Re: [chiPHPug-discuss] a bit of a sticky wicket...

[Richard L. <ce...@l-...>]

On Fri, October 12, 2007 12:28 am, jsWalter wrote:
> I have a very unusual case on my hands.
>
> I'm working with a 3rd party AR/Time mgt system from hell!
>
> The time sheets are being particularly difficult. And no, I can't
> change
> the schema, change the system, use linux, or anything else in that
> vane. I
> have to use what I have.
>
> Here is a sample record set...
>
> 	ID         DATE     line_num  seq_num  cell_num  time
> 1	APM	2007-10-01     1         0         0     8.00
> 2	APM	2007-10-01     1         1         0     8.00
> 3	APM	2007-10-02     1         0         0     0.50
> 4	APM	2007-10-02     2         0         0     6.50
> 5	APM	2007-10-02     2         0         1     6.00
> 6	APM	2007-10-02     2         0         2     5.50
> 7	APM	2007-10-02     3         0         0     1.00
> 8	APM	2007-10-02     3         0         1     0.50
> 9	APM	2007-10-02     4         0         0     1.00
> 10	APM	2007-10-02     4         0         1     1.25
> 11	APM	2007-10-03     1         0         0     1.25
> 12	APM	2007-10-03     1         0         0     2.25
>
> What I need to do is pull out it the time for each day. This example
> has 3
> days.
>
> If you look at line 1 and 2, I would use line #2 because it has 2
> LINE_NUM
> that are the same, but the second SEQ_NUM is higher. So I keep it.
>
> Lines 3 through 12 are the same date, with 5 sets of LINE_NUMs.
>  - #3 we keep, it is unique
>  - #6 we keep, because it's CEL_NUM is highest, even though it has the
>    same LINE_NUM and SEQ_NUMs
>  - #8 & #10 we keep, same reason as #6
>  - we keep both 12 & 12 because the have the same numbers all across
> the 3
> fields.
>
> Confused!?
>
> Steps, as I see them...
>   1 - collect in an array date sets
>   2 - collect into an array of LINE_NUM sub-sets by date
>   3 - collect into an array of LINE_NUM by LINE_NUM by date
>   4 - collect into an array of SEQ_NUM by LINE_NUM by LINE_NUM by date
>   5 - SEQ_NUM value has the time for that record
>
>   2007-10-01 -> 1 -> 0 -> 0 = 8.00
>              -> 1 -> 1 -> 0 = 8.00
>
>   2007-10-02 -> 1 -> 0 -> 0 = 0.50
>              -> 2 -> 0 -> 0 = 6.50
>              -> 2 -> 0 -> 1 = 5.00
>              -> 2 -> 0 -> 2 = 5.50
>              -> 3 -> 0 -> 0 = 1.00
>              -> 3 -> 0 -> 1 = 0.50
>              -> 4 -> 0 -> 0 = 1.00
>              -> 4 -> 0 -> 1 = 1.25
>
>   2007-10-03 -> 1 -> 0 -> 0 = 1.25
>              -> 1 -> 0 -> 0 = 2.25
>
> Using indexed arrays I get...
>
>   2007-10-01 -> 1 -> 1 -> 0 = 8.00
>
>   2007-10-02 -> 1 -> 0 -> 0 = 0.50
>              -> 2 -> 0 -> 2 = 5.50
>              -> 3 -> 0 -> 1 = 0.50
>              -> 4 -> 0 -> 1 = 1.25
>
>   2007-10-03 -> 1 -> 0 -> 0 = 2.25
>
> This is 90% of what I need.
>
> 10-01 and 10-02 give me the result I need form the data they have, but
> 10-03 fails. Because both SEQ_NUM and CEL_NUM are the same, I need to
> keep
> both of them.
>
> So, any minds greater than mine see anyway to do this? Without
> convoluted
> conditionals?

If you are correct about the steps, then the conditionals with good
variable names should not be that confusing...

You could also consider pulling it all into SQL.

-- 
Some people have a "gift" link here.
Know what I want?
I want you to buy a CD from some indie artist.
http://cdbaby.com/browse/from/lynch
Yeah, I get a buck. So?