discourse/spec/requests/admin/users_controller_spec.rb

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1396 lines
46 KiB
Ruby
Raw Normal View History

# frozen_string_literal: true
require 'discourse_ip_info'
require 'rotp'
RSpec.describe Admin::UsersController do
fab!(:admin) { Fabricate(:admin) }
fab!(:user) { Fabricate(:user) }
fab!(:coding_horror) { Fabricate(:coding_horror) }
it 'is a subclass of AdminController' do
expect(Admin::UsersController < Admin::AdminController).to eq(true)
end
before do
sign_in(admin)
end
describe '#index' do
it 'returns success with JSON' do
get "/admin/users/list.json"
expect(response.status).to eq(200)
expect(response.parsed_body).to be_present
end
context 'when showing emails' do
it "returns email for all the users" do
get "/admin/users/list.json", params: { show_emails: "true" }
expect(response.status).to eq(200)
data = response.parsed_body
data.each do |user|
expect(user["email"]).to be_present
end
end
it "logs only 1 entry" do
expect do
get "/admin/users/list.json", params: { show_emails: "true" }
end.to change { UserHistory.where(action: UserHistory.actions[:check_email], acting_user_id: admin.id).count }.by(1)
expect(response.status).to eq(200)
end
it "can be ordered by emails" do
get "/admin/users/list.json", params: { show_emails: "true", order: "email" }
expect(response.status).to eq(200)
end
end
end
describe '#show' do
context 'with an existing user' do
it 'returns success' do
get "/admin/users/#{user.id}.json"
expect(response.status).to eq(200)
end
it 'includes associated accounts' do
user.user_associated_accounts.create!(provider_name: 'pluginauth', provider_uid: 'pluginauth_uid')
get "/admin/users/#{user.id}.json"
expect(response.status).to eq(200)
expect(response.parsed_body['external_ids'].size).to eq(1)
expect(response.parsed_body['external_ids']['pluginauth']).to eq('pluginauth_uid')
end
end
context 'with a non-existing user' do
it 'returns 404 error' do
get "/admin/users/0.json"
expect(response.status).to eq(404)
end
end
end
describe '#approve' do
let(:evil_trout) { Fabricate(:evil_trout) }
before do
SiteSetting.must_approve_users = true
end
it "raises an error when the user doesn't have permission" do
sign_in(user)
put "/admin/users/#{evil_trout.id}/approve.json"
expect(response.status).to eq(404)
evil_trout.reload
expect(evil_trout.approved).to eq(false)
end
it "will create a reviewable if one does not exist" do
evil_trout.update!(active: true)
expect(ReviewableUser.find_by(target: evil_trout)).to be_blank
put "/admin/users/#{evil_trout.id}/approve.json"
expect(response.code).to eq("200")
expect(ReviewableUser.find_by(target: evil_trout)).to be_present
expect(evil_trout.reload).to be_approved
end
it 'calls approve' do
Jobs.run_immediately!
evil_trout.activate
put "/admin/users/#{evil_trout.id}/approve.json"
expect(response.status).to eq(200)
evil_trout.reload
expect(evil_trout.approved).to eq(true)
2019-03-12 04:16:56 -04:00
expect(UserHistory.where(action: UserHistory.actions[:approve_user], target_user_id: evil_trout.id).count).to eq(1)
end
end
describe '#approve_bulk' do
before do
SiteSetting.must_approve_users = true
end
let(:evil_trout) { Fabricate(:evil_trout) }
it "does nothing without users" do
put "/admin/users/approve-bulk.json"
evil_trout.reload
expect(response.status).to eq(200)
expect(evil_trout.approved).to eq(false)
end
it "won't approve the user when not allowed" do
sign_in(user)
put "/admin/users/approve-bulk.json", params: { users: [evil_trout.id] }
expect(response.status).to eq(404)
evil_trout.reload
expect(evil_trout.approved).to eq(false)
end
it "approves the user when permitted" do
Jobs.run_immediately!
evil_trout.activate
put "/admin/users/approve-bulk.json", params: { users: [evil_trout.id] }
expect(response.status).to eq(200)
evil_trout.reload
expect(evil_trout.approved).to eq(true)
end
end
describe '#suspend' do
fab!(:created_post) { Fabricate(:post) }
let(:suspend_params) do
{ suspend_until: 5.hours.from_now,
reason: "because of this post",
post_id: created_post.id }
end
it "works properly" do
expect(user).not_to be_suspended
expect do
put "/admin/users/#{user.id}/suspend.json", params: {
suspend_until: 5.hours.from_now,
reason: "because I said so"
}
end.not_to change { Jobs::CriticalUserEmail.jobs.size }
expect(response.status).to eq(200)
user.reload
expect(user).to be_suspended
expect(user.suspended_at).to be_present
expect(user.suspended_till).to be_present
expect(user.suspend_record).to be_present
log = UserHistory.where(target_user_id: user.id).order('id desc').first
expect(log.details).to match(/because I said so/)
end
it "checks if user is suspended" do
put "/admin/users/#{user.id}/suspend.json", params: {
suspend_until: 5.hours.from_now,
reason: "because I said so"
}
put "/admin/users/#{user.id}/suspend.json", params: {
suspend_until: 5.hours.from_now,
reason: "because I said so too"
}
expect(response.status).to eq(409)
expect(response.parsed_body["message"]).to eq(
I18n.t(
"user.already_suspended",
staff: admin.username,
time_ago: FreedomPatches::Rails4.time_ago_in_words(user.suspend_record.created_at, true, scope: :'datetime.distance_in_words_verbose')
)
)
end
it "requires suspend_until and reason" do
expect(user).not_to be_suspended
put "/admin/users/#{user.id}/suspend.json", params: {}
expect(response.status).to eq(400)
user.reload
expect(user).not_to be_suspended
expect(user).not_to be_suspended
put "/admin/users/#{user.id}/suspend.json", params: {
suspend_until: 5.hours.from_now
}
expect(response.status).to eq(400)
user.reload
expect(user).not_to be_suspended
end
context "with an associated post" do
it "can have an associated post" do
put "/admin/users/#{user.id}/suspend.json", params: suspend_params
expect(response.status).to eq(200)
log = UserHistory.where(target_user_id: user.id).order('id desc').first
expect(log.post_id).to eq(created_post.id)
end
it "can delete an associated post" do
put "/admin/users/#{user.id}/suspend.json", params: suspend_params.merge(post_action: 'delete')
created_post.reload
expect(created_post.deleted_at).to be_present
expect(response.status).to eq(200)
end
it "won't delete a category topic" do
c = Fabricate(:category_with_definition)
cat_post = c.topic.posts.first
put(
"/admin/users/#{user.id}/suspend.json",
params: suspend_params.merge(
post_action: 'delete',
post_id: cat_post.id
)
)
cat_post.reload
expect(cat_post.deleted_at).to be_blank
expect(response.status).to eq(200)
end
it "won't delete a category topic by replies" do
c = Fabricate(:category_with_definition)
cat_post = c.topic.posts.first
put(
"/admin/users/#{user.id}/suspend.json",
params: suspend_params.merge(
post_action: 'delete_replies',
post_id: cat_post.id
)
)
cat_post.reload
expect(cat_post.deleted_at).to be_blank
expect(response.status).to eq(200)
end
it "can delete an associated post and its replies" do
reply = PostCreator.create(
Fabricate(:user),
raw: 'this is the reply text',
reply_to_post_number: created_post.post_number,
topic_id: created_post.topic_id
)
nested_reply = PostCreator.create(
Fabricate(:user),
raw: 'this is the reply text2',
reply_to_post_number: reply.post_number,
topic_id: created_post.topic_id
)
put "/admin/users/#{user.id}/suspend.json", params: suspend_params.merge(post_action: 'delete_replies')
expect(created_post.reload.deleted_at).to be_present
expect(reply.reload.deleted_at).to be_present
expect(nested_reply.reload.deleted_at).to be_present
expect(response.status).to eq(200)
end
it "can edit an associated post" do
put "/admin/users/#{user.id}/suspend.json", params: suspend_params.merge(
post_action: 'edit',
post_edit: 'this is the edited content'
)
expect(response.status).to eq(200)
created_post.reload
expect(created_post.deleted_at).to be_blank
