3581 Commits

Author SHA1 Message Date
Leonardo Tegon
2e5b5fcd70 Prepare for 4.6.2 release v4.6.2 2019-03-26 13:26:56 -03:00
Leonardo Tegon
241e8077e3 Update CHANGELOG.md [ci skip] 2019-03-26 13:23:20 -03:00
Leonardo Tegon
55e726e4a7 Merge pull request #5051 from plataformatec/mf-revert-update-encrypted-password-to-nil-if-password-is-nil
Reverts both "[#4245] Allow password to nil (#4261)" and "Add more tests (#4970)"
2019-03-26 11:37:27 -03:00
Marcos Ferreira
f9d13f015a Revert "[#4245] Allowing password to nil (#4261)"
This reverts commit 3aedbf0a4d.
2019-03-26 10:29:46 -03:00
Marcos Ferreira
e704221842 Revert "Add more tests (#4970)"
This reverts commit 05bf574799.
2019-03-26 10:29:16 -03:00
Leonardo Tegon
2a6d608bd8 Merge pull request #5050 from softwaregravy/patch-1
Fix dead link
2019-03-25 18:02:20 -03:00
John Hinnegan
b2dc388556 Fix dead link
The link that's in there goes to a non-existent page. I'm guessing it was moved. I put the link in that I think it was moved to.
2019-03-24 11:07:29 -04:00
Marcos Ferreira
214ce91bf7 Merge pull request #5045 from matheusbn/small-text-fix-in-strategies/authenticatable.rb [skip ci]
Fix RDoc text redundancy in authenticatable.rb strategy
2019-03-18 09:40:53 -03:00
Matheus Berkenbrock Nedel
a460d79b08 fix text redundancy 2019-03-17 21:20:06 -03:00
emersonthis
2f3a59640b Added mention of API mode complications to README (#5041) [ci skip] 2019-03-14 18:19:45 -03:00
Felipe Renan
1102600720 Force nokogiri 1.9.1 on Rails 4.1
Sometimes our CI breaks because it tries to fetch a higher nokogiri
version that requires Ruby 2.3+.
2019-03-03 20:48:40 -03:00
Leonardo Tegon
4e4e42e1ae Include note about security issues in ISSUE_TEMPLATE.md 2019-02-19 16:04:32 -03:00
Leonardo Tegon
20e299bce0 Prepare for 4.6.1 release v4.6.1 2019-02-11 11:38:35 -02:00
Leonardo Tegon
5be5220e5a Update CHANGELOG.md [ci skip] 2019-02-11 11:04:12 -02:00
Leonardo Tegon
fcb04f5302 Check if root_path is defined with #respond_to? instead of #present (#5022)
When an application does not define a `root`, the method will be
undefined instead of returning a falsey value.
This commit also includes a new test with fake objects that mimic this
behavior.

Related resources:

* 1aab449933 (diff-c1be825bdb5f3160081e41432f83d0d7R278)
* https://github.com/plataformatec/devise/issues/5021
2019-02-11 11:00:56 -02:00
Leonardo Tegon
45016829d6 Prepare for 4.6.0 release
It was necessary to manually set the `sqlite3` Gem version after the
release of version `1.4.0`.
See https://github.com/rails/rails/issues/35153 for more info.
v4.6.0
2019-02-07 13:28:00 -02:00
Leonardo Tegon
337db5a3a2 Update CHANGELOG.md [ci skip] 2019-02-06 20:01:55 -02:00
Felipe Renan
a921c10e6a Merge pull request #5018 from plataformatec/frg-fix-webrat-warning
Fix webrat warnings
2019-02-04 15:54:36 -02:00
Felipe Renan
101e50f9df Fix webrat warnings
Forwarding methods to private methods is deprecated and produces a
warning after Ruby 2.4.

see: https://bugs.ruby-lang.org/issues/12782

To fix this issue I'm mocking patching webrat making RailsAdatper#response
method public since Webrat::Session is delegating functions to it.
2019-02-04 14:50:14 -02:00
Felipe Renan
6064ca8b73 Merge pull request #5014 from plataformatec/frg-fix-test-setup
Fix bin/test using Rails::TestUnit
2019-01-30 19:00:18 -02:00
Felipe Renan
2dd322579f Fix bin/test to use Rails::TestUnit
The command bin/test stop running single tests once Devise started to
support Rails 5.2. The problem is because we used `rails/test_unit/minitest_plugin`
and this file was moved to another place.

