Slim is a template language whose goal is to reduce the view syntax to the essential parts without becoming cryptic. It started as an exercise to see how much could be removed from a standard html template (<, >, closing tags, etc...). As more people took an interest in Slim, the functionality grew and so did the flexibility of the syntax.
A short list of the features...
- Short syntax without closing tags (Using indentation instead)
- Embedded engines like Markdown and Textile
- Configurable shortcut tags (
#fordiv idand.fordiv classin the default configuration) - Automatic HTML escaping and support for Rails'
html_safe? - HTML style mode with closing tags
- Logic less mode similar to Mustache, realized as plugin
- Translator/I18n, realized as plugin
- Highly configurable and extendable
- High performance (Comparable to ERB)
- Supported by all major frameworks (Rails, Sinatra, ...)
- Streaming support in Rails
Slim is a fast, lightweight templating engine with support for Rails 3. It has been heavily tested on all major ruby implementations. We use continous integration (travis-ci).
Slim's core syntax is guided by one thought: "What's the minimum required to make this work".
As more people have contributed to Slim, there have been syntax additions influenced from their use of Haml and Jade. The Slim team is open to these additions because we know beauty is in the eye of the beholder.
Slim uses Temple for parsing/compilation and is also integrated into Tilt, so it can be used together with Sinatra or plain Rack.
The architecture of Temple is very flexible and allows the extension of the parsing and compilation process without monkey-patching. This is used by the logic less plugin and the translator plugin which provides I18n.
Within the Rails community, Erb and Haml are without doubt the two most popular templating engines. However, Erb's syntax is cumbersome and Haml's syntax can be quite cryptic to the uninitiated.
Slim was born to bring a minimalist syntax approach with speed. If people chose not to use Slim, it would not be because of speed.
Yes, Slim is speedy! Benchmarks are provided at the end of this README file. Don't trust the numbers? That's as it should be. Therefore we provide a benchmark rake task so you could test it yourself (rake bench).
Install Slim as a gem:
gem install slim
Include Slim in your Gemfile with gem 'slim' or require it with require 'slim'. That's it! Now, just use the .slim extension and you're good to go.
Here's a quick example to demonstrate what a Slim template looks like:
doctype html
html
head
title Slim Examples
meta name="keywords" content="template language"
meta name="author" content=author
link rel="icon" type="image/png" href=file_path("favicon.png")
javascript:
alert('Slim supports embedded javascript!')
body
h1 Markup examples
#content
p This example shows you how a basic Slim file looks like.
= yield
- if items.any?
table#items
- for item in items do
tr
td.name = item.name
td.price = item.price
- else
p No items found Please add some inventory.
Thank you!
div id="footer"
= render 'footer'
| Copyright © #{year} #{author}
Indentation matters, but the indentation depth can be chosen as you like. If you want to first indent 2 spaces, then 5 spaces, it's your choice. To nest markup you only need to indent by one space, the rest is gravy.
The pipe tells Slim to just copy the line. It essentially escapes any processing. Each following line that is indented greater than the backtick is copied over.
body
p
|
This is a test of the text block.
The parsed result of the above:
<body><p>This is a test of the text block.</p></body>
The left margin is set at the indent of the backtick + one space. Any additional spaces will be copied over.
body
p
| This line is on the left margin.
This line will have one space in front of it.
This line will have two spaces in front of it.
And so on...
You can also embed html in the text line
- articles.each do |a|
| <tr><td>#{a.name}</td><td>#{a.description}</td></tr>
The single quote tells Slim to copy the line (similar to |), but makes sure that a single trailing space is appended.
You can write html tags directly in Slim which allows you to write your templates in a more html like style with closing tags or mix html and Slim style.
<html>
head
title Example
<body>
- if articles.empty?
- else
table
- articles.each do |a|
<tr><td>#{a.name}</td><td>#{a.description}</td></tr>
</body>
</html>
The dash denotes control code. Examples of control code are loops and conditionals. end is forbidden behind -. Blocks are defined only by indentation.