expect(created_post.raw).to eq("this is the edited content")
expect(response.status).to eq(200)
end
end
it "can send a message to the user" do
put "/admin/users/#{user.id}/suspend.json", params: {
suspend_until: 10.days.from_now,
reason: "short reason",
message: "long reason"
}
expect(response.status).to eq(200)
expect(Jobs::CriticalUserEmail.jobs.size).to eq(1)
job_args = Jobs::CriticalUserEmail.jobs.first["args"].first
expect(job_args["type"]).to eq("account_suspended")
expect(job_args["user_id"]).to eq(user.id)
log = UserHistory.where(target_user_id: user.id).order('id desc').first
expect(log).to be_present
expect(log.details).to match(/short reason/)
expect(log.details).to match(/long reason/)
end
it "also prevents use of any api keys" do
api_key = Fabricate(:api_key, user: user)
post "/bookmarks.json", params: {
bookmarkable_id: Fabricate(:post).id,
bookmarkable_type: "Post"
}, headers: { HTTP_API_KEY: api_key.key }
expect(response.status).to eq(200)
put "/admin/users/#{user.id}/suspend.json", params: suspend_params
expect(response.status).to eq(200)
user.reload
expect(user).to be_suspended
post "/bookmarks.json", params: {
post_id: Fabricate(:post).id
}, headers: { HTTP_API_KEY: api_key.key }
expect(response.status).to eq(403)
end
end
describe '#revoke_admin' do
fab!(:another_admin) { Fabricate(:admin) }
it 'raises an error unless the user can revoke access' do
sign_in(user)
put "/admin/users/#{another_admin.id}/revoke_admin.json"
expect(response.status).to eq(404)
another_admin.reload
expect(another_admin.admin).to eq(true)
end
it 'updates the admin flag' do
put "/admin/users/#{another_admin.id}/revoke_admin.json"
expect(response.status).to eq(200)
another_admin.reload
expect(another_admin.admin).to eq(false)
end
it 'returns detailed user schema' do
put "/admin/users/#{another_admin.id}/revoke_admin.json"
expect(response.parsed_body['can_be_merged']).to eq(true)
expect(response.parsed_body['can_be_deleted']).to eq(true)
expect(response.parsed_body['can_be_anonymized']).to eq(true)
expect(response.parsed_body['can_delete_all_posts']).to eq(true)
end
end
describe '#grant_admin' do
fab!(:another_user) { coding_horror }
after do
Discourse.redis.flushdb
end
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
it "returns a 404 when the acting user doesn't have permission" do
sign_in(user)
put "/admin/users/#{another_user.id}/grant_admin.json"
expect(response.status).to eq(404)
expect(AdminConfirmation.exists_for?(another_user.id)).to eq(false)
end
it "returns a 404 when the acting user doesn't have permission even if they have 2FA enabled" do
Fabricate(:user_second_factor_totp, user: user)
sign_in(user)
put "/admin/users/#{another_user.id}/grant_admin.json"
expect(response.status).to eq(404)
expect(AdminConfirmation.exists_for?(another_user.id)).to eq(false)
end
it "returns a 404 if the username doesn't exist" do
put "/admin/users/123123/grant_admin.json"
expect(response.status).to eq(404)
end
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
it 'sends a confirmation email if the acting admin does not have a second factor method enabled' do
expect(AdminConfirmation.exists_for?(another_user.id)).to eq(false)
put "/admin/users/#{another_user.id}/grant_admin.json"
expect(response.status).to eq(200)
expect(AdminConfirmation.exists_for?(another_user.id)).to eq(true)
end
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
it 'asks the acting admin for second factor if it is enabled' do
Fabricate(:user_second_factor_totp, user: admin)
FEATURE: Add 2FA support to the Discourse Connect Provider protocol (#16386) Discourse has the Discourse Connect Provider protocol that makes it possible to use a Discourse instance as an identity provider for external sites. As a natural extension to this protocol, this PR adds a new feature that makes it possible to use Discourse as a 2FA provider as well as an identity provider. The rationale for this change is that it's very difficult to implement 2FA support in a website and if you have multiple websites that need to have 2FA, it's unrealistic to build and maintain a separate 2FA implementation for each one. But with this change, you can piggyback on Discourse to take care of all the 2FA details for you for as many sites as you wish. To use Discourse as a 2FA provider, you'll need to follow this guide: https://meta.discourse.org/t/-/32974. It walks you through what you need to implement on your end/site and how to configure your Discourse instance. Once you're done, there is only one additional thing you need to do which is to include `require_2fa=true` in the payload that you send to Discourse. When Discourse sees `require_2fa=true`, it'll prompt the user to confirm their 2FA using whatever methods they've enabled (TOTP or security keys), and once they confirm they'll be redirected back to the return URL you've configured and the payload will contain `confirmed_2fa=true`. If the user has no 2FA methods enabled however, the payload will not contain `confirmed_2fa`, but it will contain `no_2fa_methods=true`. You'll need to be careful to re-run all the security checks and ensure the user can still access the resource on your site after they return from Discourse. This is very important because there's nothing that guarantees the user that will come back from Discourse after they confirm 2FA is the same user that you've redirected to Discourse. Internal ticket: t62183.
2022-04-13 08:04:09 -04:00
put "/admin/users/#{another_user.id}/grant_admin.json", xhr: true
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
expect(response.parsed_body["second_factor_challenge_nonce"]).to be_present
expect(another_user.reload.admin).to eq(false)
end
it 'grants admin if second factor is correct' do
user_second_factor = Fabricate(:user_second_factor_totp, user: admin)
FEATURE: Add 2FA support to the Discourse Connect Provider protocol (#16386) Discourse has the Discourse Connect Provider protocol that makes it possible to use a Discourse instance as an identity provider for external sites. As a natural extension to this protocol, this PR adds a new feature that makes it possible to use Discourse as a 2FA provider as well as an identity provider. The rationale for this change is that it's very difficult to implement 2FA support in a website and if you have multiple websites that need to have 2FA, it's unrealistic to build and maintain a separate 2FA implementation for each one. But with this change, you can piggyback on Discourse to take care of all the 2FA details for you for as many sites as you wish. To use Discourse as a 2FA provider, you'll need to follow this guide: https://meta.discourse.org/t/-/32974. It walks you through what you need to implement on your end/site and how to configure your Discourse instance. Once you're done, there is only one additional thing you need to do which is to include `require_2fa=true` in the payload that you send to Discourse. When Discourse sees `require_2fa=true`, it'll prompt the user to confirm their 2FA using whatever methods they've enabled (TOTP or security keys), and once they confirm they'll be redirected back to the return URL you've configured and the payload will contain `confirmed_2fa=true`. If the user has no 2FA methods enabled however, the payload will not contain `confirmed_2fa`, but it will contain `no_2fa_methods=true`. You'll need to be careful to re-run all the security checks and ensure the user can still access the resource on your site after they return from Discourse. This is very important because there's nothing that guarantees the user that will come back from Discourse after they confirm 2FA is the same user that you've redirected to Discourse. Internal ticket: t62183.
