurs in their db as an exact phrase
and also as part of a longer phrase like "JM's best fart fetish videos".
Either they have a tally field or they are counting the raw data everytime.
Of course, counting the occurances in longer phrases will entail counting the raw data everytime.
I like the idea of counting the raw data, but
if 50% of the daily imported data is unique,
the db is going to grow like Jack's beanstalk on Viagra.
I can't give you a sample of my data cuz then I'd have to kill you. Most of you could
build and implement this before I finish my coffee.
But you've all given me food for thought here, thanks.
Bompa
