Dynamic switch between selected and target

Hey everyone, Is it possible to check if a token is selected and if not ask for a target? Thanks for your help.

Not without the API. What are you trying to accomplish? There might be other methods.

I want to integrade a button into a charactersheet that first checks if a token is selected to grab the attributes from it and if this is not the case prompts the user to select a target.

Don't think this is possible.

Sigh, if &{noerror} suppressed the "no selected token" error message, we could do this with the new roll parsing feature.

A bit late to the party here, but the first trick in the startRoll() thread should be able to achieve this. Replace the getQuery() function with a getSelected() function - same basic idea but grab the inline label from a fake roll containing {{charid=[[0[@{selected|character_id}]]] }} If the contents doesn't match a UUID regex string, there's no valid character selected and you can use @{target} calls when you construct your main roll. I'm not 100% sure what would be returned in the startRoll() object when there's nothing selected - the whole thing might be undefined - but there should be some non-crashy way of distinguishing a successful ID grab from an error, even if you have to resort to a try{}catch{} block. I think it should work, anyway....

Oosh said: A bit late to the party here, but the first trick in the startRoll() thread should be able to achieve this. Replace the getQuery() function with a getSelected() function - same basic idea but grab the inline label from a fake roll containing {{charid=[[0[@{selected|character_id}]]] }} If the contents doesn't match a UUID regex string, there's no valid character selected and you can use @{target} calls when you construct your main roll. I'm not 100% sure what would be returned in the startRoll() object when there's nothing selected - the whole thing might be undefined - but there should be some non-crashy way of distinguishing a successful ID grab from an error, even if you have to resort to a try{}catch{} block. I think it should work, anyway.... Nice idea, but if I try it that way all I get is this chat output: You attempted to use a roll command looking for the value of a selected token, but no tokens are selected. const getSelected = async () => { const rxGrab = /^0\[(.*)\]\s*$/; try { let rollBase = `! {{charid=[[0[@{selected|character_id}]]] }}`, // just a [[0]] roll with an inline tag queryRoll = await startRoll(rollBase); queryResponse = (queryRoll.results.charid.expression.match(rxGrab) || [])[1]; console.log(queryResponse); finishRoll(queryRoll.rollId); // you can just let this time out if you want - we're done with it }catch(error) { console.log("undefined"); return undefined; } return queryResponse; };

Ah, roger that. I see the problem now, Roll20 never returns anything from the startRoll() when there's a @{selected} error, so the await becomes blocking. There is a way around that, though whether or not you consider it useful is another matter :) If we still call startRoll() as async, but don't await the result, we can manually wait for a certain amount of time to see if it returns anything. Having a manual timer in here would be a real issue for a super important roll, but the worst case here on a failure is that the player needs to manually select instead of the sheet automatically using the {selected}. I did a very quick test to make sure this worked, and 50ms seemed to be about right for the timeout - ymmv of course, and you might want a big error margin. Anyways, here's the code: // Promisified setTimeout const   aTimeout  =  async  ( ms )  =>   new   Promise ( res   =>   setTimeout (()  =>  { res ()},  ms )); const   getSelected  =  async  ()  =>  {      const   rxGrab  =  / ^ 0 \[ ( . * ) \] \s * $ / ;      let   rollBase  =  `! {{charid=[[0[@{selected|character_id}]]] }}` ;      let   selectedRoll ,  selectedResponse ;      startRoll ( rollBase ). then ( v => selectedRoll = v );      // Manually wait for a result with an async timeout      console . log ( `Start timeout` );      await   aTimeout ( 50 );      console . log ( `Leave timeout` );      // Process result if the .then() block above has had time to finish      if  ( selectedRoll ) {          selectedResponse  = ( selectedRoll . results . charid . expression . match ( rxGrab ) || [])[ 1 ];          console . log ( selectedResponse );          finishRoll ( selectedRoll.rollId );     }  else   console . log  ( `Nothing selected...` );      return   selectedResponse ; }; const   testRoll  =  async  ()  =>  {      let   val  =  await   getSelected ();      if  ( val )  console . log ( `Proceeding with {selected} roll` );      else   console . log ( `Proceeding with {target} roll` ); } on ( 'clicked:reptest' ,  testRoll ); Pretty similar setup, except the await is removed from the startRoll, and we use a .then() block to unwrap the promise if it returns anything. We don't even get the benefit of startRoll's 5 second timer here as far as I can tell - Roll20 doesn't give us peanuts if the {selected} roll fails. So feed whatever number you want into the timeout block, the code will then proceed and if the .then() has executed during the timeout, we'll have a selectedRoll to process, otherwise undefined. The rest of your code can continue as usual. Now that I look at it, if I was going to use this trick in a few places it would be a nice place to use Promise.race() ... Oh, and this does nothing for the Roll20 "nothing selected" error - it only appears for the player clicking the button, but I don't think you can suppress it without a CSS extension like Stylus.