2022-04-13 08:04:09 -04:00
put "/admin/users/#{another_user.id}/grant_admin.json", xhr: true
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
nonce = response.parsed_body["second_factor_challenge_nonce"]
expect(nonce).to be_present
expect(another_user.reload.admin).to eq(false)
post "/session/2fa.json", params: {
nonce: nonce,
second_factor_token: ROTP::TOTP.new(user_second_factor.data).now,
second_factor_method: UserSecondFactor.methods[:totp]
}
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
res = response.parsed_body
expect(response.status).to eq(200)
expect(res["ok"]).to eq(true)
expect(res["callback_method"]).to eq("PUT")
expect(res["callback_path"]).to eq("/admin/users/#{another_user.id}/grant_admin.json")
FEATURE: Add 2FA support to the Discourse Connect Provider protocol (#16386) Discourse has the Discourse Connect Provider protocol that makes it possible to use a Discourse instance as an identity provider for external sites. As a natural extension to this protocol, this PR adds a new feature that makes it possible to use Discourse as a 2FA provider as well as an identity provider. The rationale for this change is that it's very difficult to implement 2FA support in a website and if you have multiple websites that need to have 2FA, it's unrealistic to build and maintain a separate 2FA implementation for each one. But with this change, you can piggyback on Discourse to take care of all the 2FA details for you for as many sites as you wish. To use Discourse as a 2FA provider, you'll need to follow this guide: https://meta.discourse.org/t/-/32974. It walks you through what you need to implement on your end/site and how to configure your Discourse instance. Once you're done, there is only one additional thing you need to do which is to include `require_2fa=true` in the payload that you send to Discourse. When Discourse sees `require_2fa=true`, it'll prompt the user to confirm their 2FA using whatever methods they've enabled (TOTP or security keys), and once they confirm they'll be redirected back to the return URL you've configured and the payload will contain `confirmed_2fa=true`. If the user has no 2FA methods enabled however, the payload will not contain `confirmed_2fa`, but it will contain `no_2fa_methods=true`. You'll need to be careful to re-run all the security checks and ensure the user can still access the resource on your site after they return from Discourse. This is very important because there's nothing that guarantees the user that will come back from Discourse after they confirm 2FA is the same user that you've redirected to Discourse. Internal ticket: t62183.
2022-04-13 08:04:09 -04:00
expect(res["redirect_url"]).to eq("/admin/users/#{another_user.id}/#{another_user.username}")
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
expect(another_user.reload.admin).to eq(false)
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
put res["callback_path"], params: {
second_factor_nonce: nonce
}
expect(response.status).to eq(200)
expect(another_user.reload.admin).to eq(true)
end
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
it 'does not grant admin if second factor auth is not successful' do
user_second_factor = Fabricate(:user_second_factor_totp, user: admin)
FEATURE: Add 2FA support to the Discourse Connect Provider protocol (#16386) Discourse has the Discourse Connect Provider protocol that makes it possible to use a Discourse instance as an identity provider for external sites. As a natural extension to this protocol, this PR adds a new feature that makes it possible to use Discourse as a 2FA provider as well as an identity provider. The rationale for this change is that it's very difficult to implement 2FA support in a website and if you have multiple websites that need to have 2FA, it's unrealistic to build and maintain a separate 2FA implementation for each one. But with this change, you can piggyback on Discourse to take care of all the 2FA details for you for as many sites as you wish. To use Discourse as a 2FA provider, you'll need to follow this guide: https://meta.discourse.org/t/-/32974. It walks you through what you need to implement on your end/site and how to configure your Discourse instance. Once you're done, there is only one additional thing you need to do which is to include `require_2fa=true` in the payload that you send to Discourse. When Discourse sees `require_2fa=true`, it'll prompt the user to confirm their 2FA using whatever methods they've enabled (TOTP or security keys), and once they confirm they'll be redirected back to the return URL you've configured and the payload will contain `confirmed_2fa=true`. If the user has no 2FA methods enabled however, the payload will not contain `confirmed_2fa`, but it will contain `no_2fa_methods=true`. You'll need to be careful to re-run all the security checks and ensure the user can still access the resource on your site after they return from Discourse. This is very important because there's nothing that guarantees the user that will come back from Discourse after they confirm 2FA is the same user that you've redirected to Discourse. Internal ticket: t62183.
2022-04-13 08:04:09 -04:00
put "/admin/users/#{another_user.id}/grant_admin.json", xhr: true
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
nonce = response.parsed_body["second_factor_challenge_nonce"]
expect(nonce).to be_present
expect(another_user.reload.admin).to eq(false)
token = ROTP::TOTP.new(user_second_factor.data).now.to_i
token = (token == 999_999 ? token - 1 : token + 1).to_s
post "/session/2fa.json", params: {
nonce: nonce,
second_factor_token: token,
second_factor_method: UserSecondFactor.methods[:totp]
}
expect(response.status).to eq(400)
expect(another_user.reload.admin).to eq(false)
put "/admin/users/#{another_user.id}/grant_admin.json", params: {
second_factor_nonce: nonce
}
expect(response.status).to eq(401)
expect(another_user.reload.admin).to eq(false)
end
it 'does not grant admin if the acting admin loses permission in the middle of the process' do
user_second_factor = Fabricate(:user_second_factor_totp, user: admin)
FEATURE: Add 2FA support to the Discourse Connect Provider protocol (#16386) Discourse has the Discourse Connect Provider protocol that makes it possible to use a Discourse instance as an identity provider for external sites. As a natural extension to this protocol, this PR adds a new feature that makes it possible to use Discourse as a 2FA provider as well as an identity provider. The rationale for this change is that it's very difficult to implement 2FA support in a website and if you have multiple websites that need to have 2FA, it's unrealistic to build and maintain a separate 2FA implementation for each one. But with this change, you can piggyback on Discourse to take care of all the 2FA details for you for as many sites as you wish. To use Discourse as a 2FA provider, you'll need to follow this guide: https://meta.discourse.org/t/-/32974. It walks you through what you need to implement on your end/site and how to configure your Discourse instance. Once you're done, there is only one additional thing you need to do which is to include `require_2fa=true` in the payload that you send to Discourse. When Discourse sees `require_2fa=true`, it'll prompt the user to confirm their 2FA using whatever methods they've enabled (TOTP or security keys), and once they confirm they'll be redirected back to the return URL you've configured and the payload will contain `confirmed_2fa=true`. If the user has no 2FA methods enabled however, the payload will not contain `confirmed_2fa`, but it will contain `no_2fa_methods=true`. You'll need to be careful to re-run all the security checks and ensure the user can still access the resource on your site after they return from Discourse. This is very important because there's nothing that guarantees the user that will come back from Discourse after they confirm 2FA is the same user that you've redirected to Discourse. Internal ticket: t62183.
2022-04-13 08:04:09 -04:00
put "/admin/users/#{another_user.id}/grant_admin.json", xhr: true
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
nonce = response.parsed_body["second_factor_challenge_nonce"]
expect(nonce).to be_present
expect(another_user.reload.admin).to eq(false)
post "/session/2fa.json", params: {
nonce: nonce,
second_factor_token: ROTP::TOTP.new(user_second_factor.data).now,
second_factor_method: UserSecondFactor.methods[:totp]
}
res = response.parsed_body
expect(response.status).to eq(200)
expect(res["ok"]).to eq(true)
expect(res["callback_method"]).to eq("PUT")
expect(res["callback_path"]).to eq("/admin/users/#{another_user.id}/grant_admin.json")
FEATURE: Add 2FA support to the Discourse Connect Provider protocol (#16386) Discourse has the Discourse Connect Provider protocol that makes it possible to use a Discourse instance as an identity provider for external sites. As a natural extension to this protocol, this PR adds a new feature that makes it possible to use Discourse as a 2FA provider as well as an identity provider. The rationale for this change is that it's very difficult to implement 2FA support in a website and if you have multiple websites that need to have 2FA, it's unrealistic to build and maintain a separate 2FA implementation for each one. But with this change, you can piggyback on Discourse to take care of all the 2FA details for you for as many sites as you wish. To use Discourse as a 2FA provider, you'll need to follow this guide: https://meta.discourse.org/t/-/32974. It walks you through what you need to implement on your end/site and how to configure your Discourse instance. Once you're done, there is only one additional thing you need to do which is to include `require_2fa=true` in the payload that you send to Discourse. When Discourse sees `require_2fa=true`, it'll prompt the user to confirm their 2FA using whatever methods they've enabled (TOTP or security keys), and once they confirm they'll be redirected back to the return URL you've configured and the payload will contain `confirmed_2fa=true`. If the user has no 2FA methods enabled however, the payload will not contain `confirmed_2fa`, but it will contain `no_2fa_methods=true`. You'll need to be careful to re-run all the security checks and ensure the user can still access the resource on your site after they return from Discourse. This is very important because there's nothing that guarantees the user that will come back from Discourse after they confirm 2FA is the same user that you've redirected to Discourse. Internal ticket: t62183.
