Current section

Files

Jump to
phoenix_html lib phoenix_html link.ex
Raw

lib/phoenix_html/link.ex

defmodule Phoenix.HTML.Link do
@moduledoc """
Conveniences for working with links and URLs in HTML.
"""
import Phoenix.HTML.Tag
@doc """
Generates a link to the given URL.
## Examples
link("hello", to: "/world")
#=> <a href="/world">hello</a>
link("<hello>", to: "/world")
#=> <a href="/world">&lt;hello&gt;</a>
link("<hello>", to: "/world", class: "btn")
#=> <a class="btn" href="/world">&lt;hello&gt;</a>
link("delete", to: "/the_world", data: [confirm: "Really?"])
#=> <a data-confirm="Really?" href="/the_world">delete</a>
# If you supply a method other than `:get`:
link("delete", to: "/everything", method: :delete)
#=> <a href="#" data-csrf="csrf_token" data-method="delete" data-to="/everything">delete</a>
# You can use a `do ... end` block too:
link to: "/hello" do
"world"
end
## Options
* `:to` - the page to link to. This option is required
* `:method` - the method to use with the link. In case the
method is not `:get`, the link is generated inside the form
which sets the proper information. In order to submit the
form, JavaScript must be enabled
All other options are forwarded to the underlying `<a>` tag.
## Data attributes
Data attributes are added as a keyword list passed to the
`data` key. The following data attributes are supported:
* `data-confirm` - shows a confirmation prompt before
generating and submitting the form when `:method`
is not `:get`.
## JavaScript dependency
In order to support links where `:method` is not `:get`
or use the above data attributes, `Phoenix.HTML` relies
on JavaScript. You can load `priv/static/phoenix_html.js`
into your build tool.
## CSRF Protection
By default, CSRF tokens are generated through `Plug.CSRFProtection`.
"""
@valid_uri_schemes [
"http:",
"https:",
"ftp:",
"ftps:",
"mailto:",
"news:",
"irc:",
"gopher:",
"nntp:",
"feed:",
"telnet:",
"mms:",
"rtsp:",
"svn:",
"tel:",
"fax:",
"xmpp:"
]
def link(text, opts)
def link(opts, do: contents) when is_list(opts) do
link(contents, opts)
end
def link(_text, opts) when not is_list(opts) do
raise ArgumentError, "link/2 requires a keyword list as second argument"
end
def link(text, opts) do
{to, opts} = pop_required_option!(opts, :to, "expected non-nil value for :to in link/2")
to = valid_destination!(to, "link/2")
{method, opts} = Keyword.pop(opts, :method, :get)
if method == :get do
opts = skip_csrf(opts)
content_tag(:a, text, [href: to] ++ opts)
else
{csrf_data, opts} = csrf_data(to, opts)
opts = Keyword.put_new(opts, :rel, "nofollow")
content_tag(:a, text, [href: "#", data: [method: method, to: to] ++ csrf_data] ++ opts)
end
end
@doc false
# No docs since this function is only called when a `do` block is passed as
# `do:` instead of `do...end` (and that case is documented in `link/2`).
def link(opts) when is_list(opts) do
error = "link/2 requires a text as first argument or contents in the :do block"
{contents, opts} = pop_required_option!(opts, :do, error)
link(contents, opts)
end
@doc """
Generates a button that uses a regular HTML form to submit to the given URL.
Useful to ensure that links that change data are not triggered by
search engines and other spidering software.
## Examples
button("hello", to: "/world")
#=> <button class="button" data-csrf="csrf_token" data-method="post" data-to="/world">hello</button>
button("hello", to: "/world", method: "get", class: "btn")
#=> <button class="btn" data-method="get" data-to="/world">hello</button>
## Options
* `:to` - the page to link to. This option is required
* `:method` - the method to use with the button. Defaults to :post.
All other options are forwarded to the underlying button input.
"""
def button(opts, do: contents) do
button(contents, opts)
end
def button(text, opts) do
{to, opts} = pop_required_option!(opts, :to, "option :to is required in button/2")
{method, opts} = Keyword.pop(opts, :method, :post)
to = valid_destination!(to, "button/2")
if method == :get do
opts = skip_csrf(opts)
content_tag(:button, text, [data: [method: method, to: to]] ++ opts)
else
{csrf_data, opts} = csrf_data(to, opts)
content_tag(:button, text, [data: [method: method, to: to] ++ csrf_data] ++ opts)
end
end
defp skip_csrf(opts) do
Keyword.delete(opts, :csrf_token)
end
defp csrf_data(to, opts) do
case Keyword.pop(opts, :csrf_token, true) do
{csrf, opts} when is_binary(csrf) ->
{[csrf: csrf], opts}
{true, opts} ->
{[csrf: Plug.CSRFProtection.get_csrf_token_for(to)], opts}
{false, opts} ->
{[], opts}
end
end
defp pop_required_option!(opts, key, error_message) do
{value, opts} = Keyword.pop(opts, key)
unless value do
raise ArgumentError, error_message
end
{value, opts}
end
defp valid_destination!({:safe, to}, context) do
{:safe, valid_string_destination!(IO.iodata_to_binary(to), context)}
end
defp valid_destination!({other, to}, _context) when is_atom(other) do
[Atom.to_string(other), ?:, to]
end
defp valid_destination!(to, context) do
valid_string_destination!(IO.iodata_to_binary(to), context)
end
for scheme <- @valid_uri_schemes do
defp valid_string_destination!(unquote(scheme) <> _ = string, _context), do: string
end
defp valid_string_destination!(to, context) do
if String.contains?(to, ":") do
raise ArgumentError, """
unsupported scheme given to #{context}. In case you want to link to an
unknown or unsafe scheme, such as javascript, use a tuple: {:javascript, rest}
"""
else
to
end
end
end