If your ruby code needs to use multiple lines, append a backslash \ at the end of the lines.
body
- if articles.empty?
| No inventory
The equal sign tells Slim it's a Ruby call that produces output to add to the buffer. If your ruby code needs to use multiple lines, append a backslash \ at the end of the lines, for example:
= javascript_include_tag \
"jquery", \
"application"
Same as the single equal sign (=), except that it adds a trailing whitespace.
Same as the single equal sign (=), but does not go through the escape_html method.
Same as the double equal sign (==), except that it adds a trailing whitespace.
Use the forward slash for code comments - anything after it won't get displayed in the final render. Use / for code comments and /! for html comments
body
p
/ This line won't get displayed.
Neither does this line.
/! This will get displayed as html comments.
The parsed result of the above:
<body><p><!--This will get displayed as html comments.--></p></body>
Use the forward slash immediately followed by an exclamation mark for html comments (<!-- ... -->).
/[if IE]
p Get a better browser.
renders as
<!--[if IE]><p>Get a better browser.</p><![endif]-->
The doctype tag is a special tag which can be used to generate the complex doctypes in a very simple way.
XML VERSION
doctype xml
<?xml version="1.0" encoding="utf-8" ?>
doctype xml ISO-8859-1
<?xml version="1.0" encoding="iso-8859-1" ?>
XHTML DOCTYPES
doctype html
<!DOCTYPE html>
doctype 5
<!DOCTYPE html>
doctype 1.1
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
"http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
doctype strict
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
doctype frameset
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Frameset//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-frameset.dtd">
doctype mobile
<!DOCTYPE html PUBLIC "-//WAPFORUM//DTD XHTML Mobile 1.2//EN"
"http://www.openmobilealliance.org/tech/DTD/xhtml-mobile12.dtd">
doctype basic
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML Basic 1.1//EN"
"http://www.w3.org/TR/xhtml-basic/xhtml-basic11.dtd">
doctype transitional
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
HTML 4 DOCTYPES
doctype strict
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN"
"http://www.w3.org/TR/html4/strict.dtd">
doctype frameset
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Frameset//EN"
"http://www.w3.org/TR/html4/frameset.dtd">
doctype transitional
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
"http://www.w3.org/TR/html4/loose.dtd">
You can close tags explicitly by appending a trailing /.
img src="image.png"/
Note, that this is usually not necessary since the standard html tags (img, br, ...) are closed automatically.
Sometimes you may want to be a little more compact and inline the tags.
ul
li.first: a href="/a" A link
li: a href="/b" B link
For readability, don't forget you can wrap the attributes.
ul
li.first: a[href="/a"] A link
li: a[href="/b"] B link
Either start on the same line as the tag
body
h1 id="headline" Welcome to my site.
Or nest it. You must use a pipe or a backtick to escape processing
body
h1 id="headline"
| Welcome to my site.
Can make the call on the same line
body
h1 id="headline" = page_headline
Or nest it.
body
h1 id="headline"
= page_headline
You write attributes directly after the tag. For normal text attributes you must use double " or single quotes ' (Quoted attributes).
a href="http://slim-lang.com" title='Slim Homepage' Goto the Slim homepage
You can use text interpolation in the quoted attributes.
If a delimiter makes the syntax more readable for you,
you can use the characters {...}, (...), [...] to wrap the attributes.
body
h1(id="logo") = page_logo
h2[id="tagline" class="small tagline"] = page_tagline
If you wrap the attributes, you can spread them across multiple lines:
h2[id="tagline"
class="small tagline"] = page_tagline
Example:
a href="http://slim-lang.com" title='Slim Homepage' Goto the Slim homepage
You can use text interpolation in the quoted attributes:
a href="http://#{url}" Goto the #{url}
The attribute value will be escaped by default. Use == if you want to disable escaping in the attribute.
a href=="&"
Write the ruby code directly after the =. If the code contains spaces you have to wrap
the code into parentheses (...), {...} or [...]. The code in the parentheses will be evaluated.
body
table
- for user in users do
td id="user_#{user.id}" class=user.role
a href=user_action(user, :edit) Edit #{user.name}
a href={path_to_user user} = user.name
The attribute value will be escaped by default. Use == if you want to disable escaping in the attribute.
a href==action_path(:start)
The attribute values true, false and nil are interpreted
as booleans. If you use the attribut wrapper you can omit the attribute assigment
input type="text" disabled="disabled"
input type="text" disabled=true
input(type="text" disabled)
input type="text"
input type="text" disabled=false
input type="text" disabled=nil
The splat shortcut allows you turn a hash in to attribute/value pairs
.card*{'data-url'=>place_path(place), 'data-id'=>place.id} = place.name
.card *method_which_returns_hash = place.name
renders as
<div class="card" data-id="1234" data-url="/place/1234">Slim's house</div>
Similarly to Haml, you can specify the id and class attributes in the following shortcut form
body
h1#headline
= page_headline
h2#tagline.small.tagline
= page_tagline
.content
= show_content
This is the same as
body
h1 id="headline"
= page_headline
h2 id="tagline" class="small tagline"
= page_tagline
div class="content"
= show_content
You can define custom shortcuts (Similar to # for id and . for class).
In this example we add @ to create a shortcut for the role attribute.
Slim::Engine.set_default_options :shortcut => {'@' => 'role', '#' => 'id', '.' => 'class'}
We can use it in Slim code like this
.person@admin = person.name
which renders to
<div class="person" role="admin">Daniel</div>
Use standard Ruby interpolation. The text will be html escaped by default.
body
h1 Welcome #{current_user.name} to the show.
| Unescaped #{{content}} is also possible.
To escape the interpolation (i.e. render as is)
body
h1 Welcome \#{current_user.name} to the show.
Thanks to Tilt, Slim has impressive support for embedding other template engines.
Examples:
coffee:
square = (x) -> x * x
markdown:
#Header
Hello from #{"Markdown!"}
Second Line!
Supported engines:
| Filter | Required gems | Type | Description |
| ruby: | none | Shortcut | Shortcut to embed ruby code |
| javascript: | none | Shortcut | Shortcut to embed javascript code and wrap in script tag |
| css: | none | Shortcut | Shortcut to embed css code and wrap in style tag |
| sass: | sass | Compile time | Embed sass code and wrap in style tag |
| scss: | sass | Compile time | Embedd scss code and wrap in style tag |
| less: | less | Compile time | Embed less css code and wrap in style tag |
| styl: | styl | Compile time | Embed stylus css code and wrap in style tag |
| coffee: | coffee-script | Compile time | Compile coffee script code and wrap in script tag |
| markdown: | redcarpet/rdiscount/kramdown | Compile time + Interpolation | Compile markdown code and interpolate #\{variables} in text |
| textile: | redcloth | Compile time + Interpolation | Compile textile code and interpolate #\{variables} in text |
| creole: | creole | Compile time + Interpolation | Compile creole code and interpolate #\{variables} in text |
| wiki:, mediawiki: | wikicloth | Compile time + Interpolation | Compile wiki code and interpolate #\{variables} in text |
| rdoc: | rdoc | Compile time + Interpolation | Compile rdoc code and interpolate #\{variables} in text |
| builder: | builder | Precompiled | Embed builder code |
| nokogiri: | nokogiri | Precompiled | Embed nokogiri builder code |
| erb: | none | Precompiled | Embed erb code |
The embedded engines can be configured in Slim by setting the options directly on the Slim::EmbeddedEngine filter. Example:
Slim::EmbeddedEngine.default_options[:markdown] = {:auto_ids => false}
Slim and the underlying Temple framework are highly configurable. Unfortunately the way how you configure Slim depends on the compilation mechanism (Rails or Tilt). It is always possible to set default options. This can be done in Rails' environment files. For instance, in config/environments/development.rb you probably want:
# Indent html for pretty debugging and do not sort attributes (Ruby 1.8)
Slim::Engine.set_default_options :pretty => true, :sort_attrs => false
# Indent html for pretty debugging and do not sort attributes (Ruby 1.9)
Slim::Engine.set_default_options pretty: true, sort_attrs: false
You can also access the option hash directly:
Slim::Engine.default_options[:pretty] = true
For developers who know more about Slim and Temple architecture it is possible to override default options at different positions. Temple uses an inheritance mechanism to allow subclasses to override options of the superclass. The option priorities are as follows:
Options passed at engine instantination > Slim::Template > Slim::Engine > Parser/Filter/Generator (e.g Slim::Parser, Slim::Compiler)
It is also possible to set options for superclasses like Temple::Engine. But this will affect all temple template engines then.
