Skip to content
Commits on Sep 8, 2015
  1. @ggrossman

    When parsing the name parameter of Content-Disposition, support quote…

    ggrossman committed
    …d chars in the quoted-string case.
Commits on Sep 6, 2015
  1. @ggrossman

    Fix bug in parsing of Content-Disposition header where an unquoted na…

    ggrossman committed
    …me at end-of-line sucked in the trailing newline.
Commits on Sep 5, 2015
  1. @tenderlove

    pull env access in the request object to a module

    tenderlove committed
    this also tests that delegation to the request object is possible.  I
    want to see exactly how many methods we need to delegate in order to
    look like a real request object
Commits on Sep 4, 2015
  1. @tenderlove

    Move most methods on the `Rack::Request` to a module

    tenderlove committed
    Move most request methods to `Rack::Request::Helpers` and use public API
    to get values from the request object.  This enables users to mix
    `Rack::Request::Helpers` in to their own objects so they can implement
    `(get|set|fetch|each)_header` as they see fit (for example a proxy
    object).
Commits on Aug 28, 2015
  1. @tenderlove
  2. @tenderlove

    pull the tempfile factory default up to a constant

    tenderlove committed
    we don't need to create a new lambda object every time we do parsing
  3. @tenderlove

    remove the env from the multipart parser

    tenderlove committed
    I want to abstract the multipart parser from `env` so that we can get
    the data from some structure other than an env hash.
  4. @tenderlove

    wrap bounded IO objects

    tenderlove committed
    If we have a content length, wrap the IO object with a new object that
    knows about the content length and will act like an IO object with the
    length specified
Commits on Jul 18, 2015
  1. @matthewd
  2. @matthewd
  3. @sigmavirus24 @matthewd

    Omit unnecessary ?s in EXTENDED_INITIAL_VALUE

    sigmavirus24 committed with matthewd
    The ? outside the groups is redundant and Ruby's regular expression engine
    rightfully produces a warning when compiling the regular expression.
    
    Removing the ?s prevents the warning from being printed to stderr.
    
    See also: http://rubular.com/r/J4CWeIHQTK
Commits on Jun 13, 2015
  1. @burtlo @sigmavirus24

    Feature: Provided support for non-ascii character in public header

    burtlo committed with sigmavirus24
    I love that we worked on this @MadisonRuby for #OSL
  2. @sigmavirus24
  3. @sigmavirus24

    Move RFC2183 constant too

    sigmavirus24 committed
  4. @sigmavirus24
Commits on Jun 9, 2015
  1. @tenderlove

    consolidate where to find the default_query_parser make keyspace requ…

    tenderlove committed
    …ired
    
    we should only be able to find the default_query_parser in one place,
    Utils.  Also the keyspace size should be a require parameter
  2. @tenderlove

    ask the query parser to construct a params object

    tenderlove committed
    this decouples the rest of the code from knowing how a params object is
    actually constructed
Commits on May 29, 2015
  1. @jeremyevans

    Allow specifying a params hash class to use when parsing

    jeremyevans committed
    This allows you to subclass the query parser to change the
    behavior.  The main reason behind this is to allow changing
    the hash used to store params, so that indifferent access to
    params can be done efficiently without monkey patching.
    
    This adds Request#query_parser which is hard coded to use
    the default query parser (QueryParser::Default).  However,
    subclasses of Request can override this method to use an instance
    of a subclass of QueryParser for custom behavior.
    
    For backwards compatibility, Utils still defines the same
    methods as before, but now they delegate to the default
    query parser.
    
    The multipart code should use the same params class used by the
    query parser for the request, so the relevant APIs have been
    changed to support passing in the params to use as an argument.
Commits on Jul 18, 2014
  1. @raggi

    multipart content-type match now case insensitive

    raggi committed
    Closes #676
    
    This may not be 100% to spec, but it is important for compatibility and should
    not cause regressions.
Commits on Nov 27, 2013
  1. @tenderlove
Commits on Jan 4, 2013
  1. @raggi

    Fix parsing performance for unquoted filenames

    raggi committed with raggi
    Special thanks to Paul Rogers & Eric Wong
    
    Conflicts:
    	test/spec_multipart.rb
Commits on May 3, 2011
  1. @raggi
  2. @shwoodard

    refactor of multipart module

    shwoodard committed
Something went wrong with that request. Please try again.