Checking with to_s. As regexp fail with 1.8.7

Don't do remove_possible_method when delegate is used. Two reasons: 1) warnings should be shown, and fixed at the source and 2) the code is slow. Fixes #1937.

Fix a wrong assertion on url_helper_test, and add missing `#html_safe?` a

Render partial invalid check

Revert "Fix failure in test_preserves_existing_fixture_data from test/cases/fixtures_test.rb when UTC and local time occur on different dates." I am pretty sure this was an incorrect fix, and it still failed in certain circumstances anyway. I am now unable to reproduce the original failure I was experiencing so will leave it for now and see if this pops up again.

This reverts commit e4479b2f.

As we are adding db folder in it. Required in test also.

Better formatting of rake routes (issue #1991)

This route:

    match 'foo/:id' => RackApp, :id => /[A-Z]\d{5}/

is shown as:

    /foo/:id(.:format) RackApp {:id=>/[A-Z]\d{5}/}

Adds gem helper rake tasks to engines generator

Added db to gemspec so migrations are packaged

Fix wildcard route code example in changelog

Previously it was:

    {:controller=>"photos", :action=>"show", :id=>/[A-Z]\d{5}/}

Now it becomes:

    photos#show {:id=>/[A-Z]\d{5}/}

ported test from pull request #2071 to master

Patch for https://github.com/rails/rails/issues/2059

Cache key was incorrectly using timezone-dependent record#updated_at when it should be using a timezone-independent value to generate the cache key

Minor refactoring to cache_key timezone test

Closes #2059

Adds a test to validate the format of the cache_key for nil and present updated_at values

Correctly handles updated_at == nil

Bugfix by stopping find_in_batches using the records after yielding.

missing magic comments

treat fullwidth whitespace as a blank character

assets.cache_store now defaults to filesystem. You will want to share tmp/assets_cache between deploys.

Fix for SqlBypass session store (for master)

Fixed session ID fixation for ActiveRecord::SessionStore (for master)

remove extra space since comment_if already returns the space.

Two issues fixed:
1) connection_pool is not defined - needed by SessionStore#drop_table!
and create_table! since c94651f8

2) initialization of connection to the default of AR::Base.connection
only occurred at the singleton level - the instance level method defined
by cattr_accessor did not have this logic

I have found that Rails will take an invalid session ID specified by the
client and materialize a session based on that session ID. This means
that it is possible, among other things, for a client to use an
arbitrarily weak session ID or for a client to resurrect a previous used
session ID. In other words, we cannot guarantee that all session IDs are
generated by the server and that they are (statistically) unique through
time.

The fix is to always generate a new session ID in #get_session if an
existing session cannot be found under the incoming session ID.

Also added new tests that make sure that an invalid session ID is never
materialized into a new session, regardless of whether it comes in via a
cookie or a URL parameter (when :cookie_only => false).