Skip to content

Latest commit

 

History

History
143 lines (114 loc) · 12.6 KB

releases.md

File metadata and controls

143 lines (114 loc) · 12.6 KB
title permalink layout classes sidebar
Releases
/releases/
single
wide
nav
sidebar

5.10.0

What's New

  • Added string.Should().NotBeEquivalentTo - #1134
  • Added collectionOfStrings.Should().ContainMatch() to assert that the collection contains at least one string that matches a wildcard pattern - #1138
  • Added overloads of NotBeInAscendingOrder and NotBeInDescendingOrder that take a property expression - #1140
  • Include the index in the XPath information reported while comparing XDocuments for equivalence - #1181
  • Added NotHaveSameCount and HaveSameCount for dictionary assertions - #1178
  • Extended string.Should().Contain() and string.Should().ContainEquivalentOf() to test the number of times a phrase exists in a string using theString.Should().Contain("is a", MoreThan.Thrice()) - #1145
  • Added support for methods returning ValueTask and ValueTask<T> - #1158
  • Added support using SatisfyRespectively on all collections, including collections of string - #1201

Fixes

  • Updated the docs to clarify that CompleteWithin works on Func<Task> and not on Task directly - #1127.
  • Renamed collection assertion NotBeAscendingInOrder and NotBeDescendingInOrder to NotBeInAscendingOrder and NotBeInDescendingOrder (in a non-breaking way) - #1140
  • Collections containing nulls where not properly treated as equivalent by BeEquivalentTo - #1143
  • BeEquivalentTo now allows selecting an explicit interface member over a regular instance member when using RespectingDeclaredTypes and the expectation type is an interface - #1144
  • Fixed a crash that can occur when trying to determine the caller identity under .NET Native - #1149
  • Ensured that determining the caller identity works for namespaces that contain the phrase System and don't match on partial namespace segments - #1193
  • BeEquivalentTo on an XDocument could report the incorrect path to a self-closing XML tag - #1170
  • BeEquivalentTo on multi-dimensional arrays with empty elements could cause an internal error - #1167
  • Several assertion APIs did not include the reason in the failure messages - #1172
  • A self-closing XML element was not treated as equivalent to an empty element with the same name - #1174
  • Ensure that type assertion ThatAreUnderNamespace handles types in the global namespaces correctly - #1197
  • When comparing whether two objects are equal, a conversion should be precision preserving - #1202
  • BeEquivalentTo on normal Tuples should use structural equivalency instead of treating them as value types - #1206
  • Some platforms throw reflection exceptions when trying to use ConfigurationManager. This is now handled more gracefully - #1210
  • Reintroduced the package dependency on System.Xml.Linq for .NET 4.5/4.7 - #79
  • Treat enums and their numeric representations in structural object graph comparisons the same - #1208

Thanks to contributors Ronald Kroon, Daniel Petrov, @david-a-jetter, Lukas Grützmacher and Ben Randall.

And special thanks to Matthias Koch to switch us over to the awesome Nuke build system.

5.9.0

What's New

  • Added Match method to (nullable) numeric assertions - #1112

Fixes

  • Using a custom IAssertionStrategy with an AssertionScope did not always capture all assertion failures - #1118
  • Ensured null arguments are handled with clearer exception messages - #1117

Special thanks to contributors @liklainy and Amaury Levé.

5.8.0

What's New

  • Added thread-safety to tests using AssertionScope while running many async tests concurrently - #1091
  • Allow users to specify a custom IAssertionStrategy for the assertion scope, for instance, to create screenshots when a test fails - #1094 & #906.
  • Supports .NET Core 3.0 Preview 7 - #1107

Fixes

  • Excluding with BeEquivalentTo did not always work well with overridden properties - #1087 & #1077
  • Failure message formatting threw a FormatException when a nested message contains braces - #1092
  • Fixed confusing (Not)BeAssignableTo failure messages - #1104 & #1103
  • Fixed a potential leakage of failure messages between multiple assertions - #1105

Kudos to conklinb and Amaury Levé for notable contributions and my partner-in-crime Jonas Nyrup for some of the fixes, a lot of (internal) quality improvements and his critical eye.

