dhowe / rita Goto Github PK

• articlize should be silent when it cannot find phones
• articlize should only analyze first word (up to space)
• pluralize should be silent when it cannot find word
• pluralize should only analyze last word
• test 'set of choice'

oft-cited
off-site
deeply-nested

should be handled as 3 tokens in tokenizer
should be handled as a single token in tokenizer

• 1. ritajs tests
• 1. ritajs fix
• 2. sync tests with java
• 3. sync fix with java

Analysis & Grammar remain

see https://adnauseam.io as example

in /ritaweb/tmpl/*.html
as below:

expect(RiTa.evaluate('[foo](http://blah.com)', {})).eq('[foo](http://blah.com)');

• markdown URLS
• also html style
• add tests for both

and output in start if not silent (js as well)

• Analysis
• Haiku Grammar

@Real-John-Cheung can you look into creating a simple syntax highlighting mode for riscript in highlight.js -- the guide is here: https://highlightjs.readthedocs.io/en/latest/language-guide.html

Don't start working on it yet, but just a quick review and an estimate of how many hours are likely required to do the kind of basic highlighting that is in the editor

• replaceable writing ?
• pos markov

See /ritajs/src/markov.js`, line9

(ticket moved from ritajs)

We need to style these page better: phones.html, riscript.html, tokenization.html,(deleted) and postags.html

Lets add the head so that they look like this (but with the correct text content):

1. Note that the project structure has changed: rita2doc and RiTaWeb have been combined into RiTaWeb
2. You should mv away your current rita2doc and RiTaWeb and create a NEW fork of RiTaWeb
3. Running RiTaWeb/generate-docs.sh will generate only the website index page
4. However it the css needs to be fixed (see if this can be done by ONLY changing existing css files)
   to maintain the 3 column format

Create a notebook which allows us to import the riscript editor as a component in observable.
Something like:

import { editor } from "@dhowe/riscript"

The goal is to replace the text-area here: https://observablehq.com/@dhowe/scripting-rita
with a small version of the editor that does (at least) syntax highlighting

• page to hold the RiScript docs, here: https://github.com/dhowe/rita2js#ritascript
• page to hold POS tags, here: https://rednoise.org/rita/reference/PennTags.html

they can be static html pages in the data folder, but should match the visual style (using the same css) and also look correct on mobile

long: sequence, rsequence, norepeat
short: seq, rseq, nore

See Visitor (line 250 java, line 192 js)

so that markdown is not interpreted
need to research what this involves

I've made some changes to riscript that required a bunch of new tests, which I did in the js version, so the java version needs to be updated to match. You can do a search for 'SYNC:' in the js tests

We need to create a checklist like the one here for the current RiTa build process: #37
Each time there is a new build, someone will go through theses tests.

options:

• RiTa.compiler()
• RiTa.interpreter()
• RiTa.scripting()
• RiTa.interp()
• RiTa.riscript()
• ...

Pls remove the links shown in the image below (should look the same, just not be linked)

I've moved them into the main RiTa section (RiTa.grammar, RiTa.markov)

ritaweb/www/examples/dom/* and ritaweb/www/examples/p5/*

As these are bookmark links, they should be as simple as possible.

Please change as follows:

  /rita/#examples-wrapper -> /rita/#examples
  /rita/#gallery-wrapper -> /rita/#gallery
  etc

• add a test for each rule listed in the docs
• if any are incorrect, update the docs
• if a mismatch and you unsure if docs or code are wrong, discuss

make changes in ritaweb/tmpl/www-index.tmpl (test by running $ ./scripts/watch-site.sh) and viewing ritaweb/pub/index.html in the browser

currently we cannot (at least, in js) have a variable and a transform with the same name
so built-in transform functions may well get clobbered by rs or js vars in the context

to Processing

from here: /ritaweb/www/examples/p5/ReplaceableWriting/
to here: /rita/examples/processing/ReplaceableWriting/ReplaceableWriting.pde

Also please check RiTa.isPunct (and add tests for both)

should point to https://github.com/dhowe/rita

Test that rita.tokenize() matches what is listed (if not, adjust html)
Add whatever tests you create to the tests for java/js

test each example - make sure it works as expected
if either example is wrong, or code is wrong, add to issue below (also for java)

• add links to RiTaWeb for version v1.394
• move existing website to 'ritav1'
• script to publish new website
• publish new website
• add library.properties file to red
• publish RiTa v2.01
• check on npm
• check on unpkg
• check on github packages
• check processing
• add library.properties file to build script
• combine ritaweb and rita2doc into ritaweb
• check maven central issue
• check p5js site links
• fix kafgenstein example
• add note on v2 changes in both README
• test with observable.hq
• add gh workflow to auto-publish to maven central

• Create a simple editor (via CodeMirror) for RiScript
• Add line numbers and scrollbars (when needed only) to the editor
• Add an output and error areas below
• Add 'clear', 'save', and 'run' buttons
• Make lines highlight-able for errors
• Syntax highlighting (see colors below) for variables, choices, transforms
• Make code font at least as large as output text font
• Add keyboard shortcut for run (command-enter, same as p5)
• Add 'dark mode' checkbox which should be default=true
• Put all colors for each mode in separate CSS file
• Add color for comments (single and multiline)
• Add color for dynamic variables (prefixed with $$)

Regular expressions (in antlr format) are here: https://github.com/dhowe/ritajs/blob/master/grammar/RiScript.g4

Question: what to do about the fact that after many outputs, new ones disappear off the bottom ?

Note: plain text should be white (in dark mode, plain black in light mode)