biscotto
v2.3.1
Published
A CoffeeScript documentation generator.
Downloads
395
Readme
Biscotto
Biscotto is a tool for generating CoffeeScript documentation. The underlying architecture was based on codo; however, this project uses a variant of the excellent TomDoc notation, instead of the more verbose JSDoc. It also has many more niceties.
Features
- Detects classes, methods, constants, mixins & concerns.
- Generates a nice site to browse your code documentation in various ways.
- Intermediate JSON output to transform into any output
- Can generate a metadata blob for more advanced usage (see below for more information)
Installing
npm install biscotto
Comment parsing with TomDoc
API documentation is written in the TomDoc notation. Originally conceived for Ruby, TomDoc lends itself pretty nicely to CoffeeScript.
There are some slight changes in the parse rules to match CoffeeScript. Briefly, here's a list of how you should format your documentation.
Visibility
Every class and method should start with one of three phrases: Public:
, Internal:
, and Private:
. During the documentation generation process, you can flag whether or not to include Internal and Private members via the options passed in. If you don't have one of these status indicators, Biscotto will assume the global visibility (more on this below).
# Public: This is a test class with `inline.dot`. Beware.
class TestClassDocumentation
Method arguments
Each method argument must start with the argument name, followed by a dash (-
), and
the description of the argument:
argument - Some words about the arg!
Hash options are placed on a newline and begin with a colon:
options - These are the options:
:key1 - Blah blah.
:key2 - Blah
# Public: Does some stuff.
#
# something - Blah blah blah. Fah fah fah? Foo foo foo!
# something2 - Bar bar bar. Cha cha cha!!
# opts - The options
# :speed - The {String} speed
# :repeat - How many {Number} times to repeat
# :tasks - The {Tasks} tasks to do
bound: (something, something2, opts) =>
Examples
The examples section must start with the word "Examples" on a line by itself. The next line should be blank. Every line thereafter should be indented by two spaces from the initial comment marker:
# A method to run.
#
# Examples
#
# biscotto = require 'biscotto'
# file = (filename, content) ->
# console.log "New file %s with content %s", filename, content
# done = (err) ->
# if err
# console.log "Cannot generate documentation:", err
# else
# console.log "Documentation generated"
# biscotto.run file, done
run: ->
Return types
When returning from a method, your line must start with the word Returns
. You can list more than one Returns
per method by separating each type on a different line.
# Private: Do it!
#
# Returns {Boolean} when it works.
returnSingleType: ->
# Internal: Does some thing.
#
# Returns an object with the keys:
# :duration - A {Number} of milliseconds.
returnAHash: =>
Deviation from TomDoc
GitHub Flavored Markdown
Biscotto documentation is processed with GitHub Flavored Markdown.
Automatic link references
Biscotto comments are parsed for references to other classes, methods, and mixins, and are automatically linked together.
There are several different link types supported:
- Normal URL links:
{http://coffeescript.org/}
or[Try CoffeeScript](http://coffeescript.org/)
- Link to a class or a mixin:
{Animal::Lion}
or[The mighty lion]{Animal::Lion}
- Direct link to an instance method:
{Animal.Lion::walk}
or[The lion walks]{Animal.Lion::walk}
- Direct link to a class method:
{Animal.Lion.constructor}
or[A new king was born]{Animal.Lion.constructor}
If you are referring to a method within the same class, you can omit the class name: {::walk}
or {.constructor}
.
As an added bonus, wrapping default JavaScript types--like {String}
, {Number}
, {Boolean}
etc.--have links automatically generated to MDN.
Here's an example of using links:
# This links out to the `long` method of the same class.
#
# See {::internalLinkLong} for more info.
#
internalLinkShort: ->
# This links out to MDN.
#
# Returns a {Number} greater than zero.
internalLinkLong: ->
Status Blocks
Classes and methods can be marked as Public,
Private
, or Internal
.
You can flag multiple methods in a file with the following syntax:
### Public ###
That will mark every method underneath that block as Public
. You can follow the same notion for Internal
and Private
as well.