5.7.0

What's New

  • Added official support for .NET Core 3 (Preview 5 or later) - #1057
  • Introduced SatisfyRespectively on collections - #1043
  • Added an alternative fluent syntax for evaluating/invoking actions - #1017
  • Added overloads of Invoking and Awaiting for different sets of generic parameters - #1051
  • NotThrowAfter is now also available for .NET Standard 1.3 and 1.6 - #1050
  • Added CompleteWithin to assert asynchronous operations complete within a time span - #1013/#1048
  • Added NotBeEquivalent for objects - #1071
  • Added WithMessage() for async exception assertions - #1052
  • Added extension methods like 100.Nanosecond() and 20.Microsecond() to represent time spans - #1069

Fixes

  • AllBeAssignableTo and AllBeOfType did not work for list of types - #1007
  • Backslashes in subject or expected result were not correctly shown in the message - #986
  • BeOfType does not attach to the AssertionScope correctly - #1002
  • Event monitoring did not detect events on interfaces - #821
  • Fix continuation on NotThrow(After) in chained AssertionScope invocations - #1031
  • Allow nesting equivalency checks in custom assertion rules when using BeEquivalentTo - #1031
  • Removed redundant use of the thread pool in async assertions - #1020
  • Improved formatting of multidimensional arrays - #1044
  • Better handling of exceptions wrapped in AggregateExceptions - #1041
  • Fix BadImageFormatException under the .NET Core 3.0 Preview - #1057
  • ThrowExactly and ThrowExactlyAsync now support expecting an AggregateException - #1046

Kudos to Lukas Grützmacher, Matthias Lischka, Christoffer Lette, Ed Ball, David Omid, mu88, Dmitriy Maksimov and Ivan Shimko for the contributions and Jonas Nyrup to make this release possible again.

5.6.0

Fixes

  • Provide opt-out to AssertionOptions(o => o.WithStrictOrdering()) - #974
  • Add collection assertion ContainEquivalentOf - #950
  • Add Should().NotThrowAfter assertion for actions - #942

Kudos to @BrunoJuchli, @matthiaslischka and @frederik-h for these amazing additions.

5.5.3

Fixes

  • Performance fixes in BeEquivalenTo - #935
  • Reverted 5.5.0 changes to AssertionScope to ensure binary compatibility - #977

5.5.2

Fixes

  • Allows BeEquivalentTo to handle a non-generic collection as the SUT - #975, #973
  • Optimized performance of IncludeMemberByPathSelectionRule - #969

5.5.1

What's New

  • Now provides a hint when strings differ in length and contain differences - #915, #907
  • Added ThrowAsync, ThrowExactlyAsync and NotThrowAsync - #931
  • Added support for Should().Throw and Should().NotThrow for Func<T> - #951
  • Added support for private protected access modifier - #932
  • Updated BeApproximately to support nullable types for both the subject and the expectation nullable - #934
  • Added async version of ExecutionTime to - #938
  • Updated NotBeApproximately to accepting nullable subject and expectation - #939
  • type.Should().Be(type) now support open generics - #954, #955

Fixes

  • Minor performance improvements to prevent rendering messages if a test did not fail - #921, #915
  • Improve performance of Should().AllBeEquivalentTo() - #920, #914
  • Improve the presentation of enums to include the value and the number - #923, #897
  • BeEquivalentTo with WithStrictOrdering produced messy failure message - #918
  • Fixes detecting checking equivalency of a null subject to a dictionary - #933
  • Fixes duplicate conversions being mentioned in the output of the equivalency API - #941
  • Comparing an object graph against IEnumerable now works now as expected - #911
  • Selecting members during object graph assertions now better handles new overrides -#960, #956

Note In versions prior to 5.5, FA may have skipped certain properties in the equivalency comparison. #960 fixes this, so this may cause some breaking changes.

Lots of kudos @jnyrup and @krajek for a majority for the work in this release.