2022-04-13 08:04:09 -04:00
expect(res["redirect_url"]).to eq("/admin/users/#{another_user.id}/#{another_user.username}")
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
expect(another_user.reload.admin).to eq(false)
admin.update!(admin: false)
put res["callback_path"], params: {
second_factor_nonce: nonce
}
expect(response.status).to eq(404)
expect(another_user.reload.admin).to eq(false)
end
it 'does not accept backup codes' do
Fabricate(:user_second_factor_totp, user: admin)
Fabricate(:user_second_factor_backup, user: admin)
FEATURE: Add 2FA support to the Discourse Connect Provider protocol (#16386) Discourse has the Discourse Connect Provider protocol that makes it possible to use a Discourse instance as an identity provider for external sites. As a natural extension to this protocol, this PR adds a new feature that makes it possible to use Discourse as a 2FA provider as well as an identity provider. The rationale for this change is that it's very difficult to implement 2FA support in a website and if you have multiple websites that need to have 2FA, it's unrealistic to build and maintain a separate 2FA implementation for each one. But with this change, you can piggyback on Discourse to take care of all the 2FA details for you for as many sites as you wish. To use Discourse as a 2FA provider, you'll need to follow this guide: https://meta.discourse.org/t/-/32974. It walks you through what you need to implement on your end/site and how to configure your Discourse instance. Once you're done, there is only one additional thing you need to do which is to include `require_2fa=true` in the payload that you send to Discourse. When Discourse sees `require_2fa=true`, it'll prompt the user to confirm their 2FA using whatever methods they've enabled (TOTP or security keys), and once they confirm they'll be redirected back to the return URL you've configured and the payload will contain `confirmed_2fa=true`. If the user has no 2FA methods enabled however, the payload will not contain `confirmed_2fa`, but it will contain `no_2fa_methods=true`. You'll need to be careful to re-run all the security checks and ensure the user can still access the resource on your site after they return from Discourse. This is very important because there's nothing that guarantees the user that will come back from Discourse after they confirm 2FA is the same user that you've redirected to Discourse. Internal ticket: t62183.
2022-04-13 08:04:09 -04:00
put "/admin/users/#{another_user.id}/grant_admin.json", xhr: true
FEATURE: Centralized 2FA page (#15377) 2FA support in Discourse was added and grown gradually over the years: we first added support for TOTP for logins, then we implemented backup codes, and last but not least, security keys. 2FA usage was initially limited to logging in, but it has been expanded and we now require 2FA for risky actions such as adding a new admin to the site. As a result of this gradual growth of the 2FA system, technical debt has accumulated to the point where it has become difficult to require 2FA for more actions. We now have 5 different 2FA UI implementations and each one has to support all 3 2FA methods (TOTP, backup codes, and security keys) which makes it difficult to maintain a consistent UX for these different implementations. Moreover, there is a lot of repeated logic in the server-side code behind these 5 UI implementations which hinders maintainability even more. This commit is the first step towards repaying the technical debt: it builds a system that centralizes as much as possible of the 2FA server-side logic and UI. The 2 main components of this system are: 1. A dedicated page for 2FA with support for all 3 methods. 2. A reusable server-side class that centralizes the 2FA logic (the `SecondFactor::AuthManager` class). From a top-level view, the 2FA flow in this new system looks like this: 1. User initiates an action that requires 2FA; 2. Server is aware that 2FA is required for this action, so it redirects the user to the 2FA page if the user has a 2FA method, otherwise the action is performed. 3. User submits the 2FA form on the page; 4. Server validates the 2FA and if it's successful, the action is performed and the user is redirected to the previous page. A more technically-detailed explanation/documentation of the new system is available as a comment at the top of the `lib/second_factor/auth_manager.rb` file. Please note that the details are not set in stone and will likely change in the future, so please don't use the system in your plugins yet. Since this is a new system that needs to be tested, we've decided to migrate only the 2FA for adding a new admin to the new system at this time (in this commit). Our plan is to gradually migrate the remaining 2FA implementations to the new system. For screenshots of the 2FA page, see PR #15377 on GitHub.
2022-02-17 04:12:59 -05:00
nonce = response.parsed_body["second_factor_challenge_nonce"]
expect(nonce).to be_present
expect(another_user.reload.admin).to eq(false)
post "/session/2fa.json", params: {
nonce: nonce,
second_factor_token: "iAmValidBackupCode",
second_factor_method: UserSecondFactor.methods[:backup_codes]
}
expect(response.status).to eq(403)
expect(another_user.reload.admin).to eq(false)
end
end
describe '#add_group' do
fab!(:group) { Fabricate(:group) }
it 'adds the user to the group' do
post "/admin/users/#{user.id}/groups.json", params: {
group_id: group.id
}
expect(response.status).to eq(200)
expect(GroupUser.where(user_id: user.id, group_id: group.id).exists?).to eq(true)
group_history = GroupHistory.last
expect(group_history.action).to eq(GroupHistory.actions[:add_user_to_group])
expect(group_history.acting_user).to eq(admin)
expect(group_history.target_user).to eq(user)
# Doing it again doesn't raise an error
post "/admin/users/#{user.id}/groups.json", params: {
group_id: group.id
}
expect(response.status).to eq(200)
end
it 'returns not-found error when there is no group' do
group.destroy!
put "/admin/users/#{user.id}/groups.json", params: {
group_id: group.id
}
expect(response.status).to eq(404)
end
it 'does not allow adding users to an automatic group' do
group.update!(automatic: true)
expect do
post "/admin/users/#{user.id}/groups.json", params: {
group_id: group.id
}
end.to_not change { group.users.count }
expect(response.status).to eq(422)
expect(response.parsed_body["errors"]).to eq(["You cannot modify an automatic group"])
end
end
describe '#remove_group' do
it "also clears the user's primary group" do
group = Fabricate(:group, users: [user])
user.update!(primary_group_id: group.id)
delete "/admin/users/#{user.id}/groups/#{group.id}.json"
expect(response.status).to eq(200)
expect(user.reload.primary_group).to eq(nil)
end
it 'returns not-found error when there is no group' do
delete "/admin/users/#{user.id}/groups/9090.json"
expect(response.status).to eq(404)
end
it 'does not allow removing owners from an automatic group' do
group = Fabricate(:group, users: [user], automatic: true)
delete "/admin/users/#{user.id}/groups/#{group.id}.json"
expect(response.status).to eq(422)
expect(response.parsed_body["errors"]).to eq(["You cannot modify an automatic group"])
end
end
describe '#trust_level' do
fab!(:another_user) {
coding_horror.update!(created_at: 1.month.ago)
coding_horror
}
it "raises an error when the user doesn't have permission" do
sign_in(user)
put "/admin/users/#{another_user.id}/trust_level.json"
expect(response.status).to eq(404)
end
it "returns a 404 if the username doesn't exist" do
put "/admin/users/123123/trust_level.json"
expect(response.status).to eq(404)
end
it "upgrades the user's trust level" do
put "/admin/users/#{another_user.id}/trust_level.json", params: { level: 2 }
expect(response.status).to eq(200)
another_user.reload
expect(another_user.trust_level).to eq(2)
expect(UserHistory.where(
target_user: another_user,
acting_user: admin,
action: UserHistory.actions[:change_trust_level]
).count).to eq(1)
end
it "raises no error when demoting a user below their current trust level (locks trust level)" do
stat = another_user.user_stat
stat.topics_entered = SiteSetting.tl1_requires_topics_entered + 1
stat.posts_read_count = SiteSetting.tl1_requires_read_posts + 1
stat.time_read = SiteSetting.tl1_requires_time_spent_mins * 60
stat.save!