You can have as many block status flags as you want. The amount of #
s must be at least three, and you can have any text inside the block you want (for your own information). For example:
### Internal: This does some secret stuff. ###
If you explicitly specify a status for a method within a block, the status is respected. For example:
### Public ###
# Internal: A secret method
notShown: ->
shown: ->
shown
is kept as Public because of the status block, while notShown
is indeed Internal.
Delegation
If you're writing methods that do the exact same thing as another method, you can choose to copy over the documentation via a delegation. For example:
# {Delegates to: .delegatedRegular}
delegatedMethod: ->
# Public: I'm being delegated to!
#
# a - A {Number}
# b - A {String}
#
# Returns a {Boolean}
delegatedRegular: (a, b) ->
delegatedMethod
has the same arguments, return type, and documentation as delegatedRegular
. You can also choose to delegate to a different class:
# Private: {Delegates to: Another.Class@somewhere}
delegatedMethod: ->
Classes that are delegated should still set their own statuses. For example, even though Another.Class@somewhere
is Public, delegatedMethod
is still marked as Private
. The same documentation remains.
Defaults
Unlike TomDoc, there is no additional notation for default
values. Biscotto will take care of it for you, because it parses the CoffeeScript source and understands default values.
More Examples
For more technical examples, peruse the spec folder, which contains all the tests for Biscotto.
Generating and serving HTML
After installing Biscotto, you'll have a biscotto
binary that can be used to generate the documentation recursively for all CoffeeScript files within a directory.
To view a list of commands, type
$ biscotto --help
Biscotto wants to be smart and tries to detect the best default settings for the sources, the README, any extra files, and the project name, so the above defaults may be different on your project.
Project defaults
You can define your project defaults by writing your command line options to a .biscottoopts
file:
--name "Biscotto"
--readme README.md
--title "Biscotto Documentation"
--private
--quiet
--output-dir ./doc
./src
-
LICENSE
CHANGELOG.md
Put each option flag on a separate line, followed by the source directories or files, and optionally any extra file that should be included into the documentation separated by a dash (-
). If your extra file has the extension .md
, it'll be rendered as Markdown.
Website keyboard navigation
You can quickly search and jump through the documentation by using the fuzzy finder dialog:
- Open fuzzy finder dialog:
Ctrl-T
In frame mode you can toggle the list navigation frame on the left side:
- Toggle list view:
Ctrl-L
You can focus a list in frame mode or toggle a tab in frameless mode:
- Class list:
Ctrl-C
- Mixin list:
Ctrl-I
- File list:
Ctrl-F
- Method list:
Ctrl-M
- Extras list:
Ctrl-E
You can focus and blur the search input:
- Focus search input:
Ctrl-S
- Blur search input:
Esc
In frameless mode you can close the list tab:
- Close list tab:
Esc
Gulp-Biscotto
If you want to use Biscotto with Gulp, see gulp-biscotto.
Metadata generation
You can use Biscotto to generate a complete JSON representation of your Node.js module. Simply pass in the --metadata
flag and the path to your top-level module directory (where you keep your package.json).
biscotto --metadata ./path/to/module
Metadata generation should be considered an "advanced" way of using Biscotto. By generating metadata, Biscotto itself doesn't provide a way to turn it into HTML or interpret the comments in any way. It relies solely on the existing CoffeeScript AST. It allows for greater consumption and flexibility, however, because it traverses over the entire module.
You can find an example of the sort of metadata generated by looking at the test suite.
In (very) brief, take a look at this example:
"objects": {
"3": {
"0": {
"type": "class",
"name": "TextBuffer",
"classProperties": [
[
4,
10
]
],
"doc": " Public: A mutable text container with undo/redo support and the ability to\nannotate logical regions in the text.\n\n ",
}
},
"4": {
"10": {
"name": "prop2",
"type": "primitive",
"range": [
[
4,
10
],
[
4,
14
]
],
"bindingType": "classProperty"
}
}
}
Items are indexed by their row numbers, followed by thier column numbers. The idea is that you can traverse all along the metadata blob by simply following the location references.
For a deeper understanding of the syntax and rationale, see the original proposal at issue #43
License
(The MIT License)
Copyright (c) 2014 Garen J. Torikian
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the 'Software'), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.