SlimUser Guide

Overview

Slim is a library for generating HTML from concise, lightweight templates. Slim is based on Jade for javascript and Slim for Ruby.

Features include:

  • indentation driven - closing tags not needed
  • CSS shortcut notation for #id and .class attributes
  • ${...} notation to interpolate Fantom code
  • Configurable HTML, XHTML or XML tag endings
  • efan template generation
  • Template nesting with Layout pattern.

.

ALIEN-AID: Turn Slim templates into powerful HTML components with efanXtra!

Quick Start

  1. Create a text file called Example.slim
    -? using concurrent
    doctype html
    html
      head
        title afSlim Example
        meta (name="keywords" content="fantom html template language")
    
      body
        h1 Slim Example
    
        h2 Element shortcut notation:
    
        div#slimer This div has an ID of 'slimer'
        div.wombat This div has a class of 'wombat'
        div (style="color: red;") Attributes are specified in brackets
        div You can even embed <abbr>HTML</abbr> tags!
    
        | Use the pipe character for text
          that may be spanned across
          multiple lines!
    
        // This is a single line comment
    
        /* This is a block...
           .. or multiline comment
    
        /! This is a HTML comment
    
        // Use -- to execute Fantom code
        -- echo("Hello Pips!")
    
        // Use == to print the result of a Fantom expression
        == "Hello " + ctx["name"] + "!"
    
        // Use $(...) notation to embed Fantom expressions
        | Hello ${ ctx["name"] }!
    
        // Embedding Javascript is easy!
        script (type="text/javascript") |
          for (var i=0; i<3; i++) {
            console.info("Greetings from Slim!");
          }
    
        // Use ';' to condense elements onto one line
        ul; li; a (href="#") One line!
  2. Create a text file called Example.fan
    using afSlim
    
    class Example {
        Void main() {
            ctx  := ["name":"Emma"]
            html := Slim().renderFromFile(`Example.slim`.toFile, ctx)
            echo(html)
        }
    }
    
  3. Run Example.fan as a Fantom script from the command line:
    C:\> fan Example.fan
    <!DOCTYPE html>
    <html>
        <head>
            <title>afSlim Example</title>
            <meta name="keywords" content="fantom html template language">
        </head>
        <body>
            <h1>Slim Example</h1>
        ....
        ....
    

Syntax

The first non-whitespace characters of each line defines the content of the line:

doctype : <!DOCTYPE ... >
    -?  : using statement
    --  : fantom code
    ==  : fantom eval
    //  : single line comment
    /*  : block comment
    /!  : HTML comment (single line only)
   a-Z  : HTML element
     |  : plain text

ALIEN-AID: Whitespace indentation is very important! If your HTML looks wrong, click Show Whitespace in your editor / IDE and make sure you are not mixing up tabs and spaces.

Doctype

Start a line with doctype to print a document type. The common type would be html:

doctype html
  <!DOCTYPE html>

If you specify a doctype of xml then what follows is taken to be the charset and is printed with the XML processing instruction. (In this case, the directive doctype is a bit of a misnomer.)

doctype xml ISO-8859-1
  <?xml version="1.0" encoding="ISO-8859-1" ?>

Supported XHTML Doctypes:

doctype xhtml
  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
    "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">

doctype xhtml strict
  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">

doctype xhtml frameset
  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Frameset//EN"
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-frameset.dtd">

doctype xhtml mobile
  <!DOCTYPE html PUBLIC "-//WAPFORUM//DTD XHTML Mobile 1.2//EN"
  "http://www.openmobilealliance.org/tech/DTD/xhtml-mobile12.dtd">

doctype xhtml basic
  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML Basic 1.1//EN"
    "http://www.w3.org/TR/xhtml-basic/xhtml-basic11.dtd">

doctype xhtml transitional
  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

Supported HTML4 Doctypes:

doctype html4
  <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN"
    "http://www.w3.org/TR/html4/strict.dtd">

doctype html4 frameset
  <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Frameset//EN"
    "http://www.w3.org/TR/html4/frameset.dtd">

doctype html4 transitional
  <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
    "http://www.w3.org/TR/html4/loose.dtd">

Note that the doctype directive is just a short hand notation for common DOCTYPE declarations, it does not enforce constaints on your template or alter it in any way.

To print custom DOCTYPE declarations, use the | character to print a standard string:

| <!DOCTYPE wotever PUBLIC "http://www.wotever.com">

Using Statements

Start any line with -? to add a Fantom using statement.

-? using afSlim

The using statement means you don't have to use fully qualified class names:

-? using concurrent
== Actor.locals("my.value")

Elements

Element lines are formatted as:

element[#id][.class][.class] [(attributes)] [text]

div Text here           --> <div>Text here</div>
div#wombat Text here    --> <div id="wombat">Text here</div>
div.wombat Text here    --> <div class="wombat">Text here</div>
div(data-type="wombat") --> <div data-type="wombat"></div>

Whitespace before the attribute brackets is optional:

a (href="http://www.fantomfactory.org") Fantom
a(href="http://www.fantomfactory.org") Fantom

Attributes may also be enclosed in square brackets:

div[data-type="wombat"]

Use all the shortcuts together:

div#robert.juice.media (data-on="You Tube") Rap News

<div id="robert" class="juice media" data-on="You Tube">Rap News</div>

Note that attribute contents are not parsed. Whatever is inbetween the ( - ) or [ - ] is rendered exactly as is. This means you should not mix id and class attributes with shortcut notation as this would result in two id and class attributes, which would be invalid.

div#top (id="bottom")

<div id="top" id="bottom"></div>

If the text of an element needs to start with a bracket, then use empty attribute notation to avoid confusion:

div() (In Brackets)

<div>(In Brackets)</div>

If an element has no text, then it may be immediatly followed by a semi-colon ; to start a fresh line. This concise syntax prevents <li>, and other empty elements, from taking up a whole line of their own.

ul
  li; a (href="#") home
  li; span.highlight other page
  li; == ctx.otherPage

Single Line Comments

Start any line with // to add a comment.

// This is a comment
    div This is still rendered

Comments do not appear in the generated html, but do appear in the efan template.

Block Comments

Start any line with /* to add a block comment.

/* This is a Block comment
     Block comments span multiple lines
       And are great for temporarily removing chunks of HTML

Block comments do not appear in the generated html, but do appear in the efan template.

HTML Comments

Start any line with /! to add a HTML comment.

/! This is a HTML comment

becomes

<!-- This is a HTML comment -->

HTML comments do appear in the generated HTML.

Fantom Code

Start any line with -- to write Fantom code. Use to call efan helper methods.

-- echo("Hello Mum!")

Note because Slim does not have end tags, you do not specify opening or closing { curly } brackets to denote blocks of code. Due to indentation, Slim works out where they should be.

-- if (ctx.doughnuts.isEmpty)
  p You're not a *real* policeman!
-- else
  ul
    -- ctx.doughnuts.each |nut|
      li Mmm... ${nut.filling}!

Fantom Eval

Start any line with == to evaluate a line of Fantom code and print it out in the template

== ctx.doughnut.filling

The resulting string is printed raw and is not HTML escaped.

Plain Text

Any line starting with a | denotes plain text and is printed raw. You can even embed HTML:

| Look at how <small>BIG</small> I am!

Unlike other line types, text may flow / span multiple lines.

| Use the pipe character for text.
  It also lets text be spanned
  across multiple lines!

You can use | as the first character of an element. So the following:

p
  | More recently, I discovered
    Fantom
    a niffty pragmatic language

May be re-written as:

p | More recently, I discovered
    Fantom
    a niffty pragmatic language

This is handy for writing <script> tags:

script (type="text/javascript") |
  console.info("Hello...");
  console.info("     ...Pips!");

Text may be mixed with elements:

p
  | More recently, I discovered
  a (href="http://fantom.org/") FANTOM
  |  a niffty pragmatic language (*)
  | which runs on Java and .NET

(*) Note the extra leading space at the start of the line. This prevents it from butting up against the previous <a> tag:

... FANTOM a niffty pragmatic language

and not

... FANTOMa niffty pragmatic language

Slim trims 1 character of whitespace after a | and preserves trailing whitespace.

