Droplet seeks to re-envision "block programming" as "text editing". It is useful as a transitional tool for beginners using languages like Scratch, and is a go-to text editor for everyone on mobile devices (where keyboards don't work so well).
To embed, call new droplet.Editor() on a div.
<html>
<head>
<style>
#editor {
position: absolute;
top: 0; left: 0; right: 0; bottom: 0;
}
</style>
</head>
<div id="editor">
</div>
</html>require ['droplet'], (droplet) ->
editor = new droplet.Editor document.getElementById('editor'), {
# Language
mode: 'coffeescript'
# Options for the CoffeeScript parser
# (the JavaScript parser currently takes the same options)
modeOptions: {
blockFunctions: ['fd', 'rt', 'lt', 'alert'] # Override which blocks turn blue (names not here will be purple and have editable names)
valueFunctions: ['sin', 'cos'] # Same for green blocks. Defaults are the pencilcode functions.
eitherFunctions: ['write', 'confirm']
}
# Palette description
palette: [
{
name: 'Palette category'
color: 'blue' # Header color
blocks: [
{
block: "for [1..3]\n ``"
title: "Repeat some code" # title-text
}
]
}
]
}
editor.setValue '''
for i in [1..10]
document.write 'hello world'
'''Droplet uses Grunt and npm to build. Run:
git pull https://github.com/dabbler0/droplet.git
cd droplet
npm install
grunt allWhen developing, run:
grunt testserverThis will run the development server and watch the src/ and example/ directories for recompilation. Visit localhost:8000/example/example.html for a simple running environment. A view debugger is available at localhost:8000/example/test.html.
Run grunt all to run the tests.
Make a CoffeeScript (or JavaScript) file that looks like this:
define ['droplet-helper', 'droplet-parser'], (helper, parser) ->
class MyParser extends parser.Parser
markRoot: ->
return parser.wrapParser MyParserPut it in src/myparser.coffee. Add it to the build system in requirejs-paths.json:
{
// etc...
"droplet-myparser": "myparser" // meaning "myparser.coffee" -- or whatever you named your file
// etc...
}Require it from the controller:
define ['droplet-helper',
'droplet-coffee',
'droplet-javascript',
'droplet-myparser', # This is us!
'droplet-draw' # etc, etc..
# ...
], (helper,
coffee,
javascript,
myparser, # Us again!
draw, # etc, etc..
# ...
) ->
modes = {
# etc.. etc..
'mylanguage': myparser
}Then grunt. Your mode is integrated!
To have your parser actually put blocks in, you will need to do some things in the markRoot function. Fields and methods you need to know about:
# Get the raw text passed into the parser:
@text
# Get the `modeOptions` passed down from editor instantiation
@opts
# Add a Block
@addBlock({
# Configure the location of the block (all required)
bounds: {
start: {line: Number, column: Number} # Lines and columns are zero-indexed
end: {line: Number, column: Number}
}
depth: Number # Depth in the tree
# Configure the block you're about to add (all optional)
color: '#HEXCOLOR'
precedence: Number
classes: [] # Array of strings.
})
# Add a Socket
@addSocket({
# Configure the location of the socket (all required)
bounds: {
start: {line: Number, column: Number} # Lines and columns are zero-indexed
end: {line: Number, column: Number}
}
depth: Number # Depth in the tree
# Configure the block you're about to add (all optional)
precedence: Number
accepts: {'string': Number} # Maps class names (from block 'classes' array) to an acceptance level
# (see "acceptance levels")
classes: [] # Array of strings
})
# Add an Indent
@addIndent({
# Configure the location of the socket (all required)
bounds: {
start: {line: Number, column: Number} # Lines and columns are zero-indexed
end: {line: Number, column: Number}
}
depth: Number # Depth in the tree
# Configure the indent you're about to add (all optional)
prefix: ' ' # String that is a prefix of all the lines
classes: [] # Array of strings
})Call these in markRoot to insert Blocks, Sockets and Indents.
You may also want to override the following callbacks:
# Parens is called whenever a block is dropped into
# another block; you are allowed to change the leading
# and trailing text of the block at this moment (for parentheses, semicolons, etc.)
#
# The default for this is based on the precedence numbers.
MyParser.parens = (leading, trailing, node, context) ->
# "leading" is the leading text owned by the block and not its children;
# "trailing" is similar trailing text. "node" is the Block that is being dropped,
# and context is the Socket or Indent it is being dropped into.
return [newLeading, newTrailing]
# Text to fill in an empty socket when switching modes:
MyParser.empty = "blarg"
MyParser.drop = (block, context, preceding) ->
# block: the block that user is dragging
# context: the place the user is dropping that block into
# preceding: if in sequence, the block immediately before
# block, context, and preceding will have
# properties `classes` (from when you created the block),
# `precedence`, and `type` ('block', 'socket', 'indent', or 'segment')
if allowedIn(block, context)
return helper.ENCOURAGE
else if maybe(block, context)
return helper.DISCOURAGE
else
return helper.FORBID