where each hash-reference refers to what could also be an array of the remaining fields and their values (i.e. the unique combinations of the other data, with the field names only for convenience). - later...# pseudocode { hash-reference => { date => scalar, duration => scalar } }
Update: if it weren't for the fact that this allows duplicate logical keys in the upper hash ( hash references being to possibly non-unique hashes) it would have worked .... but this is the closest I got so far. I could of course encode the referenced hash into a hash key, e.g.
But it just gets messier... for example the %hash has to be guaranteed to have its keys in a consistent order when it gets co-erced into a hash key and if that happened not to be a problem, it would still be poor practice to rely on it.#pseudo-code { join( $;, %hash ) => { date => scalar, duration => scalar } }
More update... Well since so far all I get is a down-vote and no further bright ideas, I'll just go with this inelegant solution and hack the embedded hash into being ordered by key.
see y'all in about six months when I next have a puzzler of this calibre. Sadly, I don't have time to participate properly in this site and my inclination is hardly bolstered under the circumstances.
-M
Free your mind
In reply to Re: pre-organisation of test results prior to conversion to CSV
by Moron
in thread pre-organisation of test results prior to conversion to CSV
by Moron
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |