fix: Refactor the way of collecting LSP information #110
Conversation
…de of the definintion
…nown or undefined parameter
…factor # Conflicts: # parser_library/src/analyzer.cpp # parser_library/src/context/hlasm_context.cpp # parser_library/src/workspaces/workspace.h # parser_library/test/context/data_attribute_test.cpp # parser_library/test/context/macro_test.cpp # parser_library/test/context/ord_sym_test.cpp # parser_library/test/copy_mock.h # parser_library/test/expressions/expr_mocks.h # parser_library/test/processing/lookahead_test.cpp # parser_library/test/processing/opsyn_test.cpp
| std::unordered_map<parser_library::completion_item_kind, lsp_completion_item_kind> completion_item_kind_mapping() | ||
| { | ||
| using namespace parser_library; | ||
| return { { completion_item_kind::mach_instr, lsp_completion_item_kind::function }, | ||
| { completion_item_kind::asm_instr, lsp_completion_item_kind::function }, | ||
| { completion_item_kind::ca_instr, lsp_completion_item_kind::function }, | ||
| { completion_item_kind::macro, lsp_completion_item_kind::file }, | ||
| { completion_item_kind::var_sym, lsp_completion_item_kind::variable }, | ||
| { completion_item_kind::seq_sym, lsp_completion_item_kind::reference } }; | ||
| } |
There was a problem hiding this comment.
This could be a single static const map.
| { "insertText", item.insert_text() } }); | ||
| } | ||
| to_ret = json { { "isIncomplete", completion_list.is_incomplete() }, { "items", completion_item_array } }; | ||
| to_ret = json { { "isIncomplete", false }, { "items", completion_item_array } }; |
There was a problem hiding this comment.
should we perhaps impose some limits here and return isIncomplete: true?
There was a problem hiding this comment.
Have you noticed any performance issues or completion latency? Either way, I would add it to a Todo list for the future
parser_library/src/expressions/conditional_assembly/ca_expr_visitor.h
Outdated
Show resolved
Hide resolved
| @@ -15,6 +15,7 @@ | |||
| #ifndef HLASMPLUGIN_PARSERLIBRARY_LIBRARY_H | |||
| #define HLASMPLUGIN_PARSERLIBRARY_LIBRARY_H | |||
|
|
|||
| #include <regex> | |||
There was a problem hiding this comment.
perhaps split off library_local
| auto opencodes = find_related_opencodes(document_uri); | ||
| if (opencodes.empty()) | ||
| return { pos, document_uri }; | ||
| // for now take last opencode | ||
| return opencodes.back()->get_lsp_feature_provider().definition(document_uri, pos); |
There was a problem hiding this comment.
maybe we only need a function that returns that last element?
| for (const auto& i : actual) \ | ||
| { \ | ||
| auto it = std::find(expected.begin(), expected.end(), i); \ | ||
| EXPECT_NE(it, expected.end()) << "Test returned unexpected value:" << i; \ | ||
| } \ |
There was a problem hiding this comment.
this does not work for {1,1} and {1,2}
I suggest sorting if possible and then compare in order.
Also expected cannot be an expression, because it would be evaluated multiple times, so I suggest a helper functions.
Co-authored-by: slavek-kucera <53339291+slavek-kucera@users.noreply.github.com>
…-che4z-lsp-for-hlasm into lsp-context-refactor
…factor # Conflicts: # parser_library/include/c_view_array.h # parser_library/include/lib_config.h # parser_library/include/protocol.h # parser_library/src/analyzing_context.h # parser_library/src/debugging/debugger.cpp # parser_library/src/debugging/debugger.h
|
Kudos, SonarCloud Quality Gate passed! |
|
🎉 This PR is included in version 0.13.0-beta.1 🎉 The release is available on GitHub release Your semantic-release bot 📦🚀 |
|
🎉 This PR is included in version 0.13.0 🎉 The release is available on GitHub release Your semantic-release bot 📦🚀 |
Redesign the way information needed for go to definition, completion, etc are collected.
Whole statements are analyzed once they were parsed, instead of collecting the information in various places during parsing. Using visitor pattern to keep the analysis of operands in one place.
LSP context keeps all the occurences and definitions structured according to file and macro scope.