another_user.update(trust_level: TrustLevel[1])
put "/admin/users/#{another_user.id}/trust_level.json", params: {
level: TrustLevel[0]
}
expect(response.status).to eq(200)
another_user.reload
expect(another_user.trust_level).to eq(TrustLevel[0])
expect(another_user.manual_locked_trust_level).to eq(TrustLevel[0])
end
end
describe '#grant_moderation' do
fab!(:another_user) { coding_horror }
it "raises an error when the user doesn't have permission" do
sign_in(user)
put "/admin/users/#{another_user.id}/grant_moderation.json"
expect(response.status).to eq(404)
end
it "returns a 404 if the username doesn't exist" do
put "/admin/users/123123/grant_moderation.json"
expect(response.status).to eq(404)
end
it 'updates the moderator flag' do
2020-07-24 05:16:52 -04:00
expect_enqueued_with(job: :send_system_message, args: {
user_id: another_user.id,
message_type: 'welcome_staff',
message_options: { role: :moderator }
}) do
put "/admin/users/#{another_user.id}/grant_moderation.json"
end
expect(response.status).to eq(200)
another_user.reload
expect(another_user.moderator).to eq(true)
end
it 'returns detailed user schema' do
put "/admin/users/#{another_user.id}/grant_moderation.json"
expect(response.parsed_body['can_be_merged']).to eq(false)
expect(response.parsed_body['can_be_anonymized']).to eq(false)
end
end
describe '#revoke_moderation' do
fab!(:moderator) { Fabricate(:moderator) }
it 'raises an error unless the user can revoke access' do
sign_in(user)
put "/admin/users/#{moderator.id}/revoke_moderation.json"
expect(response.status).to eq(404)
moderator.reload
expect(moderator.moderator).to eq(true)
end
it 'updates the moderator flag' do
put "/admin/users/#{moderator.id}/revoke_moderation.json"
expect(response.status).to eq(200)
moderator.reload
expect(moderator.moderator).to eq(false)
end
it 'returns detailed user schema' do
put "/admin/users/#{moderator.id}/revoke_moderation.json"
expect(response.parsed_body['can_be_merged']).to eq(true)
expect(response.parsed_body['can_be_anonymized']).to eq(true)
end
end
describe '#primary_group' do
fab!(:group) { Fabricate(:group) }
fab!(:another_user) { coding_horror }
fab!(:another_group) { Fabricate(:group, title: 'New') }
it "raises an error when the user doesn't have permission" do
sign_in(user)
put "/admin/users/#{another_user.id}/primary_group.json"
expect(response.status).to eq(404)
another_user.reload
expect(another_user.primary_group_id).to eq(nil)
end
it "returns a 404 if the user doesn't exist" do
put "/admin/users/123123/primary_group.json"
expect(response.status).to eq(404)
end
it "changes the user's primary group" do
group.add(another_user)
put "/admin/users/#{another_user.id}/primary_group.json", params: {
primary_group_id: group.id
}
expect(response.status).to eq(200)
another_user.reload
expect(another_user.primary_group_id).to eq(group.id)
end
it "doesn't change primary group if they aren't a member of the group" do
put "/admin/users/#{another_user.id}/primary_group.json", params: {
primary_group_id: group.id
}
expect(response.status).to eq(200)
another_user.reload
expect(another_user.primary_group_id).to eq(nil)
end
it "remove user's primary group" do
group.add(another_user)
put "/admin/users/#{another_user.id}/primary_group.json", params: {
primary_group_id: ""
}
expect(response.status).to eq(200)
another_user.reload
expect(another_user.primary_group_id).to eq(nil)
end
it "updates user's title when it matches the previous primary group title" do
group.update_columns(primary_group: true, title: 'Previous')
group.add(another_user)
another_group.add(another_user)
expect(another_user.reload.title).to eq('Previous')
put "/admin/users/#{another_user.id}/primary_group.json", params: {
primary_group_id: another_group.id
}
another_user.reload
expect(response.status).to eq(200)
expect(another_user.primary_group_id).to eq(another_group.id)
expect(another_user.title).to eq('New')
end
it "doesn't update user's title when it does not match the previous primary group title" do
another_user.update_columns(title: 'Different')
group.update_columns(primary_group: true, title: 'Previous')
another_group.add(another_user)
group.add(another_user)
expect(another_user.reload.title).to eq('Different')
put "/admin/users/#{another_user.id}/primary_group.json", params: {
primary_group_id: another_group.id
}
another_user.reload
expect(response.status).to eq(200)
expect(another_user.primary_group_id).to eq(another_group.id)
expect(another_user.title).to eq('Different')
end
end
describe '#destroy' do
fab!(:delete_me) { Fabricate(:user) }
it "returns a 403 if the user doesn't exist" do
delete "/admin/users/123123drink.json"
expect(response.status).to eq(403)
end
context "when user has post" do
let(:topic) { Fabricate(:topic, user: delete_me) }
let!(:post) { Fabricate(:post, topic: topic, user: delete_me) }
it "returns an api response that the user can't be deleted because it has posts" do
post_count = delete_me.posts.joins(:topic).count
delete_me_topic = Fabricate(:topic)
Fabricate(:post, topic: delete_me_topic, user: delete_me)
PostDestroyer.new(admin, delete_me_topic.first_post, context: "Deleted by admin").destroy
delete "/admin/users/#{delete_me.id}.json"
expect(response.status).to eq(403)
json = response.parsed_body
expect(json['deleted']).to eq(false)
Fix i18n issues reported on Crowdin (#13191) * Pluralize `js.topics.bulk.dismiss_new_with_selected` This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/246/en-fr#57320 * Pluralize `js.topics.bulk.dismiss_read_with_selected` This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/246/en-fr#57316 * Pluralize `js.topics.bulk.dismiss_button_with_selected` * Replaces concatenated string used by `js.topic.suggest_create_topic` This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/246/en-fr#41834 * Less confusing `admin_js.admin.watched_words.test.modal_title` This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/246/en-sv#44992 * Delete unused `backup.location.*` keys This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/248/en-fr#46330 * Replace "reviewable" with "reviewable items" This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/248/en-fr#56952 * Remove "ago" from `emails.incoming.missing_attachment` This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/248/en-sv#46038 * Remove "/Posts" from `js.keyboard_shortcuts_help.application.dismiss_new_posts` Because the shortcut doesn't do anything to posts anymore. This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/246/en-de#43180 * Pluralize `user.cannot_delete_has_posts` This fixes https://discourse.crowdin.com/translate/f3230e7607a36bb0a2f97fd90605a44e/248/en-he#57490
2021-06-22 05:29:35 -04:00
expect(json['message']).to eq(I18n.t("user.cannot_delete_has_posts", username: delete_me.username, count: post_count))
end
it "doesn't return an error if delete_posts == true" do
delete "/admin/users/#{delete_me.id}.json", params: { delete_posts: true }
expect(response.status).to eq(200)
expect(Post.where(id: post.id).count).to eq(0)
expect(Topic.where(id: topic.id).count).to eq(0)
expect(User.where(id: delete_me.id).count).to eq(0)
end
context "when user has reviewable flagged post which was handled" do
let!(:reviewable) { Fabricate(:reviewable_flagged_post, created_by: admin, target_created_by: delete_me, target: post, topic: topic, status: 4) }
it "deletes the user record" do
delete "/admin/users/#{delete_me.id}.json", params: { delete_posts: true, delete_as_spammer: true }
expect(response.status).to eq(200)
expect(User.where(id: delete_me.id).count).to eq(0)
end
end
end
it "blocks the e-mail if block_email param is is true" do
user_emails = delete_me.user_emails.pluck(:email)
delete "/admin/users/#{delete_me.id}.json", params: { block_email: true }
expect(response.status).to eq(200)