See: https://github.com/rails/rails/pull/29572

I'm not sure if we should require the `minitest-plugin` directly from
Rails like we were doing, I tried it and it didn't work. So I'm
changing this `bin/test` completely based on how Rails does that [here](https://github.com/rails/rails/blob/master/tools/test.rb)
2019-01-29 17:50:50 -02:00
Felipe Renan
e72839f4bc Merge pull request #5011 from plataformatec/frg-fix-sqlite-warning
Fix SQLite3 warning when running tests
2019-01-28 15:29:25 -02:00
Felipe Renan
45438fcfc4 Fix SQLite3 warning
Before setting this option, our test suite was giving the following warning:

```
DEPRECATION WARNING: Leaving `ActiveRecord::ConnectionAdapters::SQLite3Adapter.represent_boolean_as_integer`
set to false is deprecated. SQLite databases have used 't' and 'f' to serialize
boolean values and must have old data converted to 1 and 0 (its native boolean
serialization) before setting this flag to true. Conversion can be accomplished
by setting up a rake task which runs

  ExampleModel.where("boolean_column = 't'").update_all(boolean_column: 1)
  ExampleModel.where("boolean_column = 'f'").update_all(boolean_column: 0)

for all models and all boolean columns, after which the flag must be set to
true by adding the following to your application.rb file:

  Rails.application.config.active_record.sqlite3.represent_boolean_as_integer = true
 (called from <top (required)> at $PATH/devise/test/rails_app/app/active_record/user.rb:5)
```

After configuring `represent_boolean_as_integer = true` as specified
above, we don't have this warning anymore.

More info:
https://api.rubyonrails.org/classes/ActiveRecord/ConnectionAdapters/SQLite3Adapter.html#method-c-represent_boolean_as_integer
2019-01-28 14:39:06 -02:00
Felipe Renan
f3329110ba Update README to help run tests [ci skip] (#5012)
Adds a section to help contributors run Devise test suite.
2019-01-28 14:31:42 -02:00
Hyeonseok
369ba267ef removing white space in devise generator new.html.erb (#5010) 2019-01-24 10:20:03 -02:00
Felipe Renan
a42033cb04 Update CHANGELOG.md [ci skip] 2019-01-23 16:02:12 -02:00
Felipe Renan
1aab449933 Refactor fix #4127
* Adding to check if rootpath is present on url_helpers.
* Run this code only for Rails versions lower than 5.1.
2019-01-23 15:42:17 -02:00
Alessandro
ee65cd60c9 FIX plataformatec/devise#4127 (#4700) 2019-01-23 15:32:29 -02:00
Leonardo Tegon
c000b58c56 Add Rails 6 to CI (#5009) 2019-01-22 12:43:31 -02:00
Matthew Rudy Jacobs
40e8812d75 Apply Test Helper fix to Rails 6.0 as well as 5.x (#5002) 2019-01-17 18:29:00 -02:00
Felipe Renan
cc6046b8cc Merge pull request #5005 from plataformatec/frg-fix-travis-build
Fix travis build
2019-01-17 18:26:54 -02:00
Felipe Renan
6b1ead8654 Fix travis build
Now Travis uses bundler 2.0+ and it causes some build fails since we
still support Rails 4.

This commit set Travis to not use a bundler version higher than 2.0.

