dictionary - Build Alternative Data Structures with MultiValue Arrays -

i have been programming pick few months now, , still adjusting thinking of programming solutions solely use multi-value arrays. pick doesn't natively offer other data structures either database or in-memory operations * (ignoring dimensioned arrays sake of argument since still type of "array")

what creative ways of imitating other data structures? can imagine ways of implementing hash/dictionary, or linked list, using arrays , routines manipulate them. concern here gets complicated since interface data structure can't conceptually simplified object.

are there easy implement algorithms people have used create unique data structures?

---

my system: writing routines in pick add-ons epicor's eclipse, sits atop universe. know there c# interface universe, not usable of work.

the pick/multivalue dbms allows multiple values stored in single field.

a common example have single attribute street address, street1, street2, , perhaps street3 values. structure this: 01 company name 02 street1]street2]street3 03 city 04 state...

those fields of same type don't relate else. similar structure have multiple attributes values correlated. there typically 1 controlling attribute , number of dependent attributes (atb 26 used random example):

26 shipdate]shipdate]shipdate 27 shipqty]shipqty]shipqty 28 shipvia]shipvia]shipvia 

the application program manages ri, not database, though triggers can used allow database enforce ri.

a rdbms developer might prefer not use structure, can use traditional foreign keys related data. in case might see single attribute keys records in shipping file:

26 1234*01]1234*02]1234*03 

so each of values key file. *00 value application-maintained sequential numeric id each delivery. might used here in case 1 of deliveries later marked unconfirmed, value deleted, leaving this:

26 1234*02]1234*03 

the common mvdbms platforms allow 1 more level subvalues. these used reservation can quite valuable. consider above shipping data in single attribute:

26 shipdate\shipqty\shipvia ] shipdate\shipqty\shipvia 

the shipvia second shipment referenced in code item<26,2,3>.

some developers prefer use non-system delimiters. valid long won't found in data. you'll see asterisk, slash, pipe, comma, or other alphanumeric:

26 shipdate|shipqty|shipvia ] shipdate|shipqty|shipvia 

the code second shipvia might be:

field(  item<26,2> , "|" , 3  ) 

that shipvia field might foreign key table file has id key , shippername in atb1. we're talking same data structure being used 2 kinds of file: transactional data, , static tables. it's no different relational database.

yes, can create linked lists, example company record has single attribute departments. translates departments file there's attribute managers. translates employees file attribute allocated team members. each of ids translate employees, of employees might managers own teams. structure used bill of materials, multi-level marketing companies individuals have down-line, or insurance risk broken multiple companies each further break risk.

so there lot of versatility structures. can rdbms or key/value structure can done mv.

how closely answer match question? , other kind of example see?

with regard objects, qm mvdbms have oo basic, , caché dbms includes mvdbms implementation mv data can exposed objects. also, mv.net , other tools, can generate typed classes dictionary definitions (schema) allows exact same kind of orm relational platform.