Slim::Engine > Temple::Engine
Slim::Compiler > Temple::Filter
The following options are exposed by the Slim::Engine and can be set with Slim::Engine.set_default_options.
There are a lot of them but the good thing is, that Slim checks the configuration keys and reports an error if you try to use an invalid configuration key.
| Type | Name | Default | Purpose |
| String | :file | nil | Name of parsed file, set automatically by Slim::Template |
| Integer | :tabsize | 4 | Number of whitespaces per tab (used by the parser) |
| String | :encoding | "utf-8" | Set encoding of template |
| String | :default_tag | "div" | Default tag to be used if tag name is omitted |
| Hash | :shortcut | \{'.' => 'class', '#' => 'id'} | Attribute shortcuts |
| Symbol list | :enable_engines | nil (All enabled) | List of enabled embedded engines (whitelist) |
| Symbol list | :disable_engines | nil (None disabled) | List of disabled embedded engines (blacklist) |
| Boolean | :disable_capture | false (true in Rails) | Disable capturing in blocks (blocks write to the default buffer |
| Boolean | :disable_escape | false | Disable automatic escaping of strings |
| Boolean | :escape_quoted_attrs | false | Escape quoted attributes |
| Boolean | :use_html_safe | false (true in Rails) | Use String#html_safe? from ActiveSupport (Works together with :disable_escape) |
| Symbol | :format | :xhtml | HTML output format (Possible formats :xhtml, :html4, :html5, :html) |
| String | :attr_wrapper | '"' | Character to wrap attributes in html (can be ' or ") |
| Hash | :attr_delimiter | \{'class' => ' '} | Joining character used if multiple html attributes are supplied (e.g. class="class1 class2") |
| Boolean | :sort_attrs | true | Sort attributes by name |
| Boolean | :pretty | false | Pretty html indenting (This is slower!) |
| String | :indent | ' ' | Indentation string |
| Boolean | :streaming | false (true in Rails > 3.1) | Enable output streaming |
| Class | :generator | Temple::Generators::ArrayBuffer/RailsOutputBuffer | Temple code generator (default generator generates array buffer) |
Additionally the code generator options can be set (used by the :generator class). The standard generators support the options :buffer and :capture_generator. See the code generator code documentation for details on that (Temple::Generators).
There are more options which are supported by the Temple filters but which are not exposed and are not officially supported. You have to take a look at the Slim and Temple code for that.
Logic less mode is inspired by Mustache. Logic less mode uses a dictionary object e.g. a recursive hash tree which contains the dynamic content.
If the object is not false or empty?, the content will show
- article
h1 = title
If the object is false or empty?, the content will show
-! article
p Sorry, article not found
If the object is an array, the section will iterate
- articles
tr: td = title
Example code:
- article
h1 = title
In wrapped dictionary acccess mode (the default, see the options), the dictionary object is accessed in the following order.
- If article.respond_to?(:title), Slim will execute article.send(:title)
- If article.respond_to?(:has_key?) and article.has_key?(:title), Slim will execute article[:title]
- If article.instance_variable_defined?(@title),Slim will execute article.instance_variable_get @title
If all the above fails, Slim will try to resolve the title reference in the same order against the parent object. In this example, the parent would be the dictionary object you are rendering the template against.
As you might have guessed, the article reference goes through the same steps against the dictionary. Instance variables are not allowed in the view code, but Slim will find and use them. Essentially, you're just using dropping the @ prefix in your template. Parameterized method calls are not allowed.
Install:
$ gem install slim
Require:
gem 'slim', :require => 'slim/logic_less'
Sinata has built-in support for Slim. All you have to do is require the logic less Slim plugin. This can be done in your config.ru:
require 'slim/logic_less'
You are then ready to rock!
| Type | Name | Default | Purpose |
| Boolean | :logic_less | true | Enable logic less mode (Enabled if 'slim/logic_less' is required) |
| String | :dictionary | "self" | Dictionary where variables are looked up |
| Symbol | :dictionary_access | :wrapped | Dictionary access mode (:string, :symbol, :wrapped) |
The translator plugin provides automatic translation of the templates using Gettext, Fast-Gettext or Rails I18n. Static text in the template is replaced by the translated version.
Example:
h1 Welcome to #{url}!
Gettext translates the string from english to german where interpolations are replaced by %1, %2, ...
"Welcome to %1!" -> "Willkommen auf %1!"
and renders as
<h1>Willkommen auf slim-lang.com!</h1>
Enable the translator plugin with
require 'slim/translator'
| Type | Name | Default | Purpose |
| Boolean | :tr | true | Enable translator (Enabled if 'slim/translator' is required) |
| Symbol | :tr_mode | :dynamic | When to translate: :static = at compile time, :dynamic = at runtime |
| String | :tr_fn | Depending on installed translation library | Translation function, could be '_' for gettext |
Slim uses Tilt to compile the generated code. If you want to use the Slim template directly, you can use the Tilt interface.
Tilt.new['template.slim'].render(scope)
Slim::Template.new('template.slim', optional_option_hash).render(scope)
Slim::Template.new(optional_option_hash) { source }.render(scope)
The optional option hash can have to options which were documented in the section above.
require 'sinatra'
require 'slim'
get('/') { slim :index }
__END__
@@ index
doctype html
html
head
title Sinatra With Slim
body
h1 Slim Is Fun!
Rails generators are provided by slim-rails. slim-rails
is not necessary to use Slim in Rails though. Just install Slim and add it to your Gemfile with gem 'slim'.
Then just use the .slim extension and you're good to go.
HTTP streaming is enabled enabled by default if you use a Rails version which supports it.
The gem 'slim' comes with the small tool 'slimrb' to test Slim from the command line.
$ slimrb --help
Usage: slimrb [options]
-s, --stdin Read input from standard input instead of an input file
--trace Show a full traceback on error
-c, --compile Compile only but do not run
-r, --rails Generate rails compatible code (Implies --compile)
-t, --translator Enable translator plugin
-l, --logic-less Enable logic less plugin
-p, --pretty Produce pretty html
-o, --option [NAME=CODE] Set slim option
-h, --help Show this message
-v, --version Print version
Start 'slimrb', type your code and press Ctrl-d to send EOF. Example usage:
$ slimrb markdown: First paragraph. Second paragraph. * one * two * three //Enter Ctrl-d <p>First paragraph </p> <p>Second paragraph </p> <ul> <li>one</li> <li>two</li> <li>three</li> </ul>
There are plugins for various text editors (including the most important ones - Vim, Emacs and Textmate):
The benchmarks demonstrate that Slim in production mode is nearly as fast as Erubis (which is the fastest template engine). So if you choose not to use Slim it is not due to its speed.
Run the benchmarks with rake bench. You can add the option slow to
run the slow parsing benchmark which needs more time. You can also increase the number of iterations.
