-
Notifications
You must be signed in to change notification settings - Fork 1
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Remove some colossal strings in the compiler tests #1137
Remove some colossal strings in the compiler tests #1137
Conversation
Skip in case we're running tests to remain compatible with JDT
…n count Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
Javac sometimes add virtual nodes to the tree (usually with endPos == -1). Ignore then when computing diagnostic position.
Log instead
This reverts commit 13618d9 as upstream as applied similar/better change when merging Java 23 support
* Treat DCRawText as raw text (don't both splitting lines) * Some fixes in TagElement range correction
Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
…ings is set to true Signed-off-by: Rob Stryker <[email protected]>
…r in parsing or resolution Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
…tion Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]> Add helpful comments in code, and add test annotations Signed-off-by: Rob Stryker <[email protected]> More annotations Signed-off-by: Rob Stryker <[email protected]> More test annotations Signed-off-by: Rob Stryker <[email protected]> Missing import Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
Signed-off-by: Rob Stryker <[email protected]>
This causes 6 regerssions because 'possible this escape' is now reported. Signed-off-by: David Thompson <[email protected]>
- fix a bug related to ordering of tag names - get `@param| ` completion to work as expected - use `DOMCompletionUtil.findParent` instead of only checking the direct parent when calculating the node to use as context for completion (in the case of nodes within `Javadoc`) - split out `MemberRef` and `MethodRef` to their own completion cases - for now these are stubs - move the logic for type completion for `@see` and `@link` under the TagElement section, as it should be handled differently than the "default completion": - improve handling of qualified references - for the time being, suggest nothing when completing a type from a specific module. i.e. `@see java.base/java.util.|` no longer gives any results. Signed-off-by: David Thompson <[email protected]>
Also needed but missed.
Note that there are two code paths: one through `MemberRef` and another through an `@link` or `@see` `TagElement`. This is because the AST conversion tests expect the following to be a `TextElement` instead of a `MemberRef`: ```java /** * @see MyClass# */ ``` (and also:) ```java /** * @see # */ ``` However, the completion tests expect completion for MyClass' members to work in this case. Fixes eclipse-jdt#1103 Signed-off-by: David Thompson <[email protected]>
* Type parameters * Annotations * Relevance
* Support type restriction for SwitchCase * Support for QualifiedType * Fix CompletionContext.getExpectedTypesSignatures()
* Return more proposals * Do no repeat extra import in proposals
- Fixes for the relevance numbers for methods in javadoc - Suggest constructors (including the default constructor) in `@see` and `@link` - Do not show completion when completing parameter name in method reference in Javadoc - when referencing a method from Javadoc, and the cursor is in the method parameters, specifically right after a `(` or `,`, complete the entire method signature This should fix 37 test cases Signed-off-by: David Thompson <[email protected]>
As there are such cases often I raised it to error and enable the save action.
Makes it slightly easier to use javac branch on big codebases as hundreds of explicitly suppressed such warnings are now properly not displayed anymore.
javac is incapable of compiling these files due to the large string literals. By recreating the contents using `StringBuilder`, the files can be compiled with javac. Fixes eclipse-jdt#1136 Signed-off-by: David Thompson <[email protected]>
Should this also be contributed upstream? |
Ideally, yes. I don't know if they'll accept it, because this is only a compiler error under javac and not ECJ, and for now, the tests are only ever compiled using ECJ. However, I'll open a PR for it. I already have a similar PR open for the builder test suite. Who should I ask to review it? |
I think if you can demonstrate with a simple CLI that |
I don't think they care about whether it compiles under javac, since they only ever compile the test suite under ECJ. However, I think there is a solid argument to be made that these string literals make the tests hard to read and will be much easier to understand if they are generated by a |
9c17bb6
to
4516c31
Compare
4516c31
to
329b8ed
Compare
merged upstream |
javac is incapable of compiling these files due to the large string literals. By recreating the contents using
StringBuilder
, the files can be compiled with javac.Fixes #1136