expect(ScreenedEmail.exists?(email: user_emails)).to eq(true)
end
it "does not block the e-mails if block_email param is is false" do
user_emails = delete_me.user_emails.pluck(:email)
delete "/admin/users/#{delete_me.id}.json", params: { block_email: false }
expect(response.status).to eq(200)
expect(ScreenedEmail.exists?(email: user_emails)).to eq(false)
end
it "does not block the e-mails by default" do
user_emails = delete_me.user_emails.pluck(:email)
delete "/admin/users/#{delete_me.id}.json"
expect(response.status).to eq(200)
expect(ScreenedEmail.exists?(email: user_emails)).to eq(false)
end
it "blocks the ip address if block_ip param is true" do
ip_address = delete_me.ip_address
delete "/admin/users/#{delete_me.id}.json", params: { block_ip: true }
expect(response.status).to eq(200)
expect(ScreenedIpAddress.exists?(ip_address: ip_address)).to eq(true)
end
it "does not block the ip address if block_ip param is false" do
ip_address = delete_me.ip_address
delete "/admin/users/#{delete_me.id}.json", params: { block_ip: false }
expect(response.status).to eq(200)
expect(ScreenedIpAddress.exists?(ip_address: ip_address)).to eq(false)
end
it "does not block the ip address by default" do
ip_address = delete_me.ip_address
delete "/admin/users/#{delete_me.id}.json"
expect(response.status).to eq(200)
expect(ScreenedIpAddress.exists?(ip_address: ip_address)).to eq(false)
end
context "with param block_url" do
before do
@post = Fabricate(:post_with_external_links, user: delete_me)
TopicLink.extract_from(@post)
@urls = TopicLink.where(user: delete_me, internal: false)
.pluck(:url)
.map { |url| ScreenedUrl.normalize_url(url) }
end
it "blocks the urls if block_url param is true" do
delete "/admin/users/#{delete_me.id}.json", params: { delete_posts: true, block_urls: true }
expect(response.status).to eq(200)
expect(ScreenedUrl.exists?(url: @urls)).to eq(true)
end
it "does not block the urls if block_url param is false" do
delete "/admin/users/#{delete_me.id}.json", params: { delete_posts: true, block_urls: false }
expect(response.status).to eq(200)
expect(ScreenedUrl.exists?(url: @urls)).to eq(false)
end
it "does not block the urls by default" do
delete "/admin/users/#{delete_me.id}.json", params: { delete_posts: true, block_urls: false }
expect(response.status).to eq(200)
expect(ScreenedUrl.exists?(url: @urls)).to eq(false)
end
end
it "deletes the user record" do
delete "/admin/users/#{delete_me.id}.json"
expect(response.status).to eq(200)
expect(User.where(id: delete_me.id).count).to eq(0)
end
end
describe '#activate' do
fab!(:reg_user) { Fabricate(:inactive_user) }
it "returns success" do
put "/admin/users/#{reg_user.id}/activate.json"
expect(response.status).to eq(200)
json = response.parsed_body
expect(json['success']).to eq("OK")
reg_user.reload
expect(reg_user.active).to eq(true)
end
it "should confirm email even when the tokens are expired" do
reg_user.email_tokens.update_all(confirmed: false, expired: true)
reg_user.reload
expect(reg_user.email_confirmed?).to eq(false)
put "/admin/users/#{reg_user.id}/activate.json"
expect(response.status).to eq(200)
reg_user.reload
expect(reg_user.email_confirmed?).to eq(true)
end
end
describe '#deactivate' do
fab!(:reg_user) { Fabricate(:active_user) }
it "returns success" do
put "/admin/users/#{reg_user.id}/deactivate.json"
expect(response.status).to eq(200)
json = response.parsed_body
expect(json['success']).to eq("OK")
reg_user.reload
expect(reg_user.active).to eq(false)
end
end
describe '#log_out' do
fab!(:reg_user) { Fabricate(:user) }
it "returns success" do
post "/admin/users/#{reg_user.id}/log_out.json"
expect(response.status).to eq(200)
json = response.parsed_body
expect(json['success']).to eq("OK")
end
it "returns 404 when user_id does not exist" do
post "/admin/users/123123drink/log_out.json"
expect(response.status).to eq(404)
end
end
describe '#silence' do
fab!(:reg_user) { Fabricate(:user) }
it "raises an error when the user doesn't have permission" do
sign_in(user)
put "/admin/users/#{reg_user.id}/silence.json"
expect(response.status).to eq(404)
reg_user.reload
expect(reg_user).not_to be_silenced
end
it "returns a 404 if the user doesn't exist" do
put "/admin/users/123123/silence.json"
expect(response.status).to eq(404)
end
it "punishes the user for spamming" do
put "/admin/users/#{reg_user.id}/silence.json"
expect(response.status).to eq(200)
reg_user.reload
expect(reg_user).to be_silenced
expect(reg_user.silenced_record).to be_present
end
it "can have an associated post" do
silence_post = Fabricate(:post, user: reg_user)
put "/admin/users/#{reg_user.id}/silence.json", params: {
post_id: silence_post.id,
post_action: 'edit',
post_edit: "this is the new contents for the post"
}
expect(response.status).to eq(200)
silence_post.reload
expect(silence_post.raw).to eq("this is the new contents for the post")
log = UserHistory.where(
target_user_id: reg_user.id,
action: UserHistory.actions[:silence_user]
).first
expect(log).to be_present
expect(log.post_id).to eq(silence_post.id)
reg_user.reload
expect(reg_user).to be_silenced
end
it "will set a length of time if provided" do
future_date = 1.month.from_now.to_date
put "/admin/users/#{reg_user.id}/silence.json", params: {
silenced_till: future_date
}
expect(response.status).to eq(200)
reg_user.reload
expect(reg_user).to be_silenced
expect(reg_user.silenced_till).to eq(future_date)
end
it "will send a message if provided" do
expect do
put "/admin/users/#{reg_user.id}/silence.json", params: {
message: "Email this to the user"
}
end.to change { Jobs::CriticalUserEmail.jobs.size }.by(1)
expect(response.status).to eq(200)
reg_user.reload
expect(reg_user).to be_silenced
end
it "checks if user is silenced" do
put "/admin/users/#{user.id}/silence.json", params: {
silenced_till: 5.hours.from_now,
reason: "because I said so"
}
put "/admin/users/#{user.id}/silence.json", params: {
silenced_till: 5.hours.from_now,
reason: "because I said so too"
}
expect(response.status).to eq(409)
expect(response.parsed_body["message"]).to eq(
I18n.t(
"user.already_silenced",
staff: admin.username,
time_ago: FreedomPatches::Rails4.time_ago_in_words(user.silenced_record.created_at, true, scope: :'datetime.distance_in_words_verbose')
)
)
end
end
describe '#unsilence' do
fab!(:reg_user) { Fabricate(:user, silenced_till: 10.years.from_now) }
it "raises an error when the user doesn't have permission" do
sign_in(user)
put "/admin/users/#{reg_user.id}/unsilence.json"
expect(response.status).to eq(404)
end
it "returns a 403 if the user doesn't exist" do
put "/admin/users/123123/unsilence.json"
expect(response.status).to eq(404)
end
it "unsilences the user" do
put "/admin/users/#{reg_user.id}/unsilence.json"
expect(response.status).to eq(200)
reg_user.reload
expect(reg_user.silenced?).to eq(false)
log = UserHistory.where(
target_user_id: reg_user.id,
action: UserHistory.actions[:unsilence_user]
).first
expect(log).to be_present
end
end
describe '#ip_info' do
it "retrieves IP info" do
ip = "81.2.69.142"
DiscourseIpInfo.open_db(File.join(Rails.root, 'spec', 'fixtures', 'mmdb'))
Resolv::DNS.any_instance.stubs(:getname).with(ip).returns("ip-81-2-69-142.example.com")
get "/admin/users/ip-info.json", params: { ip: ip }
expect(response.status).to eq(200)
expect(response.parsed_body.symbolize_keys).to eq(
city: "London",
country: "United Kingdom",
country_code: "GB",
geoname_ids: [6255148, 2635167, 2643743, 6269131],
hostname: "ip-81-2-69-142.example.com",
location: "London, England, United Kingdom",
region: "England",
latitude: 51.5142,
longitude: -0.0931,
)
end
end
describe '#delete_other_accounts_with_same_ip' do
it "works" do
user_a = Fabricate(:user, ip_address: "42.42.42.42")
user_b = Fabricate(:user, ip_address: "42.42.42.42")
delete "/admin/users/delete-others-with-same-ip.json", params: {
