Fabian =?utf-8?q?Keßler?= <[email protected]>, Fabian =?utf-8?q?Keßler?= <[email protected]>, Fabian =?utf-8?q?Keßler?= <[email protected]>, Fabian =?utf-8?q?Keßler?= <[email protected]> Message-ID: In-Reply-To: <llvm.org/llvm/llvm-project/pull/[email protected]>
Febbe wrote: > > My plan was just to move the existing "ast view" treeview widget to a new > > menu, add a text bar above, and allow it to show multiple matches for a > > first prototype. > > I'm wondering, what is the use case for showing the entire AST subtree (and > indeed, the protocol sending an entire `ASTNode`) for each match? > > Might it be sufficient to just send a text file range for each match (much > like ordinary search results), and then if the user wants to see the AST for > a particular result they can select the range and invoke "Show AST"? Interesting question, the tree of ASTNode's contain your search result and should also contain the source code range. For the most cases, just the source code range might be sufficient. But, you explicitly searched for a matching AST, so you might also view that AST result. It was also an idea, to fullfill the last point of https://github.com/clangd/clangd/issues/1945#issue-2149756432 : > * Optionally replace them with a sub AST This would only work, if you can modify parts of the AST. Whether this is text based in the end might not matter, but in the intermediate state, e.g. while replacing static_cast's with a specific custom cast function, you need the subranges of each node. https://github.com/llvm/llvm-project/pull/156090 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