rake bench slow=1 iterations=1000
Linux + Ruby 1.9.3, 1000 iterations
user system total real
(1) erb 0.020000 0.000000 0.020000 ( 0.017383)
(1) erubis 0.020000 0.000000 0.020000 ( 0.015048)
(1) fast erubis 0.020000 0.000000 0.020000 ( 0.015372) <===
(1) temple erb 0.030000 0.000000 0.030000 ( 0.026239)
(1) slim pretty 0.030000 0.000000 0.030000 ( 0.031463)
(1) slim ugly 0.020000 0.000000 0.020000 ( 0.018868) <===
(1) haml pretty 0.130000 0.000000 0.130000 ( 0.122521)
(1) haml ugly 0.110000 0.000000 0.110000 ( 0.106640)
(2) erb 0.030000 0.000000 0.030000 ( 0.035520)
(2) erubis 0.020000 0.000000 0.020000 ( 0.023070)
(2) temple erb 0.040000 0.000000 0.040000 ( 0.036514)
(2) slim pretty 0.040000 0.000000 0.040000 ( 0.040086)
(2) slim ugly 0.030000 0.000000 0.030000 ( 0.028461)
(2) haml pretty 0.150000 0.000000 0.150000 ( 0.145618)
(2) haml ugly 0.130000 0.000000 0.130000 ( 0.129492)
(3) erb 0.140000 0.000000 0.140000 ( 0.134953)
(3) erubis 0.120000 0.000000 0.120000 ( 0.119723)
(3) fast erubis 0.100000 0.000000 0.100000 ( 0.097456)
(3) temple erb 0.040000 0.000000 0.040000 ( 0.035916)
(3) slim pretty 0.040000 0.000000 0.040000 ( 0.039626)
(3) slim ugly 0.030000 0.000000 0.030000 ( 0.027827)
(3) haml pretty 0.310000 0.000000 0.310000 ( 0.306664)
(3) haml ugly 0.250000 0.000000 0.250000 ( 0.248742)
(4) erb 0.350000 0.000000 0.350000 ( 0.350719)
(4) erubis 0.310000 0.000000 0.310000 ( 0.304832)
(4) fast erubis 0.300000 0.000000 0.300000 ( 0.303070)
(4) temple erb 0.910000 0.000000 0.910000 ( 0.911745)
(4) slim pretty 3.410000 0.000000 3.410000 ( 3.413267)
(4) slim ugly 2.880000 0.000000 2.880000 ( 2.885265)
(4) haml pretty 2.280000 0.000000 2.280000 ( 2.292623)
(4) haml ugly 2.170000 0.000000 2.170000 ( 2.169292)
(1) Compiled benchmark. Template is parsed before the benchmark and
generated ruby code is compiled into a method.
This is the fastest evaluation strategy because it benchmarks
pure execution speed of the generated ruby code.
(2) Compiled Tilt benchmark. Template is compiled with Tilt, which gives a more
accurate result of the performance in production mode in frameworks like
Sinatra, Ramaze and Camping. (Rails still uses its own template
compilation.)
(3) Cached benchmark. Template is parsed before the benchmark.
The ruby code generated by the template engine might be evaluated every time.
This benchmark uses the standard API of the template engine.
(4) Parsing benchmark. Template is parsed every time.
This is not the recommended way to use the template engine
and Slim is not optimized for it. Activate this benchmark with 'rake bench slow=1'.
Temple ERB is the ERB implementation using the Temple framework. It shows the
overhead added by the Temple framework compared to ERB.
Slim provides an extensive test-suite based on minitest. You can run the tests with 'rake test' and the rails integration tests with 'rake test:rails'.
Travis-CI is used for continous integration testing: {http://travis-ci.org/#!/stonean/slim}
Slim is working well on all major Ruby implementations:
- Ruby 1.8.7
- Ruby 1.9.2
- Ruby 1.9.3
- Ruby EE
- JRuby
- Rubinius 2.0
If you'd like to help improve Slim, clone the project with Git by running:
$ git clone git://github.com/stonean/slim
Work your magic and then submit a pull request. We love pull requests!
Please remember to test against Ruby versions 1.9.2 and 1.8.7.
If you find the documentation lacking (and you probably will), help us out The docs are located in the gh-pages branch:
$ git checkout gh-pages
If you don't have the time to work on Slim, but found something we should know about, please submit an issue.
Slim is released under the MIT license.
- Google Group
- IRC Channel #slim-lang on freenode.net
Template compilation framework:
Framework support:
Syntax highlighting:
Template Converters (HAML, ERB, ...):
Language ports/Similar languages: