939 Commits (a89ed0b0ebbacc2591870f1a33fd2011f507f413)

Author	SHA1	Message	Date
Alex Coles	ae59a6985a	Teach User.authorize_within ActiveRecord::Relation ... Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	81de05e5fe	Remove deprecated Relation#all in specs ... Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	b2790b8db1	Migrate deprecated .update_all with conditions ... Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	0f9240037a	Fix TimeEntry cleanup on work package bulk delete ... Fix `WorkPackage.cleanup_associated_before_destructing_if_required` semantics so that the first argument can either be an individual `WorkPackage` (as the spec originally expected) or a collection (as in actual usages). Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	ce93701029	Fix syntax (w/Rubocop) in specs ... Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	57618b25ec	Replace dynamic finder usages with #find_by, etc. ... This patch replaces all dynamic finders, for the sake of consistency, although only some methods are deprecated. See: https://github.com/rails/activerecord-deprecated_finders#active-record-deprecated-finders * Revert some `User#find_by_login` usages in cuke steps accidentally removed in 74228b59. User Story # 20325 Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Jan Sandbrink	620c55bf3d	change the default preference for mail hiding	10 years ago
Oliver Günther	cb123c83be	Fix and specs for missing plus in login attribute ... Fixes the merge issue from 105934a5f9	10 years ago
Alex Coles	abf4a011f0	Fix copying of versions on Project copy ... Adds a more robust solution for copying closed versions and associating with copied work packages (see original patch 10ec20a8). Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	d060e654a5	Use vanilla Rails validations for CustomField#name ... Globalize should support regular Rails validations. Additionally `validate_name` was no longer working as expected, possibly due to changes in `CustomField::Translation` lifecycle. * Do not test translation implementation details in CustomFieldController spec. Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Jan Sandbrink	c075a7b401	only allow simple space in login ... also add specs to verify that	10 years ago
Alex Coles	3276c4ebb5	Add/update Copyright headers in specs ... Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Jan Sandbrink	8e1970b9bb	add roundtrip tests for more datatypes	10 years ago
Jan Sandbrink	27d6abfa35	add integrative tests for typed_value for booleans ... first charge of tests. I intend those to fail for postgres, but can't verify that locally, so travis will have to do it for me. If you see this build failing for the newly added specs: This is intentional!	10 years ago
Jan Sandbrink	f9b242884d	let custom values accept their native type ... - add specs for strategies to parse their typed representation as valid (instead of only their stringly typed counterpart) - add specs to int and float strategy regarding accepting eachother - fix int custom values accepting float inputs (which would work, but omit the fractional part) - fix date custom values not accepting on Date	10 years ago
Alex Coles	2efd6f4e45	Move legacy specs under spec/legacy ... Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	19ed4dc9ab	Replace more deprecated shoulda matchers in specs ... Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Jan Sandbrink	cdeb533904	allow to disable modules in project factory ... replaces specific "wiki-only" trait that existed before	10 years ago
Jan Sandbrink	88b99db223	only respond to valid custom_field_n accessors ... it is not enough that the CF exists, it actually has to be a CF that can be used on that work package	10 years ago
Jan Sandbrink	ad5b7b3c62	make spec localization independent	10 years ago
Jan Sandbrink	122384b8f2	use "be" instead of "eql"	10 years ago
Jan Sandbrink	05bdba6a29	fix preview of true custom field in form ... :typed_value would not return true if the internal value was set to true (it only ever expected it to be '1')	10 years ago
Jens Ulferts	13e0940171	only display scm options if they are configured	10 years ago
Jens Ulferts	fce72938e3	prevent path traversal via filesystem scm ... Setting up a filesystem scm now requires to specify a whitelist via the configuration	10 years ago
Alex Coles	ebbc9e3d7d	Prefer #not_to over #to_not in specs ... Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	33748ede31	Remove by qualifier in neg. #change expectation ... No longer supported. Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	6cd0cad516	Remove #errors_on usages in specs ... This behaviour would now require `rspec-collection_matchers`. See https://github.com/yujinakayama/transpec#note-about-expectmodelto-havenerrors_onattr Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alexander Bach	15f52bcdc6	Added validation for estimated hours values	10 years ago