ip: "42.42.42.42", exclude: -1, order: "trust_level DESC"
}
expect(response.status).to eq(200)
expect(User.where(id: user_a.id).count).to eq(0)
expect(User.where(id: user_b.id).count).to eq(0)
end
end
describe '#sync_sso' do
let(:sso) { DiscourseConnectBase.new }
let(:sso_secret) { "sso secret" }
before do
SiteSetting.email_editable = false
FEATURE: Rename 'Discourse SSO' to DiscourseConnect (#11978) The 'Discourse SSO' protocol is being rebranded to DiscourseConnect. This should help to reduce confusion when 'SSO' is used in the generic sense. This commit aims to: - Rename `sso_` site settings. DiscourseConnect specific ones are prefixed `discourse_connect_`. Generic settings are prefixed `auth_` - Add (server-side-only) backwards compatibility for the old setting names, with deprecation notices - Copy `site_settings` database records to the new names - Rename relevant translation keys - Update relevant translations This commit does **not** aim to: - Rename any Ruby classes or methods. This might be done in a future commit - Change any URLs. This would break existing integrations - Make any changes to the protocol. This would break existing integrations - Change any functionality. Further normalization across DiscourseConnect and other auth methods will be done separately The risks are: - There is no backwards compatibility for site settings on the client-side. Accessing auth-related site settings in Javascript is fairly rare, and an error on the client side would not be security-critical. - If a plugin is monkey-patching parts of the auth process, changes to locale keys could cause broken error messages. This should also be unlikely. The old site setting names remain functional, so security-related overrides will remain working. A follow-up commit will be made with a post-deploy migration to delete the old `site_settings` rows.
2021-02-08 05:04:33 -05:00
SiteSetting.discourse_connect_url = "https://www.example.com/sso"
SiteSetting.enable_discourse_connect = true
SiteSetting.auth_overrides_email = true
SiteSetting.auth_overrides_name = true
SiteSetting.auth_overrides_username = true
SiteSetting.discourse_connect_secret = sso_secret
sso.sso_secret = sso_secret
end
it 'can sync up with the sso' do
sso.name = "Bob The Bob"
sso.username = "bob"
sso.email = "bob@bob.com"
sso.external_id = "1"
user = DiscourseConnect.parse(sso.payload, secure_session: read_secure_session).lookup_or_create_user
sso.name = "Bill"
sso.username = "Hokli$$!!"
sso.email = "bob2@bob.com"
post "/admin/users/sync_sso.json", params: Rack::Utils.parse_query(sso.payload)
expect(response.status).to eq(200)
user.reload
expect(user.email).to eq("bob2@bob.com")
expect(user.name).to eq("Bill")
expect(user.username).to eq("Hokli")
end
it 'should create new users' do
sso.name = "Dr. Claw"
sso.username = "dr_claw"
sso.email = "dr@claw.com"
sso.external_id = "2"
post "/admin/users/sync_sso.json", params: Rack::Utils.parse_query(sso.payload)
expect(response.status).to eq(200)
user = User.find_by_email('dr@claw.com')
expect(user).to be_present
expect(user.ip_address).to be_blank
end
it "triggers :sync_sso DiscourseEvent" do
sso.name = "Bob The Bob"
sso.username = "bob"
sso.email = "bob@bob.com"
sso.external_id = "1"
user = DiscourseConnect.parse(sso.payload, secure_session: read_secure_session).lookup_or_create_user
sso.name = "Bill"
sso.username = "Hokli$$!!"
sso.email = "bob2@bob.com"
events = DiscourseEvent.track_events do
post "/admin/users/sync_sso.json", params: Rack::Utils.parse_query(sso.payload)
end
expect(events).to include(event_name: :sync_sso, params: [user])
end
it 'should return the right message if the record is invalid' do
sso.email = ""
sso.name = ""
sso.external_id = "1"
post "/admin/users/sync_sso.json", params: Rack::Utils.parse_query(sso.payload)
expect(response.status).to eq(403)
expect(response.parsed_body["message"]).to include("Primary email can't be blank")
end
it 'should return the right message if the signature is invalid' do
sso.name = "Dr. Claw"
sso.username = "dr_claw"
sso.email = "dr@claw.com"
sso.external_id = "2"
correct_payload = Rack::Utils.parse_query(sso.payload)
post "/admin/users/sync_sso.json", params: correct_payload.merge(sig: "someincorrectsignature")
expect(response.status).to eq(422)
FEATURE: Rename 'Discourse SSO' to DiscourseConnect (#11978) The 'Discourse SSO' protocol is being rebranded to DiscourseConnect. This should help to reduce confusion when 'SSO' is used in the generic sense. This commit aims to: - Rename `sso_` site settings. DiscourseConnect specific ones are prefixed `discourse_connect_`. Generic settings are prefixed `auth_` - Add (server-side-only) backwards compatibility for the old setting names, with deprecation notices - Copy `site_settings` database records to the new names - Rename relevant translation keys - Update relevant translations This commit does **not** aim to: - Rename any Ruby classes or methods. This might be done in a future commit - Change any URLs. This would break existing integrations - Make any changes to the protocol. This would break existing integrations - Change any functionality. Further normalization across DiscourseConnect and other auth methods will be done separately The risks are: - There is no backwards compatibility for site settings on the client-side. Accessing auth-related site settings in Javascript is fairly rare, and an error on the client side would not be security-critical. - If a plugin is monkey-patching parts of the auth process, changes to locale keys could cause broken error messages. This should also be unlikely. The old site setting names remain functional, so security-related overrides will remain working. A follow-up commit will be made with a post-deploy migration to delete the old `site_settings` rows.
2021-02-08 05:04:33 -05:00
expect(response.parsed_body["message"]).to include(I18n.t('discourse_connect.login_error'))
expect(response.parsed_body["message"]).not_to include(correct_payload["sig"])
end
it "returns 404 if the external id does not exist" do
sso.name = "Dr. Claw"
sso.username = "dr_claw"
sso.email = "dr@claw.com"
sso.external_id = ""
post "/admin/users/sync_sso.json", params: Rack::Utils.parse_query(sso.payload)
expect(response.status).to eq(422)
FEATURE: Rename 'Discourse SSO' to DiscourseConnect (#11978) The 'Discourse SSO' protocol is being rebranded to DiscourseConnect. This should help to reduce confusion when 'SSO' is used in the generic sense. This commit aims to: - Rename `sso_` site settings. DiscourseConnect specific ones are prefixed `discourse_connect_`. Generic settings are prefixed `auth_` - Add (server-side-only) backwards compatibility for the old setting names, with deprecation notices - Copy `site_settings` database records to the new names - Rename relevant translation keys - Update relevant translations This commit does **not** aim to: - Rename any Ruby classes or methods. This might be done in a future commit - Change any URLs. This would break existing integrations - Make any changes to the protocol. This would break existing integrations - Change any functionality. Further normalization across DiscourseConnect and other auth methods will be done separately The risks are: - There is no backwards compatibility for site settings on the client-side. Accessing auth-related site settings in Javascript is fairly rare, and an error on the client side would not be security-critical. - If a plugin is monkey-patching parts of the auth process, changes to locale keys could cause broken error messages. This should also be unlikely. The old site setting names remain functional, so security-related overrides will remain working. A follow-up commit will be made with a post-deploy migration to delete the old `site_settings` rows.
