Hi

In my crossing from FileMaker to Postgresql, I have a question that so
far I can't answer from reading the O'Reilly book. BTW: most of the
'calculations' I'm referring to here are in text; I'm building a tool
for a Linguistics application. If I use numbers, they are usually a
code from which to calculate textual variables.

e.g.:

In languages which combine STEMS and a variety of ENDINGS to
differentiate genders, cases, tenses etc ... I could build text
operators in FileMaker to generate these.

1) a couple of fields [columns] might derive from data entry, while the
rest would be filled automatically via a number of 'calculations'
(while, if, case, etc)

2) I might like to split a word into sections (some of which may change
and others not) and then have a final field [column] that reassembles
them, according to a contextual variable.

In the O'Reilly book, I can find a number of functions described which
look like tools for the job, but they are all mentioned as methods to
SELECT FROM, rather than INSERT INTO. I haven't found anything (so
far) on columns that can depend on/derive from other columns. Have I
missed something, or doesn't PostgreSQL do this sort of thing?

If anyone can help me out here, I'd be most grateful. If the worse come
to the worse I could always build those calculations in FileMaker and
then export them to Postgresql. BUT, that's an inelegant solution!

Thanks in anticipation

Neacal

In article <270920040324199058%deb (AT) mac (DOT) ppc>, Neacal <deb (AT) mac (DOT) ppc> wrote:

Guess that's a NO then; in this case silence does not imply assent.

While I'm here I'll say this: I wonder if those criticisms of
PostgreSQL for being SLOW (or slower than mySQL etc) relate to what
seems to me - see my above question - to be an overemphasis on making
all calculations at the SELECT time?? That reminds me of the days when
disc capacity was small.

Maybe not, I'm no expert.

Neacal