Character Sheet Enhancements!

Hi Keagan,

Assuming you are talking about the D&D 5th Edition by Roll20 Sheet, you can either create a thread in the Bug reports forum, or post to the D&D 5th Edition by Roll20 thread. If you are talking about any other D&D sheet, the new thread option would be best.

CSE has parsing issues with inputs that utilize nested quotes and/or queries that require special character substitutions which can end an attribute value prematurely(ie nested quotes) and/or break macros(ie nested queries).
I know the Dev's are working on a fix, but...
as a temporary workaround, it looks like you can substitute the leading ampersand with `&`.  So `"` would become `"`  Seems like I recall we used to have to do something similar when posting macros with special characters on the forums.   
example;

Legacy Mode:
<button type="roll" name="roll_abilities" class="command-button" value="/w &quot;@{character_name}&quot; @{ability_buttons_macro}">
CSE Mode:
<button type="roll" name="roll_abilities" class="command-button" value="/w &amp;quot;@{character_name}&amp;quot; @{ability_buttons_macro}">

Nicholas said:

Hey keithcurtis -

I've made sure to bring this up to our developers and we now have a potential fix in the pipeline for deployment this week or next. Thank you for the heads up here, it's greatly appreciated!

Thanks for the update, Nicholas!

@Nicholas

Ditto on Keith's sentiments. 

Any progress on the parsing issues?  Thanks

Rolltemplates seem to be throwing new errors when using the conditional rolltemplate functions where the roll is missing. 

Just made a quick test sheet using the following: 

<button type="roll" value="&{template:test} {{testroll=[[1d4]]}}">Shoot</button>
<button type="roll" value="&{template:test}">Blanks</button>
<rolltemplate class="sheet-rolltemplate-test">
    <h1>Roll Template Errors</h1>
    {{#rollTotal() testroll 1}}
        You rolled a 1
    {{/rollTotal() testroll 1}}
    {{#rollTotal() testroll 2}}
        You rolled a 2
    {{/rollTotal() testroll 2}}
    {{#rollTotal() testroll 3}}
        You rolled a 3
    {{/rollTotal() testroll 3}}
    {{#rollTotal() testroll 4}}
        You rolled a 4
    {{/rollTotal() testroll 4}}
    
</rolltemplate>

Checking the console when pressing the buttons, no errors are thrown when the rolltemplate is provided a roll, but each conditional throws an error if the Blank button is pressed. My working sheet piles up a lot of console logs very quickly right now and this seems like it should be an exception to the error checks. 

Nicholas said:

Hey folks - 

Just wanted to drop another quick update here.

keithcurtis, vÍnce, and others - The fix for the bug discussed in this post is live and should be good to go. If you run into this problem again or see any other reports of continued issues, please don't hesitate to let us know!

Sweet! Thanks, I'll check it out.

I am still having these same issues on my Chromebook, that I can see the items and journal pages but when I try to open my DnD 5e Character sheet it only shows a loading symbol (Though you can barely see the Bio&Info, Character Sheet and Attributes & Abilities tabs like it wants to load). I'm new to both roll20 and this Chromebook so I'm not sure what to do, any help would be appreciated

Nic B. said:

Hi Richard,

Just wanted to note that this has been added to some continuing work the team is doing around Roll Templates.

Thanks for bringing it to our attention.

While you're doing work on rolltemplates, could you fix the issue with logic functions and number below zero.

In legacy mode, the number comparison functions stop working if the number you are comparing is below zero.

I could've sworn I raised it a bug, but it will have been a couple years back.

I hadnt noticed any comments about this, but in sheet workers, does the case of attribute names matter now.

In another thread, I just posted a code example, and needed to tweak what used to be working code but was now unworking. This is the code:

The important line is this one:

      const score = values[stat.toLowerCase()];

In the past, the following line would have been used instead:

      const score = values[stat];

Because attribute names were case-insenstitive. The only time it mattered was in the event line, where you had to use lower case, regardless of the attribute in the sheet.

But now it seems attribute names are case sensitive. Looking for the attribute "Agility" fails, but "agility" works.

I could see this breaking sheet workers in a lot of sheets (though admittedly, if it did there have probably been posts about it by now). Was there a notice about this that I've missed?

So what is the status with the sheet enhancements? Is it stable, now? I wonder whether I could start migrating my sheet from legacy to the new sanitization system...

It's been stable since around mid-April.

I've refactored 2 sheets to CSE, and created 4 sheet right away to CSE since.

Still some parsing issues that the Devs are working on...

I've converted the PF Community sheet to CSE.

Thanks for the answer. Since you've already done it, did you notice any particular pitfall, or do you have any advice on the conversion process?

Laurent said:

Thanks for the answer. Since you've already done it, did you notice any particular pitfall, or do you have any advice on the conversion process?

This thread, as well as the CSE on the wiki has the conversion to CSE pretty much covered.  Just plan on doing a lot of testing before you go live with it to make sure everything is acceptable.  Depending on the size and complexity of the sheet, this can take some time.  Seems like I'm still fiddling with the css weeks after converting.  Granted, this could be due to the css/html that's spanned over five years, so there are definitely a few curmudgeons in there that CSE uncovers. ;-)  There's still some parsing issues where special characters used within sheet macros can be prematurely converted.  (IIRC; nested quotes, nested queries)  I think you can workaround this until the Devs rework/revisit the parser by "double-substituting" the leading ampersand. ie "
Good luck.

Hi everybody,

I am not sure whether I am here in the right thread, but I couldn't find a better.

Question: is there a way to get the D&D 5E variant for critical damage working in the 5E OGL sheet? I want to have "max die + die" as crit damage instead of "2 x die". E.g. crit damage of a battleaxe would be "8 + d8" instead of "2d8".

