It has been suggested that writing filter files would be made easier if
some kind of "scoring" facility were available. It has also been
suggested that user variables be implemented. I feel that a full-blown
implementation of variables would amount to a mini programming language
and would be getting out of hand. On the other hand, a very simple
scoring system with a command like "score <n>" provides in effect just
one variable, which is a bit restricting (though less so than no
variables :-).
What do people think about the following compromise:
10 numeric variables are available with the names n0 - n9. They start
out with a value of zero when a filter file is entered. Their values
can be changed by the command
add <number> to <name>
where negative numbers are permitted. Their values can be used in
other commands by variable expansions such as $n0.
A simple scoring scheme could be implemented by something like
if <condition 1> then add 1 to n1 endif
if <condition 2> then add 2 to n1 endif
if <condition 3> then add -1 to n1 endif
if <condition 4> then add 1 to n1 endif
if $n1 above 3 then seen finish endif
if $n1 above 0 then freeze endif
with more complicated schemes making use of more than one variable.
Opinions, please!
--
Philip Hazel University Computing Service,
ph10@??? New Museums Site, Cambridge CB2 3QG,
P.Hazel@??? England. Phone: +44 1223 334714
--
*** Exim information can be found at
http://www.exim.org/ ***
