getTokens - without check every token?

[Nertht]

Hello again.
Is it possible, to get array of tokens in XY coordinations without checking every token in map? Wiki: getTokens() function works well, but with larger amount of tokens in map, it deffers macro quite lot.

[aliasmask]

Yes and no. If you read Wiki: getTokens() you can specify a range around a token. So, yes you can specify filters to the getTokens, but no you can't specify with just an x,y coordinate. It has to be centered on a token. One method would be to move a token to area and make the check. Also, you can get visible tokens and filter those with canSeeToken from that token's position. I don't know what your application is, so I can't give you specifics on your needs.

[Nertht]

First of all, thanks for instant reply,
but my problem isnt, that I cant define tokens in XY. Problem is, that i cant do that without checking every token in the map. When I use Wiki: getTokens() and define condition (for example - State), MT will check every token with defined state. But when I define area as condition, MT will check every token in the map and after it will define which tokens from whole map is in XY. And when i have about 10000 tokens, using of this function will deffer execution really lot.
Is there any better way to do it?

[wolph42]

No. If you define area correctly then it will only retrieve tokens in that area. That method can however be insanely slow. Fastest method is using states, next you can move a token to the centre of the area you wish to grab and use that token together with the range option.
I usually just cycle through all tokens, using a slice method in case their are more than 1000 tokens(which is the loop limit in mt)

[Nertht]

Slice method? Could you tell me more about it?

[wolph42]

Slice method? Could you tell me more about it?

Argh on.the.autocorrect. splice method. Anyway, it's a function on the bag of tricks that creates an array of lists of exactly 1000 elements per list, allowing for a loop upto 1000000 tokens. You can cycle through them with two for each loops. Usually you then run into the other limit of max 2 {} nesting, which can be omitted by the '' method(read the working with two code levels article, linked in Sig) .

[wolph42]

examples are always so much more useful, im currently working on something so its a copy paste for me:

Code: Select all

[h:allToks    = getTokenNames(",", '{layer:["TOKEN", "HIDDEN", "OBJECT", "BACKGROUND"]}')]
[h:splices    = bot_createSplices(allToks)]
[h,foreach(splice, splices), CODE:{
    [foreach(tok, splice), CODE:{
        [x        = getTokenX(0, tok)]
        [y        = getTokenY(0, tok)]
        [if(x >= 0 && y >= 0 && x <= 35 && y <= 35), CODE:{
            <!-- your code here -->
        ''
        };{}]
    ''
    }]
''
}] 

bot_createSplices() is a function available as soon as you have lib:ontokenmove installed in your framework.
The '' are single quotes (two of them. And placed where they are, they allow for more than two code levels.

[Nertht]

Thanks a lot, it saves me again...