Fantom Interpolation

Similar to Fantom Str interpolation, you can output Fantom expressions anywhere in the template using the standard ${...} notation;

div Mmmm... ${ctx.doughnut.filling} is my favourite!

By default all text rendered via ${...} is XML escaped. To print raw / unescaped text use $${...}. Backslash escape any expression to ignore it and print it as is.

To summarise:

.
  ${...} : XML escaped
 \${...} : ignored
 $${...} : raw / unescaped
\$${...} : ignored

For simple expressions, the curly brackets may be omitted:

div Mmmm... $ctx.doughnut.filling is my favourite!

Localisation

Similar to Fantom Locale Literals, you can output localised Fantom strings anywhere in the template using the standard $<key> notation;

div Mmmm... $<doughnut.filling.raspberry> is my favourite!

Values pertaining to the key must be defined in locale props files as specified by Fantom Localized Properties. Property resolution follows the usual fallbacks of locale -> lang -> en. The key itself is printed should a localised value not exist.

By default all text rendered via $<...> is XML escaped. To print raw / unescaped text use $$<...>. Backslash escape any expression to ignore it and print it as is.

To summarise:

.
  $<...> : XML escaped
 \$<...> : ignored
 $$<...> : raw / unescaped
\$$<...> : ignored

Arg Interpolation

Arguments may be passed as fantom literals / code to be interpolated into the localised values. Up to 4 arguments may be provided.

div Mmmm... $<doughnuts, "cream", raspberry()> are my favourites!

Arguments are specified by ${x} notation in the property value.

doughnuts = Doughnuts with ${1} and ${2}

If more arguments are required, a list may be passed as the first argument.

div Mmmm... $<doughnuts, ["cream", "raspberry", "vanilla", "strawberry", "chocolate"]> are my favourites!

A map may also be passed as the first arguments.

div Mmmm... $<doughnuts, ["1st":"raspberry", "2nd":"cream"]>

Map arguments are interpolated via their keys.

doughnuts = Doughnuts with ${1st} and ${2nd}

No checks are made to ensure all args are used, and no error is thrown should an arg does not exist.

HTML vs XHTML vs XML

HTML

By default Slim renders tags as HTML5 elements; that is, all tags representing void elements such as meta, br and input are printed without an end tag:

<input type="submit">
<br>

Warnings are logged should a void element NOT be empty.

HTML5 documents should be served up (from a web / app server such as BedSheet) with a Content-Type of:

text/html

While HTML is nice for browsers, this format doesn't lend itself to XML parsing; should you wish to use Sizzle for instance. So Slim offers alternative renderings of tag endings.

XHTML

By creating Slim with a TagStyle of xhtml all void elements are rendered as a self-closing tags. Warnings are logged should a void element NOT be empty.

<input type="submit" />
<br />

All non void elements are NOT rendered as self closing, even when empty.

<script></script>

XHTML documents should be served up with a Content-Type of:

application/xhtml+xml

Note that Internet Explorer versions 8 and below are reported not to accept this MimeType, resulting in a download dialogue box shown to the user.

Note that XHTML files must declare the xhtml namespace in the html tag or browsers will not render the page. Example:

<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml">
  ...
</html>

XML

If you create Slim with a TagStyle of xml then ALL empty tags are self-closing and void tags have no special meaning. Use this style when Slim is to create pure XML documents.

Depending on usage XML documents may be served up with a Content-Type of either:

text/xml
application/xml

Custom Components

If standard HTML tags aren't enough, you may register your own tags with Slim that render custom templates.

Pass your cutsom SlimComponents to the Slim ctor and they get invoked whenever they are encountered in the Slim template.

div:componentName#id.class (attributes)

or more simply:

componentName#id.class (attributes)

IoC

Slim makes use of the non-invasive module feature of IoC 3.

When Slim is added as a dependency to an IoC enabled application, such as BedSheet or Reflux, then:

  • Slim becomes an injectable service
  • efan Xtra becomes aware of, and can parse, slim templates and .slim files.