This builds support for tracking the effect of changes to imports on
other documents, and completes our model for being able to keep
generated code up to date.
Addresses a blocking issue for FAR of types when used in user-code in a
directive. The FAR infrastructure is skipping over the directive code
because it's mapped to `#hidden`. As you can see in the code, the token
provided by the user is already included in the projection mappings.
I think we didn't do this before because we didn't expect this code to
need line numbers - it's not really debuggable, and design-time codegen
doesn't happen when you build the project.
I think it's OK for now that we don't line-map (or include) directives
based on view imports. If you trigger FAR on an `@inject ...` in an
import for instance, you'll find the reference for the view import file.
That seems pretty good, and the only cases I can really imagine it being
broken would be for go-to-definition (within a Razor view). Lets revisit
in the future based on feedback.
* Auto generate CSharp syntax tree nodes
- Auto generated syntax nodes, visitors and factories
- Renamed RazorComment to RazorCommentLiteral to avoid a conflict
- Modified the Visit() contract
- Added new SyntaxKinds
- Regenerated baselines
* Updated the tokenizer and parser to operate on SyntaxTokens directly
- The tokenizer and parser now operates directly on SyntaxTokens (Green)
- The SyntaxToken(Red) is now created when the Span is built with the correct parent and position
- All other passes that run after the parsing is complete(TagHelperRewriter etc) will operate on the Red tokens.
- There is now only one type for all SyntaxTokens. They are differentiated by their SyntaxKind.
- Added equivalence checking for tokens
- Updated test code to react
- Regenerated baselines
The new Formatter is used in debugger display, to resolve#2264
The next iteration (assuming we like this) will be to replat our testing
baseline infrastructure on top of this, and then start sharing that with
Blazor - related to #2265.
I found when implementing debugger display that I was duplicating a lot
of details between testing and this.
I also want to convert the Blazor tests to use shared infrastructure.
Right now the Blazor tests use a modified version of the Razor
infrastructure that has different features.
Initial infrastrucure for a better html parser
- Added new syntax node types that follow Roslyn's red/green pattern
- Modified the html text node in the current syntax tree to use the new
nodes
- Updated test infrastructure
- Generated baselines
- When a surrogate pair is about to be split down the middle we reduce the size of our split by 1 character. This way we split right before a surrogate pair. In the case of zero width joiners, if we split on a zero width joiner we still render valid bytes because the zero width joiner by itself can stand alone.
- Added tests for the various split cases.
#2470
- Added the necessary infrastructure to serialize and verify the syntax
tree
- Updated ParserVisitor to be overrideable
- Made CSharpVerbatimBlockTest to use baselines
- Prior to this change we weren't strict enough with where our smart indenter ran. We made the assumption that every code block could be smart indented and then Roslyn would "do the right thing". However, in nested code block scenarios we found that Roslyn and us would both indent resulting in extra newlines. These changes make the criteria for applying smart indentation a little stricter.
- Updated directive code block parsing to add a C# marker symbol in cases of an empty code block directive.
- Added unit tests to verify new smart indenter behavior.
- Updated existing tests to expect new syntax tree marker symbol for empty directive bits.
- Regenerated baselines.
#2410
- Added ability to understand valid inserts, deletes and replacements for the `@functions` directive (and any other directive that uses our extensible code block bits).
- Added unit tests.
- Updated existing tests.
- Found an issue when completing some C# items the auto-completions would impact the underlying snapshot after we'd captured the change. Fixed this by forcing a reparse when we detect that our understanding of the latest snapshot and the actual latest snapshot diverge.
#2408
- Allow identical `@using` directives in the primary document. This means that if a user types `@using System` twice in the primary document (i.e. `Index.cshtml`) then we'll generate that using statement twice; this enables proper C# IntelliSense to light up on both using statements and allows auto-completion to work because there's no more magical distinction logic for Razor code that is in the current document.
- Prior to this change if you had two identical using statements in the same document one would be in no-mans land mapped to nothing (not C#, not HTML) and the other would be mapped to C#; the second a user differentiated the two statements (i.e. adding a `.`) we'd distinctify the using statements resulting in an invalid completion.
- This PR has an end-user impact where they will now receive the normal C# warning about having duplicate using. I treated this prior behavior more as a bug because we threw away the first using directive instance which impacted editing and in general was a silly thing to correct for the user.
- Added new integration test showing how using directives are not de-duplicated in the primary document.
#1255
- This is part of a fix for #1255 - this change enables signature help in implicit expressions by improving the partial parsing. We're now smart enough about the contents of an implicit expression and attempt to balance parenthesis to determine if we should not full parse.
#1255
- We need to pre-emptively generate line pragmas to handle C# scenarios which are session based. For instance, if a C# completion session starts at a location without a pragma they capture the origination point; at which point if a user keeps typing we generate a line pragma drastically changing where the copmletion started resulting in a broken C# completion experience.
- Updated test files to reflect new behavior.
#2299
We want to have a way to specify the taghelper descriptors and imports to use while
processing a specific document.
- Added an overload to Process and ProcessDesignTime to take in a list
TagHelperDescriptors and a list of imports
- Added the corresponding CreateCodeDocumentCore overload
- Added GetTagHelpers and SetTagHelpers extension methods for
CodeDocument
- Added the necessary plumbing to use the taghelpers from the
CodeDocument when available and fallback logic.
- Added DocumentImportsTracker and updated background code generation
logic to use the new overload
- Added/updated tests
The hash code implementation here is exhaustive when it doesn't need to
be. Slimming this down to a much more reasonable set of things for perf
reasons.
The project snapshot now maintains a RazorProjectEngine as well as set
of Tag Helpers that are known for that snapshot.
Pivoted some more services to be snapshot-centric.
Also added the ability to track .cshtml documents to the project system.
For now most components just ignore document changes.
This change will allow someone extending Razor to use generic type
parameters in generated classes.
There's no user-level extensibility provided here yet, as in there is no
language support for adding type parameters.
Ajay Bhargav Baaskaran
Ryan Nowak
Nate McMaster
N. Taylor Mullen
Pranav K
Ajay Bhargav Baaskaran
Marcin Burak
Ryan Nowak
Artak
Daniel Eisterhold
Ryan Brandenburg