* Simplify or optimize regexes with polynomial time worst cases
* PR feedback & cleanup
Co-authored-by: David Michon <dmichon-msft@users.noreply.github.com>
* Use builtin scanner function for checking whitespace in fallback method (its faster)
Co-authored-by: David Michon <dmichon-msft@users.noreply.github.com>
* Revert "Revert "Editor support for link tag (#41877)" (#43302)"
This reverts commit 451d4354b94e533907ef3b6362a89435f67577a3.
* Fix parsing @link at end of comment
* Parse comments as string when no @link occurs
* fix lint
* Initial scribbles
* Compiles but provides spans instead of location pairs
Probably need to fork the services/server types and provide a conversion
with Session.toFileSpan. Not sure where to put the conversion.
* Switch to DocumentSpan
In theory this is already better supported, but not sure practise bears
that out.
* Builds w/protocol types + conversions
* cleanup:better names and scrub TODOs
* fix test harness too
* Misc
1. Simplify protocol after talking to @mjbvz.
2. Add more tests.
3. Initial notes about where to add parsing.
* Parse and store links in the compiler
The text of the link is still stored in the comment text, but that's now
kept in an object instead of just a string. Each link has the parse for
the entity reference, if there is one.
Needs lots more tests -- this just makes all the existing jsdoc tests
pass.
* more tests and some fixes
* Fix other failing tests
* fix bad merge
* polish parser
* improve names and array types
* slight tweaks
* remove some done TODOs
* more tests + resulting fixes
* add+fix cross-module tests
* Support `@see {@link`
Plus find-all-refs support equivalent to @see's.
* add server test
* Make comments actually part of the AST
* Add span for link.name in language service/protocol
* Make checker optional in getJSDocTags
Also change to JSDocCommentText from JSDocCommentComment
* Use getTokenValue instead of getTokenText
Measure twice, slice once
* Add missing support for top-level links
The language service and protocol were missing support for top-level
links. This commit adds that plumbing.
* add string back to comment type in node constructors
* Full parse of link tags and jsdoc comment text
- Doesn't pass fourslash yet, I'm going to switch to baselines for
failures there.
- Still needs some work on the protocol to convert file+offset to
file+line+offset.
* fix lint
* Fix missing newlines in inferFromUsage codefix
* Parse jsdoc comments as text node/link array
And switch to line+character offsets in the protocol
* Fix fourslash tests
Mostly ones that can't be baselined, but I switched a couple more over
to baselines
* Improve types and documentation
* Test+fix @link emit, scrub other TODOs
* update API baselines
* test that goto-def works with @link
* Split link displaypart into 3
One for link prefix and suffix, one for link name, and one for link
text.
* update baselines
* Provide JSDocTagInfo.text: string to full clients by default
Instead of upgrading them to displayparts.
* Real server tests
* Disambiguate {@link} and @param x {type}
They are ambiguous; previously the parser preferred the type
interpretation, but will now look ahead and parse links instead when the
prefix is `{@link`.
* Add explanatory comment in test
* fix location in richResponse in protocol
* update API baseline
* Address PR comments
1. Add a cross-file goto-def test.
2. Switch from per-message args to UserPreference.
* use arraysEqual from core
* Refactor node factory API, use node factory in parser
* Move UnparsedSource nodes to factory
* Make most Node properties read-only
* Make pos/end/parent and JSDoc 'comment' read-only
* Update function/constructor-type factories
* Remove treeStateObserver
* Simplify Debug.deprecate
* Remove unused factory methods, simplify lazy factory methods
* Fix base factory used for source file updates
* Update test baseline due to merge from master
* Rename factory methods to be more consistent (#39058)
* migrate nullish coalescing commit
* add more test case
* add branch type check test
* add more tests
* fix nullish precedence
* update public api
* add rescan question question token to fix regression
* update public api baseline
* Added tests that emit for nullish coalescing operator conforming with grammar restrictions when assertions are used.
* Fixed emit to hoist temporary variables (they previously went undeclared).
Added tests to ensure calls and property accesses are only called once.
* use not equal to null
* rename factory
* add grammar check
* fix more cases
* Fix handling of nullish coalescing oprator in expando objects.
* Fixed classifier to support ?? operator.
* update baseline
* accept baseline
* fix review
* update emitter and more testcase
* update control flow
* make linter happy
* update libs
* avoid unnecessary assert
* fix typooo
* Fixes for control-flow analysis
It was overly permissive and ended up making a mess of C#-style
comments:
`/// <summary>Text</summary>`
Now it checks the element name. Attribute names remain unchecked.
* Enable '--strictNullChecks'
* Fix API baselines
* Make sys.getEnvironmentVariable non-nullable
* make properties optional instead of using `| undefined` in thier type
* reportDiagnostics should be required
* Declare firstAccessor as non-nullable
* Make `some` a type guard
* Fix `getEnvironmentVariable` definition in tests
* Pretend transformFlags are always defined
* Fix one more use of sys.getEnvironmentVariable
* `requiredResponse` accepts undefined, remove assertions
* Mark optional properties as optional instead of using `| undefined`
* Mark optional properties as optional instead of using ` | undefined`
* Remove unnecessary null assertions
* Put the bang on the declaration instead of every use
* Make `createMapFromTemplate` require a parameter
* Mark `EmitResult.emittedFiles` and `EmitResult.sourceMaps` as optional
* Plumb through undefined in emitLsit and EmitExpressionList
* `ElementAccessExpression.argumentExpression` can not be `undefined`
* Add overloads for `writeTokenText`
* Make `shouldWriteSeparatingLineTerminator` argument non-nullable
* Make `synthesizedNodeStartsOnNewLine` argument required
* `PropertyAssignment.initializer` cannot be undefined
* Use one `!` at declaration site instead of on every use site
* Capture host in a constant and avoid null assertions
* Remove few more unused assertions
* Update baselines
* Use parameter defaults
* Update baselines
* Fix lint
* Make Symbol#valueDeclaration and Symbol#declarations non-optional to reduce assertions
* Make Node#symbol and Type#symbol non-optional to reduce assertions
* Make `flags` non-nullable to reduce assertions
* Convert some asserts to type guards
* Make `isNonLocalAlias` a type guard
* Add overload for `getSymbolOfNode` for `Declaration`
* Some more `getSymbolOfNode` changes
* Push undefined suppression into `typeToTypeNodeHelper`
* `NodeBuilderContext.tracker` is never `undefined`
* use `Debug.assertDefined`
* Remove unnecessary tag
* Mark `LiteralType.freshType` and `LiteralTupe.regularType` as required
* Set startPos at EOF in jsdoc token scanner to node end positions for nodes terminated at EoF are right
* More complete nonwhitespace token check, fix syntactica jsdoc classifier
* Use loop and no nested lookahead
* Do thigns unrelated to the bug in the test
* Fix typo move return
* Patch up typedef end pos
* Fix indentation, make end pos target more obvious
* Add typedef declaration space, unify typedef name gathering, strengthen errorUnusedLocal
* Bonus round: make jsdoc presence way mroe typesafe
* Be exhaustive in nameForNamelessJSDocTypedef
* Remove nonrequired casts
* Replace more casts with guards
* Cannot be internal
* Debug.fail returns never, assert never no longer needs unreachable throw to satisfy checker
* Rename type
* Add replacement message as in 18287
* And lint rule to check that `Debug.assert` calls do not eagerly interpolate strings
* Use more specific 'assert' functions to avoid callbacks
* Respond to PR feedback
* Add 'name' property to Identifier
* Rename to unescapedText
* Rename 'id.text' to 'id.escapedText'
* Rename 'id.unescapedText' to 'id.text'
* Make escapeIdentifier and unescapeIdentifier do nothing
Now Typescript supports the creation of anonymous types using successive
`@param` lines in JSDoc:
```js
/**
* @param {object} o - has a string and a number
* @param {string} o.s - the string
* @param {number} o.n - the number
*/
function f(o) { return o.s.length + o.n; }
```
This is equivalent to the Typescript syntax `{ s: string, n: number }`,
but it allows per-property documentation, even for types that only need
to be used in one place. (`@typedef` can be used for reusable types.)
If the type of the initial `@param` is `{object[]}`, then the resulting
type is an array of the specified anonymous type:
```js
/**
* @param {Object[]} os - has a string and a number
* @param {string} os[].s - the string
* @param {number} os[].n - the number
*/
function f(os) { return os[0].s; }
```
Finally, nested anonymous types can be created by nesting the pattern:
```js
/**
* @param {Object[]} os - has a string and a number
* @param {string} os[].s - the string
* @param {object} os[].nested - it's nested because of the object type
* @param {number} os[].nested.length - it's a number
*/
function f(os) { return os[0].nested.length; }
```
Implementation notes:
1. I refactored JSDocParameterTag and JSDocPropertyTag to
JSDocPropertyLikeTag and modified its parsing to be more succinct. These
changes make the overall change easier to read but are not strictly
required.
2. parseJSDocEntityName accepts postfix[] as in `os[].nested.length`,
but it doesn't check that usages are correct. Such checking would be
easy to add but tedious and low-value.
3. `@typedef` doesn't support nested `@property` tags, but does support
`object[]` types. This is mostly a practical decision, backed up by the
fact that usejsdoc.org doesn't document nested types for `@typedef`.
* Created a branded type for escaped strings
Then flowed it throughout the compiler, finding and fixing a handful of
bugs relating to underscore-prefixed identifiers in the process.
Includes a test for two cases noticed - diagnostics from conflicting
symbols from export *'s, and enum with underscore prefixed member emit.
* Correctly double underscores WRT mapped types
* Add fourslash tests for other fixed issues
* use function call over cast
* Update forEachEntry type accuracy
* Just use escaped names for ActiveLabel
* Remove casts from getPropertyNameForPropertyNameNode
* This pattern has occurred a few times, could use a helper function.
* Remove duplicated helper
* Remove unneeded check, use helper
* Identifiers list is no longer escaped strings
* Extract repeated string-getting code into helper
* Rename type and associated functions
* Make getName() return UnderscoreEscapedString, add getUnescapedName()
* Add list of internal symbol names to escaped string type to cut back on casting
* Remove outdated comments
* Reassign interned values to nodes, just in case
* Swap to string enum
* Add deprecated aliases to escapeIdentifier and unescapeIdentifier
* Add temp var
* Remove unsafe casts
* Rename escaped string type as per @sandersn's suggestion, fix string enum usages
* Reorganize double underscore tests
* Remove jfreeman from TODO
* Remove unneeded parenthesis
* Support completions for JSDoc @param tag names
* Undo change to finishNode
* Don't include trailing whitespace in @param range; instead, specialize getJsDocTagAtPosition