I really don't like all the crits have less damage than some normal damage rolls anymore. But I couln't find a way to change that. Only thing is to change the crit range.

Thanks!
Bernd

Hi Bernd,

For this sort of question, your best bet is to make your own thread, although it could be asked in the D&D 5th Edition by Roll20 feedback thread. This question comes up every now and then. Here's a recent thread, with some suggested workarounds.

While working on a sheet today I realized that I'd really like to see a new feature. It would be awesome if the repeating section html parsing would add the rowID to any IDs created in that section or referenced in that section. Preferably this would be done similar to how attribute names are done where it looks for the base id in the row, if it doesn't find it, it doesn't add the row id.

This would allow us to use things like labels referencing checkboxes/buttons/radios without needing to use some of the old work arounds for pseudo-buttons.

Example code of html that this would allow:

<fieldset class="repeating_skills">
  <input type='checkbox' id='skill-enable' value='1' class='hidden' checked>
  <input name="attr_name" type="text" val="" data-i18n-placeholder="skill name" title="@{repeating_skills_$0_name"/>
  <input class="boxed" name="attr_mod" type="number" val="" title="@{repeating_skills_$0_mod"/>
  <button name="roll_roll" value="@{template_start} {{title=@{name}}} {{roll=[[1d20+@{mod}]]}}" title="%{repeating_skills_$0_roll}" type="roll">
  </button>
  <label for='skill-enable'>On/Off</label>
</fieldset>

Great suggestion, Scott. A feature like that would be so helpful.

Wow, if you do fix nested queries so they work naturally, that will be huge.

Nicholas said:

1) We're fixing the old implementation that uses character codes so that older sheets aren't broken.

2) We're implementing smarter and better detection of nested roll syntax - i.e. nested usage of ?{ } syntax. We originally thought about adding new syntax for nested roll queries specifically, but we decided against that since this feels more natural and easier to use.

Does that also mean that user macros will also have an easier time of coding roll queries?

Given that the thread Now on Dev Server: Custom Roll Parsing for Character Sheets was immediately closed without any further feedback once the update had been pushed to the Live server...  I decided to post this here;

Since the new roll parsing was added to the live servers, macros that include a roll template with just an empty key ie {{=foo}} display the line twice with the duplicate line substituting "computed::" for the empty key. 

&{template:default}{{name=test}}{{=[[1d20]]}}

BTW: I've found you can include an empty space(or multiple spaces) as a workaround. Prevents the extra "computed::" line and still has a similar expected behavior that depended on the roll templates allprops() formatting.

&{template:default}{{name=test}}{{ =[[1d20]]}}

I think I'm seeing a bug with the CSE and the charactermancer. The show/hide functions don't seem to work anymore. I am testing this in the custom sheet sandbox. It could also just be me, but I created a simple test below.

The behavior:

1) Sections tagged with the "sheet-choice" class hide on load.

2) The showChoices function only works if the "choice" class is also added.

3) hideChoices does not appear to work at all once the choices are shown.

Some HTML to reproduce:

<div>
    <button type="action" name="act_create">Create</button>
</div>

<charmancer class="sheet-charmancer-intro">
    <div>
        <label>Choice</label>
        <select class="choice testselect" name="comp_testselect" required>
            <option />
            <option value="1">1</option>
            <option value="2">2</option>
            <option value="3">3</option>
        </select>

        <div class="sheet-choice choice testchoice1">
            <label>Choice 1</label>
            <input type="text" name="comp_choice1" />
        </div>

        <div class="sheet-choice choice testchoice2">
            <label>Choice 2</label>
            <input type="text" name="comp_choice2" />
        </div>

        <div class="sheet-choice choice testchoice3">
            <label>Choice 3</label>
            <input type="text" name="comp_choice3" />
        </div>

        <button type="finish" value="apply">Apply Changes</button>
    </div>
</charmancer>
<charmancer class="sheet-charmancer-final">
    <div>
        <h2>Please wait...<h2>
    </div>
</charmancer>

<script type="text/worker">
    on('clicked:create', function() {
        startCharactermancer('intro');
    });

    on('mancerchange:testselect', function(eventinfo) {
        hideChoices();
        if (eventinfo.newValue) {
            let choice = parseInt(eventinfo.newValue)||0;
            switch (choice) {
                case 1:
                    showChoices(['testchoice1']);
                    break;
                case 2:
                    showChoices(['testchoice2']);
                    break;
                case 3:
                    showChoices(['testchoice3']);
                    break;
            }
        }
    });

    on('mancerfinish:apply', function() {
        finishCharactermancer();
    });
</script>

I did some more testing and came up with a workaround (in case anyone else or the Roll20 team need it). It looks like there is a css selector that looks like:

.sheet-choice {
    display: none;
}

However, the functions (hideChoices, showChoices) have been updated for CSE. They omit the "sheet-" prefix when updating the DOM. So you see a toggle back and forth of "choice" and "choice-showing". The toggle works, but there isn't any corresponding CSS. So I added a simple workaround:

.choice {
    display: none;
}

And now it works great! The required fields are ignored when hidden, etc.

Possible bug:

attribute names including "_repeating" within the name, do not trigger API events.

Not sure if this is related to the CSE update or one of the other recent updates...  

Example;

Using "delete_repeating_ability" no longer works to trigger API events

        on('change:delete_repeating_ability', function(eventInfo) {		
            console.log('~~~~| DING |~~~~: '+eventInfo.sourceAttribute);
	});

but changing the attribute name to "delete_rep_ability" works

	on('change:delete_rep_ability', function(eventInfo) {
	    console.log('~~~~| DING |~~~~: '+eventInfo.sourceAttribute);
	});

It appears the API is now interpreting "_repeating" in an attribute's name as a repeating section/or ignoring it all together.