«

Does anyone know of a fast implementation of a "table" data structure--
an arbitrary number of columns and rows--with the features of fast field
location and update, along with the ability to quickly add and delete
columns and rows? Something including the algorithm analysis would be
an even niftier thing to see.

It's gotta be sound enough to be reasonably equivalent in performance at
the two extremes: lots of rows and only a few columns, or a lot of
columns with only a few rows.

I'm thinking I'm smoking crack here just to think there is such a thing,
but then I'm thinking odds are pretty good that I don't know everything.

GNU GCC 3.3, but I'd prefer not to use any funny language extensions.

"Richard" <> wrote in message
news:MPG.196ec04c224c22eb989b43@199.45.49.11...
>
> Does anyone know of a fast implementation of a "table" data structure--
> an arbitrary number of columns and rows--with the features of fast field
> location and update, along with the ability to quickly add and delete
> columns and rows? Something including the algorithm analysis would be
> an even niftier thing to see.
>
> It's gotta be sound enough to be reasonably equivalent in performance at
> the two extremes: lots of rows and only a few columns, or a lot of
> columns with only a few rows.
>
> I'm thinking I'm smoking crack here just to think there is such a thing,
> but then I'm thinking odds are pretty good that I don't know everything.
>
> GNU GCC 3.3, but I'd prefer not to use any funny language extensions.

You really want news:comp.programming

If the columns are all the same type, look do a web search for "sparse
matrix"

If you are going to model a database, then model a field first. Then model
a row as a list of fields. Then model a table as a list of rows. That way,
row operations are fast. Column operations will be a bit tedious, but not
atrocious.
--
C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
"The C-FAQ Book" ISBN 0-201-84519-9
C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Anal...ject%20FAQ.htm