- Updated `TagHelper` errors to no longer highlight the entire tag as an error, instead just the tag name is marked as an error. This is now consistent with nested tags in `@{ ... }` errors.
- Updated `RazorError` and corresponding error logging constructs to disallow creation without providing lengths.
- Updated `TagHelperDescriptorResolver` and related classes to properly determine assembly locations within directives. This allows for exact error locations in the `@addTagHelper` directive.
#386
- Renamed CodeGenerators to ChunkGenerators.
- Updated location of TestFiles from TestFiles/CodeGenerator/CS/{Output|Source} => TestFiles/CodeGenerator/{Output|Source}.
- Removed ChunkTree test; it was a legacy test used to experiment with Razor rendering (not a real test).
- Removed CSharpRazorCodeGenerator; it's now replaced with RazorCodeGenerator. It was an empty class that did nothing.
- Updated ChunkBlock => ParentChunk. Also updated several patterns throughout the code base that referenced these blocks as blocks and not parents.
- Moved Chunks and ChunkGenerators into the Chunks/Chunks.Generators namespace/folder structure. Updated test project to reflect the same.
- Moved CodeBuilders and CodeVisitors to the CodeGeneration/CodeGeneration.Visitors namespace/folder structure. Updated test project to reflect the same.
- Moved several TagHelper assets outside of their own namespaces and into Razors more general structures; such as CodeGeneration and Chunks/Chunks.Generators.
#140
- add `[NotNull]` in some `public` or `protected` callers as well
- add `[NotNull]` in `SeekableTextReader` constructors
- add `where TSymbolType : struct` to replace incorrect `null` checks
- remove `T` type parameters in changed files e.g. change to `TWriter`
- remove tests of removed code
nits:
- change `TextReaderExtensions` to consistently call other extensions as statics
- wrap some long doc comments
This is a perf improvement of about 500ms for our razor code generation
benchmark on my dev box. That's about .8% of the overall execution time of
this benchmark.
This change will remove a bunch of unnessary allocations from the
parsing/code-generation path, and should improve responsiveness.
For reference Enum.HasFlags performs boxing of the enum value, and then
does a type comparison to see if the types are the same. This is
significantly more costly than a normal bitwise and comparison, and it
results in allocations.
Pranav K
N. Taylor Mullen
Doug Bunting
Ryan Nowak
Andrew Peters
David Fowler