2021-02-08 05:04:33 -05:00
expect(response.parsed_body["message"]).to include(I18n.t('discourse_connect.blank_id_error'))
end
end
describe '#disable_second_factor' do
let(:second_factor) { user.create_totp(enabled: true) }
2018-06-28 04:12:32 -04:00
let(:second_factor_backup) { user.generate_backup_codes }
let(:security_key) { Fabricate(:user_security_key, user: user) }
describe 'as an admin' do
before do
sign_in(admin)
second_factor
2018-06-28 04:12:32 -04:00
second_factor_backup
security_key
expect(user.reload.user_second_factors.totps.first).to eq(second_factor)
end
it 'should able to disable the second factor for another user' do
expect do
put "/admin/users/#{user.id}/disable_second_factor.json"
end.to change { Jobs::CriticalUserEmail.jobs.length }.by(1)
expect(response.status).to eq(200)
2018-06-28 04:12:32 -04:00
expect(user.reload.user_second_factors).to be_empty
expect(user.reload.security_keys).to be_empty
job_args = Jobs::CriticalUserEmail.jobs.first["args"].first
expect(job_args["user_id"]).to eq(user.id)
expect(job_args["type"]).to eq('account_second_factor_disabled')
end
it 'should not be able to disable the second factor for the current user' do
put "/admin/users/#{admin.id}/disable_second_factor.json"
expect(response.status).to eq(403)
end
describe 'when user has only one second factor type enabled' do
it 'should succeed with security keys' do
user.user_second_factors.destroy_all
put "/admin/users/#{user.id}/disable_second_factor.json"
expect(response.status).to eq(200)
end
it 'should succeed with totp' do
user.security_keys.destroy_all
put "/admin/users/#{user.id}/disable_second_factor.json"
expect(response.status).to eq(200)
end
end
describe 'when user does not have second factor enabled' do
it 'should raise the right error' do
2018-06-28 04:12:32 -04:00
user.user_second_factors.destroy_all
user.security_keys.destroy_all
put "/admin/users/#{user.id}/disable_second_factor.json"
expect(response.status).to eq(400)
end
end
end
end
describe "#penalty_history" do
fab!(:moderator) { Fabricate(:moderator) }
let(:logger) { StaffActionLogger.new(admin) }
it "doesn't allow moderators to clear a user's history" do
sign_in(moderator)
delete "/admin/users/#{user.id}/penalty_history.json"
expect(response.code).to eq("404")
end
def find_logs(action)
UserHistory.where(target_user_id: user.id, action: UserHistory.actions[action])
end
it "allows admins to clear a user's history" do
logger.log_user_suspend(user, "suspend reason")
logger.log_user_unsuspend(user)
logger.log_unsilence_user(user)
logger.log_silence_user(user)
sign_in(admin)
delete "/admin/users/#{user.id}/penalty_history.json"
expect(response.code).to eq("200")
expect(find_logs(:suspend_user)).to be_blank
expect(find_logs(:unsuspend_user)).to be_blank
expect(find_logs(:silence_user)).to be_blank
expect(find_logs(:unsilence_user)).to be_blank
expect(find_logs(:removed_suspend_user)).to be_present
expect(find_logs(:removed_unsuspend_user)).to be_present
expect(find_logs(:removed_silence_user)).to be_present
expect(find_logs(:removed_unsilence_user)).to be_present
end
end
describe "#delete_posts_batch" do
describe 'when user is is invalid' do
it 'should return the right response' do
put "/admin/users/nothing/delete_posts_batch.json"
expect(response.status).to eq(404)
end
end
context "when there are user posts" do
before do
post = Fabricate(:post, user: user)
Fabricate(:post, topic: post.topic, user: user)
Fabricate(:post, user: user)
end
it 'returns how many posts were deleted' do
put "/admin/users/#{user.id}/delete_posts_batch.json"
expect(response.status).to eq(200)
expect(response.parsed_body["posts_deleted"]).to eq(3)
end
end
context "when there are no posts left to be deleted" do
it "returns correct json" do
put "/admin/users/#{user.id}/delete_posts_batch.json"
expect(response.status).to eq(200)
expect(response.parsed_body["posts_deleted"]).to eq(0)
end
end
end
describe "#merge" do
fab!(:target_user) { Fabricate(:user) }
fab!(:topic) { Fabricate(:topic, user: user) }
fab!(:first_post) { Fabricate(:post, topic: topic, user: user) }
it 'should merge source user to target user' do
Jobs.run_immediately!
post "/admin/users/#{user.id}/merge.json", params: {
target_username: target_user.username
}
expect(response.status).to eq(200)
expect(topic.reload.user_id).to eq(target_user.id)
expect(first_post.reload.user_id).to eq(target_user.id)
end
end
describe '#sso_record' do
fab!(:sso_record) { SingleSignOnRecord.create!(user_id: user.id, external_id: '12345', external_email: user.email, last_payload: '') }
it "deletes the record" do
FEATURE: Rename 'Discourse SSO' to DiscourseConnect (#11978) The 'Discourse SSO' protocol is being rebranded to DiscourseConnect. This should help to reduce confusion when 'SSO' is used in the generic sense. This commit aims to: - Rename `sso_` site settings. DiscourseConnect specific ones are prefixed `discourse_connect_`. Generic settings are prefixed `auth_` - Add (server-side-only) backwards compatibility for the old setting names, with deprecation notices - Copy `site_settings` database records to the new names - Rename relevant translation keys - Update relevant translations This commit does **not** aim to: - Rename any Ruby classes or methods. This might be done in a future commit - Change any URLs. This would break existing integrations - Make any changes to the protocol. This would break existing integrations - Change any functionality. Further normalization across DiscourseConnect and other auth methods will be done separately The risks are: - There is no backwards compatibility for site settings on the client-side. Accessing auth-related site settings in Javascript is fairly rare, and an error on the client side would not be security-critical. - If a plugin is monkey-patching parts of the auth process, changes to locale keys could cause broken error messages. This should also be unlikely. The old site setting names remain functional, so security-related overrides will remain working. A follow-up commit will be made with a post-deploy migration to delete the old `site_settings` rows.
2021-02-08 05:04:33 -05:00
SiteSetting.discourse_connect_url = "https://www.example.com/sso"
SiteSetting.enable_discourse_connect = true
delete "/admin/users/#{user.id}/sso_record.json"
expect(response.status).to eq(200)
expect(user.single_sign_on_record).to eq(nil)
end
end
describe "#anonymize" do
it "will make the user anonymous" do
put "/admin/users/#{user.id}/anonymize.json"
expect(response.status).to eq(200)
expect(response.parsed_body['username']).to be_present
end
it "supports `anonymize_ip`" do
Jobs.run_immediately!
sl = Fabricate(:search_log, user_id: user.id)
put "/admin/users/#{user.id}/anonymize.json?anonymize_ip=127.0.0.2"
expect(response.status).to eq(200)
expect(response.parsed_body['username']).to be_present
expect(sl.reload.ip_address).to eq('127.0.0.2')
end
end
end