Solves #7044 #8156
Draft
Solves #7044 #8156
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…ition adds tests to show the error adds inner and outer record in ui template text for setting java format spaces solves the error by moving spaces(cs.spaceWithinMethodDeclParens() ? 1 : 0, true); to the correct place. solves apache#7043 formatter to handle record closing braces correctly squashed commits of test and solution.
Add expected values for passing test in InnerOuterRecordTest. The existing javaformatting did not conform to any standard. To make it possible to work with that file, default netbeans formatting has been applied. That leads to some white space changes only, so be it. Further added method assertLinesEqual to get better readable output when comparing expected and acyla multi line string values, as in comparing or diffing files. Part of effort to find the problem described in apache#7044
…tion the original comparison of expected and actual by collapsing all white space chars into one space is kept to keep original tests intact.
Refactoring of records is still broken as described in apache#7044. These tests should help to find the error. The error might be something small as forgetting to consider a record as a class in the same way as enum.
Case labels for RECORD were missing in two places in WorkingCopy.java
Faster without scan and wait
New tests to separate out more issues. Current recognized issues : Compact construtor in inner is not properly moved. Record header of inner is not properly moved. Current understanding is that code refactoring to source for inner record is incomplete (Not considering RECORD and or RECORD_COMPONENT).
Part of the solution ma
Part of the solution ma
Inner record has compact constructor that does a meaningful validation. Test runs without issues. Rationale: Trying to find who mishandles the inner record as described in apache#7044.
since Netbeans requires 17 to build and run, it is reasonable to expect that in source levelutils too. In RefactoringTestBase the repo is always reindexed, even when not required. As per documentation this is an expensive operation and taxes your patience. remove modifiers in variable member, so the do not show up Modifiers (private final in particular) show up in record header, making the resulting source code illegal and different from expected. The removal is done by rewriting the member tree where variables occur. make sure static fields of inner records survive field removal. ensure static fields remain in inner record. normal fields are modified by removing private final, but this should not happen to static fields. This is part of making inner records work. avoid use of raw types. pass declared type further down. Original code always passes down CLASS, now RECORD is passed down as such. It should not make a difference for class, interface, or enum. spelling RECORD should be equivalent to class, interface and enum. Add RECORD_COMPONENT as allowed element types add isInnerRecord boolean field. Some reformatting. For the reformatting: sorry. Stick to the formatting rules. For the field: Record does not fit the standard class definition mold because of the record header which has parenthesis before the first curly brace and the compact constructor that needs special attention. Somehow the information that ithe inner tis a inner record must be forwardable to the processors. The author comment is modified by some part. The @author line is substituted by some yet unknown processor, so exclude it from the result comparison to make the tests less britle because of dependence of the user property. generateOuter sounds like a boolean, but it is the name of the new inner match expected to actual after formatting. This test passes, showing that the Compact constructor in the outer survives the refactoring. formatting in expected outer cleanup. Permanently remove deprecated boolean cancelRequest cancelRequest is not only deprecated but has no use at all and was already replaced (function wide) by and atomicBoolean cancelRequested This commit cleans up the tests, may add a print that needs removing. For now: I am able to refactor inner record to outer but are still lacking the understandig of how to modify a MethodTree or the equivalent source code, and how to write a compact constructor instead. remove deprecated boolean no need to handle method special here. make test helper static to make it usable in more places.
The refactoring of inner record to outer now works as expected. The implementation tries to detect candidate 'compact constructor' even if the original constructor declaration was not. I consider that acceptable.
|
@homberghp hi! I forgot to mention it on the last PR you created #8106. It is much better to create a branch for each PR. This decouples master and any modifications to it. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This pull request solves #7044.
The work helped me to improve my understanding of the refactoring API.