refactor inlining code and add tests for it #987
Merged
Conversation
This file contains hidden or 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
guillep
requested changes
Jun 19, 2025
Comment on lines
1032
to
1033
| inliner := SLInlinerStrategy new | ||
| codeGenerator: self; | ||
| yourself. |
There was a problem hiding this comment.
Ok so this PR already replaces the old one by the new one right?
There was a problem hiding this comment.
yes, but just because i wanted to see the CI for the new one
Comment on lines
2
to
8
| #name : 'SLInliner', | ||
| #superclass : 'Object', | ||
| #instVars : [ | ||
| 'codeGenerator', | ||
| 'currentMethod', | ||
| 'sLNodeAnnotatorVisitor' | ||
| ], | ||
| #category : 'Slang', | ||
| #package : 'Slang' |
There was a problem hiding this comment.
What's the difference between an inliner and an InlinerStrategy?
Comment on lines
646
to
677
| TStatementListNode >> removeFirst [ | ||
|
|
||
| | first | | ||
| first := self firstNonCommentStatement. | ||
| statements := statements select: [ :e | e ~= first ] | ||
| ] |
There was a problem hiding this comment.
I don't particularly like the collection interface in the statement list...
Also, removeFirst makes me think that it will remove the first, but it's removing the first non comment, right?
There was a problem hiding this comment.
yes, it needs a better name
This was referenced Jun 19, 2025
guillep
previously approved these changes
Jul 11, 2025
guillep
reviewed
Jul 11, 2025
guillep
reviewed
Jul 11, 2025
|
@RenaudFondeur can you move to another PR the new tests that show regression bugs but that are not fixed in this PR? |
add test for inlining handle potential bug now that we have more generated comments
add a visitor to properly determine what is an expression and what is a statement add the tests for the visitor, it is not used yet
start to integrate the visitor in the inlining to get exitVar and direct return clean some code now unused add tests
…lly buggy hack change the code to fix it
add a test for a bug in both inlining strategy add handle of comments in one more checks
small refactor of inlining
…restore the tests for it + 2 other tests that shouldn't have been supress
…e bug with the non-necessary nil
Cleanup dead code and comments
guillep
force-pushed
the
betterInliningStable
branch
from
eb7c214 to
bdda5f9
Compare
|
Thanks @RenaudFondeur ! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
We now have two ways to do inlining, accessible via their own strategy (SLInlinerStrategy or SLOldInlineStrategy) in doBasicInlining:.
The old inlining hasn't changed, it has currently 8 failing tests for existing bugs.
The refactored one has 6 failing tests.
I managed to solve some of the remaining bugs, but it ended up changing more things than expected so it will come in another PR.
I also couldn't exactly reproduce yet two differences in the generated code from the two strategies which appear close to existing test cases.