* request.formats is much simpler now
* For XHRs or Accept headers with a single item, we use the Accept header
* For other requests, we use params[:format] or fallback to HTML
* This is primarily to work around the fact that browsers provide completely
broken Accept headers, so we have to whitelist the few cases we can
specifically isolate and treat other requests as coming from the browser
* For APIs, we can support single-item Accept headers, which disambiguates
from the browsers
* Requests to an action that only has an XML template from the browser will
no longer find the template. This worked previously because most browsers
provide a catch-all */*, but this was mostly accidental behavior. If you
want to serve XML, either use the :xml format in links, or explicitly
specify the XML template: render "template.xml".
* only one of partial_name or :as will be available as a local
* `object` is removed
* Simplify _layout_for in most cases.
* Remove <% render :partial do |args| %>
* <% render :partial do %> still works fine
* Call _evaluate_assigns_and_ivars at the two entry points so we don't have to
do a check at every render.
* Make template.render viable without having to go through a wrapper method
* Remove old TemplateHandler#render(template, local_assigns) path so we don't have
to set self.template every time we render a template.
* Move Template rescuing code to Template#render so it gets caught every time.
* Pull in some tests from Pratik that test render @object in ActionView
* Move #set_cookie and #delete_cookie inline to optimize. These optimizations should
almost certainly be sent back upstream to Rack. The optimization involves using
an ivar for cookies instead of indexing into the headers each time.
* Was able to use a bare Hash for headers now that cookies have their own joining
semantics (some code assumed that the raw cookies were an Array).
* Cache blankness of body on body=
* Improve expand_cache_key for Arrays of a single element (common in our case)
* Use a simple layout condition check unless conditions are used
* Cache visible actions
* Lazily load the UrlRewriter
* Make etag an ivar that is set on prepare!
* Tentatively replaced HeaderHash with SimpleHeaderHash, which does not preserve
case but does handle converting Arrays to Strings in to_hash. This requires
further discussion.
* Moved default_charset to ActionDispatch::Response to avoid having to hop over
to ActionController. Ideally, this would be a constant on AD::Response, but
some tests expect to be able to change it dynamically and I didn't want to change
them yet.
* Completely override #initialize from Rack::Response. Previously, it was creating
a HeaderHash, and then we were creating an entirely new one. There is no way to
call super without incurring the overhead of creating a HeaderHash.
* Override #write from Rack::Response. Its implementation tracks Content-Length,
and doing so adds additional overhead that could be mooted if other middleware
changes the body. It is more efficiently done at the top-level server.
* Change sending_file to an instance_variable instead of header inspection. In
general, if a state is important, it should be set as a property of the response
not reconstructed later.
* Set the Etag to @body instead of .body. AS::Cache.expand_cache_key handles
Arrays fine, and it's more efficient to let it handle the body parts, since
it is not forced to create a joined String.
* If we detect the default cache control case, just set it, rather than setting
the constituent parts and then running the normal (expensive) code to generate
the string.
To discuss: What are the desired semantics (if any) for layouts in a collection. There are no
tests for it at present, and I'm not sure if it's needed at all.
Deprecated on this branch: `object` pointing at the current object in partials. You can still
use the partial name, or use :as to achieve the same thing. This is obviously up for discussion.
The offset in codepoints was being passed directly to the wrapped string's index/rindex method. Now we translate the offset into bytes first.
[#3028 state:committed]
Signed-off-by: Jeremy Kemper <jeremy@bitsweat.net>