Jan Sandbrink	857bcf4636	some refactoring around the WP schema ... - schema does not decide anymore whether to include allowed values (solely decided by "form_embedded" of representer) - schema will answer as many "assignable_*" methods as possible - work package does not need those methods anymore (except few cases)	10 years ago
Alex Coles	c13b9a59ff	Convert specs to RSpec 3.2.0 syntax with Transpec ... This conversion is done by Transpec 3.0.7 with the following command: transpec -f * 11 conversions from: it { should ... } to: it { is_expected.to ... } * 3 conversions from: be_true to: be_truthy * 1 conversion from: be_false to: be_falsey * 1 conversion from: failure_message_for_should { } to: failure_message { } * 1 conversion from: failure_message_for_should_not { } to: failure_message_when_negated { } * 1 conversion from: obj.stub(:message) to: allow(obj).to receive(:message) * 1 conversion from: obj.stub_chain(:message1, :message2) to: allow(obj).to receive_message_chain(:message1, :message2) For more details: https://github.com/yujinakayama/transpec#supported-conversions	10 years ago
Alex Coles	3785fdc9ed	Use #describe, #context with String arg not symbol ... The semantics of `describe` have changed in RSpec 3: passing a symbol rather than a string will change the value of the described object. See discussion here: https://github.com/rspec/rspec-core/issues/1114 Signed-off-by: Alex Coles <alex@alexbcoles.com>	10 years ago
Alex Coles	944b89d83e	Convert specs to RSpec 3.2.0 syntax with Transpec ... This conversion is done by Transpec 3.0.7 with the following command: transpec -f * 34 conversions from: it { should ... } to: it { is_expected.to ... } * 14 conversions from: obj.stub_chain(:message1, :message2) to: allow(obj).to receive_message_chain(:message1, :message2) * 7 conversions from: obj.stub(:message) to: allow(obj).to receive(:message) * 4 conversions from: be_true to: be_truthy * 3 conversions from: obj.should_receive(:message) to: expect(obj).to receive(:message) * 2 conversions from: failure_message_for_should { } to: failure_message { } * 2 conversions from: failure_message_for_should_not { } to: failure_message_when_negated { } * 1 conversion from: Klass.any_instance.stub(:message => value) to: allow_any_instance_of(Klass).to receive_messages(:message => value) * 1 conversion from: be_false to: be_falsey * 1 conversion from: obj.should_not_receive(:message) to: expect(obj).not_to receive(:message) * 1 conversion from: stub('something') to: double('something') * 1 addition of: RSpec.configure { |c| c.infer_spec_type_from_file_location! } For more details: https://github.com/yujinakayama/transpec#supported-conversions	10 years ago
Jan Sandbrink	3168182499	"fix" a previously "good" spec	10 years ago
Jan Sandbrink	348d3c746e	fix submission of "false" as CF value ... it was not possible to set a custom field to false via APIv3. Apparently the way the value is set in other places does not involve "real" boolean types	10 years ago
Markus Kahl	de309f0eee	polish	10 years ago
Markus Kahl	104b70cc62	work package notification mail specs	10 years ago
Jan Sandbrink	925a1fb706	don't pick the (random) last journal ... pick the first (and hopefully only) journal that has customizable journals	10 years ago
Jan Sandbrink	264c58980b	change expectation back to falsey ... be_false == be_falsey, both expect falsey results alternative would be "be false", which expects real false... *confused*	10 years ago
Jan Sandbrink	9986466e15	use direct CF accessors in CF injector	10 years ago
Jan Sandbrink	1aa234ae58	fix another broken spec	10 years ago
Jan Sandbrink	0dcd0a0c84	fix weird specs ... probably only worked because of unintended validation error from automatic validation: - testing to pass nil as value, though nil-values were explicitly ignored - assuming a single journal entry, though the WP is saved twice really needed to be fixed, because of new implementation: - valdiation errors now indicate CF id and are more specific	10 years ago
Jan Sandbrink	25dec31c5a	extract some case statements into fancy strategies ... - 17 new files - lots of new lines of code - specs! - same functionality as before	10 years ago
Alexander Bach	f403d6aaaa	Increase category validation strictness	10 years ago
Alexander Bach	03494dae0a	Some style improvements	10 years ago
Alexander Bach	92618be3ed	Added tests for category writing	10 years ago
Jens Ulferts	1bf67b644e	apply shared acts_as_watchable spec to all watchable models	10 years ago
Jens Ulferts	a70bbd7045	improve performance of watcher determination	10 years ago
Jens Ulferts	fdd3bd80b6	fix test_unit tests	10 years ago
Florian Kraft	1a2ea133f2	fix tests for wiki and permitted params	10 years ago
Jens Ulferts	c65c60c9b7	style work package form	10 years ago