"origin" locations in Yul AST
#12005
Conversation
| ret["src"] = formatLocationWithFileIndex(_debugData->nativeLocation, m_sourceIndex); | ||
| if (_debugData->originLocation.isValid()) | ||
| { | ||
| if (_debugData->nativeLocation.sourceName == _debugData->originLocation.sourceName) | ||
| ret["origin"] = formatLocationWithFileIndex(_debugData->originLocation, m_sourceIndex); | ||
| else | ||
| ret["origin"] = formatLocationWithFileName(_debugData->originLocation); | ||
| } |
There was a problem hiding this comment.
The problem with this PR is that we don't have indexes for the original Solidity files when compiling Yul.
This does work because currently we do not support outputting Yul AST (#9590) but will break when we implement it so I don't think this is a good solution.
cameel
force-pushed
the
origin-location-in-yul-ast-json
branch
4 times, most recently
from
28b4092 to
2e3b316
Compare
|
I think we need to provide the source-name-to-index-mapping to the asm json converter in the same way we do it for the ASTJsonConverter - does this not work? |
|
But that will work only if we have access to Solidity sources. When we're importing inline assembly AST we do, but this is the case where |
cameel
force-pushed
the
origin-location-in-yul-ast-json
branch
from
2e3b316 to
fff703e
Compare
@cameel was this solved now? |
|
Not really. I think this PR does not really make sense without implementing Yul AST output first (#9590). These origin locations will never show up for inline assembly. Originally I wanted to implement #9590 and then put this PR on top of that. But I got into other PRs and didn't really have time for this one. I could close it but @chriseth seems to think this should actually work. I didn't want to close it without clarifying that first. |
|
I'm going to close this. I definitely won't be working on this in the near future and like I said above, I'm not even sure if this PR even makes sense - it's only needed for Yul ASTs and we don't support printing those at this point. |
Originally requested in #11998 (comment).
Depends on #11998.
Related to #11043.