The following questions will seem painfully simple to some, I know.
Grateful for some basic pointers.

I keep large numbers of texts (around 3,000), each in a separate
record, in an FMP file called "Documents". As well as the "Documents"
table, this file has a table called "Keywords." I use lots of keywords
(over 80) so they need management, sorting and classifying (some
categories: "democracy," "ownership," "climate change," and many more
finely-tuned ones). Each document can have up to 10 keywords, though
on average I would probably use no more than 3 or 4.

To retrieve items as needed, I use a self-lookup relationship defined
between the "Keyword" field and a global called "Lookup_Keyword." The
latter appears in a box which I have associated with a value list. So
if I click on this field and select "democracy" from the drop-down
list, a portal defined on the self-lookup relationship becomes
populated with the titles of the documents which have been given (by
me the user) the keyword "democracy."

Here are my issues:

1) I want to be able to do the same thing with groups of keywords. Say
I had four lookup fields: Lookup_Keyword_1, Lookup_Keyword_2,
Lookup_Keyword_3 and Lookup_Keyword_4. As I select from the drop-down
menu of each of these, I want the portal to show a progressively
constrained list.

Thus, I select "democracy" in Lookup_Keyword_1, "ownership" in
Lookup_Keyword_2, and "climate change" in Lookup_Keyword_3. As I do
so, I want the list of titles in the portal to get whittled down to
just the texts bearing those three keywords.

2) I want the value list for each of the Lookup_Keyword fields to be
itself a constrained sub-set of the 80-odd members of my Keywords (as
stored in the Keywords table). I have categorized the keywords by a
set of themes ("buzzword", "politics", "economics", "environment").
Categories and keywords are themselves in a many-many relationship:
the keyword "climate change" belongs in both the "environment" and
"buzzword" categories, for instance.

I have also set up routines to break the list up alphabetically --
"Keywords a-f", "g-m", "n-t","u-z").

So when I start operating my lookup function I would like to get to
choose one keyword from each of these categories. It would be nice to
have a choice between the thematic breakup and the alphabetic one, but
that can wait. The point is to break up the list into manageable
chunks, so as to choose keywords quickly and easily.

A Dai

"Dai Kaili" <daikaili (AT) gmail (DOT) com> schreef in bericht
news:1184772504.354680.183450 (AT) j4g2000prf (DOT) googlegroups.com...
I haven't tried this but here's my thoughts. Create a chain of linked
relations.

1) LookUP1 is mirrored in the second TO with a calculation. This calc
returns empty when the keyword is not related to the current record. But
returns the complete text when the relation to the keyword is valid.
2) LookUP2 is mirrored in the second TO with a new calculation. This calc
returns empty when the keyword is not related to the current record. But
returns the complete text when the relation LookUP2::keyword is valid.

etc etc.

Since the relations are chained, and are completely inherited every choise
should limit the amount of records returned. You should built all the calcs
so that they also return everything when the LookUP is empty.

Once again, I haven't tried it, but I would start along these lines.

Keep well, Ursus