We have all of these executors but they aren't really
documented/supported for extensibility today. This change introduces a
pattern for action result executors so we can make them extensible.
- #6648
- a different take on #4871
- `DateTime` can also round-trip `DateTimeKind.UTC` with `[DataType("datetimeoffset")]` or `[UIHint("datetimeoffset")]`
- since they're now handled differently by default, add more `DateTime` tests
- expand tests involving `Html5DateRenderingMode.CurrentCulture`
nits: make VS-suggested changes to files updated in this PR
If you give ModelExpressionProvider a lambda with a private property
you'll end up here. This wasn't common before, but it seems like users
are more likely to try it with pages.
Model Metadata and Model Binding don't handle private properties, so
supporting it in Model Expressions seems less than useful.
This isn't a breaking change because this case would have resulted in a
null-ref. Addresses #6400
- #5655
- also make `ExpressionTextCache` more robust for defence-in-depth
nits:
- two `null` expression nodes are equal
- declare data properties as `TheoryData<T>`
* This allows specifying all it's dependencies rather than for consuming projects to do this
* Remove unused APIs
* Fix weird downgrade warnings that show up due to P2P references
- #5347
- inconsistent bounds checking caused problems after `ArrayPool<char>` fell back to `new char[2048]`
- would fail a `Debug` assertion in Debug builds and loop endlessly in Release builds
- change to `CacheTagHelper+CharBufferHtmlContent` is for correctness only
- always uses a `CharArrayBufferSource` and that returns arrays of the exact size requested
- #3918
- precompute size of `StringBuilder` in `ExpressionHelper`
- reduce `string` allocations in `ViewDataEvaluator`
- also get rid of `Enumeration` state machines
- reduce the size of a few objects; use more expression-valued properties
- e.g. don't store `_modelType` in `ModelExplorer`
- add `EmptyArray<TElement>`; make empty arrays consistently `static`
- avoid `string.Split()` in HTML and tag helpers
nits:
- make `ExpressionHelperTest` tests more stringent
- correct `Message` for an `ArgumentNullException`
- remove excess `using`s in classes I touched (but often ended up leaving otherwise unchanged)
- improve doc comments
- remove `ToString()` call on a `string`
- avoid encoding `string.Empty`
- fix test file name
- remove useless variables
- correct spelling
- improve whitespace
- #5317
- previously worked only because `TagBuilder` cannot be empty and `HtmlString` overrides `ToString()`
- `TagBuilder.ToString()` (now the type's `FullName`) is also never empty
- copy `NullHtmlEncoder` from Razor and give it a better name (`PassThroughHtmlEncoder`)
- not available in this project and (from its namespace) not intended for general use
* Added consistent model property/tests to ViewResult, PartialViewResult, ViewComponentResult. This resolves#4813.
* Removed unnecessary model asserts
* Removed redundant model checking
- Simplify things that used to rely on HtmlTextWriter. This behavior is
now the default.
- Simplify a whole mess of Razor TextWriter code.
- Integration test for merging of TagHelper buffers back into the 'main'
buffer.
This change removes a layer of abstraction. These validators now just do
what they do now without creating a bunch of intermediate objects.
ModelClientValidationRule has been removed, and client validations
manipulate the attributes collection of a tag directly.
These changes are aimed at significantly improving the performance of
MVC/Razor when a large amount of content is in play or a large number of
TagHelpers are used.
A few issues addressed:
- Buffer sync writes after a flush has occurred so that we can write them
asyncronously. The issue is that an IHtmlContent can only do sync
writes. This is very bad for Kestrel in general. Doing these writes
async is better for our overall perf, and the buffer that we use for it
is from the pool.
- 'Flatten' ViewBuffers when possible. A page with lots of TagHelpers can
end up renting a ViewBuffer and only write 2-3 things into it. When a
ViewBuffer sees another ViewBuffer we can either steal its pages, or
copy data out and 'return' its pages. This lets us use 3-4 buffers for a
large Razor page instead of hundreds.
Pranav K
Ryan Nowak
Kiran Challa
Doug Bunting
bchavez
Nate McMaster
Jass Bagga
Ryan Brandenburg
Ajay Bhargav Baaskaran
Yves57
Medeni Baykal
Crystal Qian
jacalvar
N. Taylor Mullen
mnltejaswini