String to array?:)

There's no such thing as JSON.parse in connect iq.
What you can do I this (I'm writing this in the browser, so it's not tested code):

function toArray(string, splitter) {
var array = new [5]; //Use maximum expected length
var index = 0;
var location;
do
{
location = string.find(splitter);
if (location != null) {
array[index] = string.substring(0, location);
string = string.substring(location + 1, string.length());
index++;
}
}
while (location != null);
array[index] = string;

var result = new [index];
for (var i = 0; i < index; i++) {
result= array;
}
return result;
}
[/CODE]

This code doesn't need to know how many elements is in string.
function toArray(text, delimiter)
{
var dict = {};
var len = text.length();
var delLen = delimiter.length();
var idx = 0;

while (len > 0)
{
var iend = text.find(delimiter);
iend = iend != null? iend: len;
dict.put(idx++, text.substring(0, iend).toNumber());
text = text.substring(iend + delLen, len);
len -= iend + delLen;
}
return dict.values();
}


True, but a dictionary is far less memory efficient as an array.

Yes but in this case, in fact, it does not matter because this memory will be freed very soon.

Peter

Yes but in this case, in fact, it does not matter because this memory will be freed very soon.

I would expect that the dictionary uses significantly more compute power; I would expect that Garmin probably implemented the dictionary by using the hash of the string value of the key as the indexing. That is, the key is converted to a string (surprisingly expensive for numbers), then hashed, then dropped into the storage structure.

I suggest coding up a linked list class for these sorts of tasks.

I suggest coding up a linked list class for these sorts of tasks.

I'd stick with array, as long as you had a reasonable limit to work with. Building a linked list for this would impact not only available memory, but available objects (there's a limit on those too), and if there is no limit for the data and using a linked list, the app could crash when there is still free memory, but "max objects" has been hit.

I would expect that the dictionary uses significantly more compute power

"Power" means CPU time. So make some tests and show as results.

I would expect that Garmin probably implemented the dictionary by using the hash of the string value of the key as the indexing.

I do not think so. Every object has its own 'hashCode' function and can be a key. So if the number is a key than its hashcode it is just that number.

Peter

thx! :)

Thx for help :)