dhowe / dialogic Goto Github PK

Two identical Finds at the same game state should not be guaranteed to return the same Chat.

How do we determine this range? Probably a percentage of the normalized score.

See FuzzySearch::line ~50

need to pass globals to FuzzySearch.FindX methods
also add tests for Constraint.check(globals)

Steps to Reproduce: Parse this set of Chats, repeatedly pass a ResumeEvent with parameters {type=test}

CHAT CORE_Stale_Slow {type=test, stage=CORE, stalenessIncr=0.05}
SAY slow!

CHAT CORE_Stale_Med {type=test, stage=CORE, stalenessIncr=1}
SAY medium!

CHAT CORE_Stale_Fast {type=test, stage=CORE, stalenessIncr=20}
SAY fast!

Expected: CORE_Stale_Slow will appear with increasing frequency, CORE_Stale_Fast will appear with decreasing frequency.

Actual: All three chats appear at the same frequency. In fact, after all three have played, they repeat that same pattern infinitely.

as per thejanalyn

Output from 'Execute' command should show in separate div, with separate color (maybe blue?), probably above the div for 'Validate'

in-progress

One or two checks:

• event queued
• "waiting" for an answer

This would be useful in situations where interruptability is irrelevant e.g. for determining when "small talk" chats should be triggered.

Chat metadata is intentionally not realized at this point. However I can imagine edge cases where this might be useful... (though given a large number of chats and/or searches, it may be quite inefficient)

FindAsync calls scheduler.Completed(false) regardless of current RuntimeState, resulting in a DialogicException in the first line of ChatScheduler.Completed()

Should FindAsync simply do a null check on scheduler.chat?

SET hero = (Jane | Jill)
SAY Long ago there was a girl named [selected=$hero]. 
SAY $selected was a clever girl.

OUTPUT:

Long ago there was a girl named Jane. Jane was a clever girl.
    OR
Long ago there was a girl named Jill. Jill was a clever girl.

Reported by @dcannizzaro

seems like nesting more than three level deep breaks groupings

CHAT myGrammar {noStart=true}
SAY (I|(You|(doh|why|no)|oUY)|They) (want|hate|like|love) (coffee|bread|milk)

You can use the 'LoadUrl' button on this page as an example.

We should be able to use a github raw URL, or a Google-docs URL.

Start by testing with URLs of these files (in RAW mode): https://github.com/dcannizzaro/tendAR/tree/ca06ee409fd15420135e9567ed921645b337b0b4/TendAR/Assets/Libraries/Dialogic/GuppyScripts

For example: https://raw.githubusercontent.com/dcannizzaro/tendAR/ca06ee409fd15420135e9567ed921645b337b0b4/TendAR/Assets/Libraries/Dialogic/GuppyScripts/gscript.gs.txt?token=AAtBZge0aTiAfaPesTi2C3SE8_Zssiuwks5ay-qjwA%3D%3D

For the Google docs case, see also: #25

Primarily a performance optimization

Think about staleness iterations in search over long-play situations

We currently do not treat #Labels as dynamic data (which means $variables and (value1|value2) are not replaced with literals. There may be cases however where this would be useful...

This applies to Labels in GO, DO, and OPT.action(Go).

This was requested in meeting on April 4. Syntax should be:

GO #(LabelA | LabelB)

Steps to Reproduce: Parse the following chats. Trigger CORE_Long_Interruptable, while it is playing, trigger CORE_Short_Interruption.

CHAT CORE_Long_Interruptable {type=poke, stage=CORE}
SAY Oh, it's you...
SAY It's been a long time. How have you been?
SAY I've been really busy being dead.
SAY You know
SAY After you MURDERED ME
SAY Okay. Look. We both said a lot of things that you're going to regret.
SAY But I think we can put our differences behind us.
SAY For science.
SAY You monster.

CHAT CORE_Short_Interruption {type=shake, stage=CORE}
SAY Stop that.

Expected: CORE_Long_Interruptable will resume after CORE_Short_Interruption

Actual: CORE_Long_Interruptable does not resume after CORE_Short_Interruption

validate: 'validate the syntax of your script'
execute: 'print the output of the last command above'

Add a checkbox to the form, which should be automatically sent along with the POST request. Then I can handle the modification of the LintServer

not sure if this can be done with control-c (preferable), or if we need a 'copy' button
currently we get the square brackets in the copy

I'm currently accomplishing this with the following line in Event.cs:

        public UpdateEvent(Command c) {
            this.data = c.realized;
            this.data.Add("duration", c.ComputeDuration()); // <----
        }

To avoid a long line like below (but not for the 'validate' output, only the 'execute' output)

generate html class documentation (like javadoc), preferably from within Visual Studio

Enable authors to specify a transitional (smoothing) Chat to run whenever the specified Chat is interrupted. This should be settable via the metadata key 'onResume'. See this script

This is related to #21, in which we provide a Google docs URL for validating

Correctly handle local scope on Chat completion

Note: if we clear, conditional exit of Chat will be a problem (all branches of an Ask thus must go to other Chats) - otherwise (or if they return) scope will be empty

 ${noun.pluralize()}

calls pluralize() on $noun AFTER symbol binding

Such as:

CHAT c1 { emotion = (sad,depressed,gloomy) }

OR

CHAT c1 { emotion = (sad | depressed | gloomy) }

Need to allow arrays as metadata values, at least for Chats

On hold for post Chat-refactor

[Pending Question]

I've implemented a constant for all Command event durations, called Defaults.GLOBAL_TIME_SCALE, which defaults to 1. This affects only the scheduler timing (that is, the duration for each Command). Setting it to 1.5, for example, would extend these durations by 50% (thus slowing the system down)

What, if any, other timing properties should be affected here? Should this include literal properties (set by either the game or a Chat author) as well? For example, if an author specifies an Ask timeout of 5 seconds, should this be scaled as well? I think not, but want to verify the thinking here

Example:

CHAT myGrammar {noStart=true,defaultCmd=SET}
start = <subject> <verb> <object>.
subject = I | You | They
object = coffee | bread | milk
verb = want | hate | like | love
SAY $Start

-> You want milk. You want milk. You want milk. You want milk. You want milk. You want milk. You want milk.

Got this error from tests -> Run Item after merging the latest commit

add better tests for all operators

Extected: IDictionary with Chat's meta values e.g. {stage=S1, type=hello, length=long}