More info: https://docs.travis-ci.com/user/languages/ruby/#bundler-20
2019-01-17 17:45:12 -02:00
Patrick Jones
518f6f39bf Fixed broken README link in changelog [ci skip] (#4999) 2019-01-10 11:09:51 -02:00
Rafael França
6edbeb82cd Merge pull request #4998 from Atul9/update-license
Update copyright notice to 2019 [ci skip]
2019-01-07 14:23:26 -05:00
Atul Bhosale
a8d0a8f63d Update copyright notice to 2019 [ci skip] 2019-01-07 23:28:19 +05:30
Leonardo Tegon
62703943be Make #increment_failed_attempts concurrency safe (#4996)
As reported in #4981, the method `#increment_failed_attempts` of `Devise::Models::Lockable` was
not concurrency safe. The increment operation was being done in two steps: first the value was read from the database, and then incremented by 1. This may result in wrong values if two requests try to update the value concurrently. For example:

```
Browser1 -------> Read `failed_attempts` from DB (1) -------> Increment `failed_attempts` to 2
    Browser2 -------> Read `failed_attempts` from DB (1) -------> Increment `failed_attempts` to 2
```

In the example above, `failed_attempts` should have been set to 3, but it will be set to 2. 

This commit handles this case by calling `ActiveRecord::CounterCache.increment_counter` method, which will do both steps at once, reading the value straight from the database.

This commit also adds a `ActiveRecord::AttributeMethods::Dirty#reload` call to ensure that the application gets the updated value - i.e. that other request might have updated. 
Although this does not ensure that the value is in fact the most recent one - other request could've updated it after the `reload` call - it seems good enough for this implementation. 
Even if a request does not locks the account because it has a stale value, the next one - that updated that value - will do it. That's why we decided not to use a pessimistic lock here.

Closes #4981.
2018-12-28 17:00:50 -02:00
kenji kobayashi
e3a00b27d1 Add an option to not automatically sign in a user after changing a password (#4569) 2018-12-28 11:29:58 -02:00
Pavel Pestau
8266e85576 Removing extra characters (#4991) 2018-12-21 10:18:05 -02:00
Pavel Pestau
4372dd402b Removed extra characters (#4988) 2018-12-20 11:51:19 -02:00
Marcos Ferreira
6e4b0b94bc Don't run gem update --system and gem install bundler on CI
There were some issues with the combination of some `bundler` and `rubygems` versions.
We got around that by always updating `bundler` in the CI but since those issues
were fixed we can quit doing that.

Related to `gem update --system`:
- travis-ci/travis-ci#8978
- rubygems/rubygems#2123

Related to `gem install bundler`:
- travis-ci/travis-ci#9333
- travis-ci/travis-ci#8969
- bundler/bundler#6340
2018-12-19 17:04:56 -02:00
Leonardo Tegon
bf74a506c8 Update CHANGELOG.md [ci skip] 2018-12-19 10:32:50 -02:00
Marcos Ferreira
e01fb8bb22 Merge pull request #4989 from plataformatec/mf-fix-unclosed-tag-error-messages-partial
Fix unclosed html tag and add deprecation warn
2018-12-19 09:52:10 -02:00
Marcos Ferreira
14785afe8d Add a deprecation warn for DeviseHelper.devise_error_messages!
To customize the error style one should modifying the
`devise/shared/error_messages` partial.
2018-12-18 21:15:13 -02:00
Marcos Ferreira
66a96ab9d6 Fix unclosed li tag in error_messages partial 2018-12-18 21:15:11 -02:00
Graham Brereton
76b87dc0e8 Refactor the devise_error_messages! helper to render a partial (#4616)
This should make it more intuitive and easier for developers using the
project to implement their own layout for errors.
2018-12-18 21:01:16 -02:00
Leonardo Tegon
cb663e96a3 Update CHANGELOG.md [ci skip] 2018-12-04 14:37:20 -02:00
Stan Hu
1192c76f62 Fix corner case when confirmation_sent_at is equal to 0.days.ago (#4529)
If `Confirmable#confirmation_sent_at` is equal to `0.days.ago`, then
`confirmation_period_valid?` will be deemed valid even if the setting is
configured to disable this outright. To prevent this error, we explicitly
check the configuration setting to be `0.days.ago`.
2018-12-04 14:34:32 -02:00
Leonardo Tegon
ecd2187652 Update CHANGELOG.md [ci skip] 2018-11-30 17:43:28 -02:00
ihatov08
d1571627b7 Add deprication waring if use options argument at DatabaseAuthenticatable#update_with_password,#update_without_password (#4935) 2018-11-28 17:58:10 -02:00
Leonardo Tegon
52b24e41de Fix typo [ci skip] 2018-11-23 13:26:26 -02:00