From: Zach Bartels <no@...>
Date: 2010-02-14T09:04:26+09:00
Subject: Re: Generating all possible combinations of a 5 digit pattern.

Wow, maybe its just the size of your posting, but that looks even more
complicated,  LOL...

I do appreciate the other ideas and opinions though.   I am beginning
to wonder if I shouldn't approach it differently,  just from a method
of expedience (both in execution and iteration through pattern match
definitions).  Or at least keep my options open. (not intending to
flip flop on what i'm doing after people have given me their time )

Would it make more sense to, for instance,   parse through the project
file,  in 5 frame groups as originally intended,  but instead of
matching each group against pattern definitions,  simply record the
pattern it is currently reading instead?   If the pattern is new, then
a record of it and the frame ranges associated with it will be
recorded into some kind of data structure.

Like so.

"Pattern found:  CCCNN"  no existing definition, saving unique pattern
to database.

"Pattern found:  CNCCN" no existing definition, saving.... etc

And for any time it comes across a group of frames,  if the pattern is
already there.

"Pattern found: CCCNN"  Since it already exists as encountered
earlier,  it simply  notes the frame numbers for this particular
group,  and saves them in a data structure referring to a specific
decimation type  (i.e 5th frame).   Every group matching that pattern
gets the 5'th frame decimated from the group, be it  frame number 4,
9, 14, 49, 104, etc..  (in video, 0 is actually first, like with array
indices )


I don't necesarrily have to, or want to change from my original
proposal.  But I got to thinking about it,  and it seems like what I
was going to do was  brute-force  matching;  and when I think of
brute-force,  I think "slow"   :p   

Of course if my second idea might be easier to implement that would be
something to consider as well.    But for now I'm just playing catch
up with the previous posts and studying the code examples.   I
appreciate them greatly,   even the one liners ;)

-Zach


On 13 Feb 2010 23:12:31 GMT, Seebs <usenet-nospam@seebs.net> wrote:

>On 2010-02-13, Robert Klemme <shortcutter@googlemail.com> wrote:
>> I'm sorry, but this is overly complicated.  If I want to know whether 
>> the fourth bit is set I would use Fixnum#[]:
>
>Oh, neat!
>
>-s