Add indexed include definition navigation

Myriad-Dreamin · Myriad-Dreamin · commit b589f587b6bc · 2026-06-18T19:59:12.000+08:00
diff --git a/src/index/merged_index.cpp b/src/index/merged_index.cpp
@@ -5,6 +5,7 @@
 
 #include "index/serialization.h"
 #include "support/filesystem.h"
+#include "syntax/lexer.h"
 
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/Support/raw_os_ostream.h"
@@ -443,6 +444,44 @@ void MergedIndex::lookup(this const Self& self,
     }
 }
 
+std::optional<std::uint32_t> MergedIndex::find_include_definition(this const Self& self,
+                                                                  std::uint32_t offset) {
+    if(self.impl) {
+        auto& index = *self.impl;
+        auto argument = find_directive_argument_at(index.content, offset);
+        if(!argument) {
+            return std::nullopt;
+        }
+        for(auto& [_, context]: index.compilation_contexts) {
+            for(auto& location: context.include_locations) {
+                if(location.include == static_cast<std::uint32_t>(-1) &&
+                   location.line == argument->line) {
+                    return location.path_id;
+                }
+            }
+        }
+    } else if(self.buffer) {
+        auto index = fbs::GetRoot<binary::MergedIndex>(self.buffer->getBufferStart());
+        llvm::StringRef content;
+        if(auto* stored_content = index->content()) {
+            content = stored_content->string_view();
+        }
+        auto argument = find_directive_argument_at(content, offset);
+        if(!argument) {
+            return std::nullopt;
+        }
+        for(auto context: *index->compilation_contexts()) {
+            for(auto location: *context->include_locations()) {
+                if(location->include_id() == static_cast<std::uint32_t>(-1) &&
+                   location->line() == argument->line) {
+                    return location->path_id();
+                }
+            }
+        }
+    }
+    return std::nullopt;
+}
+
 void MergedIndex::lookup(this const Self& self,
                          SymbolHash symbol,
                          RelationKind kind,
diff --git a/src/index/merged_index.h b/src/index/merged_index.h
@@ -3,6 +3,7 @@
 #include <chrono>
 #include <cstdint>
 #include <memory>
+#include <optional>
 #include <vector>
 
 #include "index/tu_index.h"
@@ -48,6 +49,10 @@ class MergedIndex {
                 std::uint32_t offset,
                 llvm::function_ref<bool(const Occurrence&)> callback);
 
+    /// Find the included path id for an include directive argument at `offset`.
+    std::optional<std::uint32_t> find_include_definition(this const Self& self,
+                                                         std::uint32_t offset);
+
     /// Lookup the relations of given symbol.
     void lookup(this const Self& self,
                 SymbolHash symbol,
diff --git a/src/server/compiler/compiler.cpp b/src/server/compiler/compiler.cpp
@@ -761,6 +761,8 @@ kota::task<> Compiler::run_compile(std::uint32_t pid, std::shared_ptr<Session::P
         OpenFileIndex ofi;
         ofi.file_index = std::move(tu_index.main_file_index);
         ofi.symbols = std::move(tu_index.symbols);
+        ofi.include_paths = std::move(tu_index.graph.paths);
+        ofi.include_locations = std::move(tu_index.graph.locations);
         ofi.content = sess->text;
         ofi.mapper.emplace(ofi.content, lsp::PositionEncoding::UTF16);
         sess->file_index = std::move(ofi);
diff --git a/src/server/compiler/indexer.cpp b/src/server/compiler/indexer.cpp
@@ -26,6 +26,23 @@ namespace clice {
 
 namespace lsp = kota::ipc::lsp;
 
+namespace {
+
+auto zero_range() -> protocol::Range {
+    auto pos = protocol::Position{.line = 0, .character = 0};
+    return protocol::Range{.start = pos, .end = pos};
+}
+
+auto file_location(llvm::StringRef path) -> std::optional<protocol::Location> {
+    auto uri = lsp::URI::from_file_path(std::string(path));
+    if(!uri) {
+        return std::nullopt;
+    }
+    return protocol::Location{.uri = uri->str(), .range = zero_range()};
+}
+
+}  // namespace
+
 void Indexer::merge(const void* tu_index_data, std::size_t size) {
     auto tu_index = index::TUIndex::from(tu_index_data);
     if(tu_index.graph.paths.empty()) {
@@ -243,10 +260,57 @@ Indexer::CursorHit Indexer::resolve_cursor(llvm::StringRef path,
     return {};
 }
 
+std::optional<protocol::Location>
+    Indexer::find_include_definition(llvm::StringRef path,
+                                     const protocol::Position& position,
+                                     Session* session) {
+    if(session && session->file_index && session->file_index->mapper) {
+        auto offset = session->file_index->mapper->to_offset(position);
+        if(!offset) {
+            return std::nullopt;
+        }
+        auto target = session->file_index->find_include_definition(*offset);
+        if(target && *target < session->file_index->include_paths.size()) {
+            return file_location(session->file_index->include_paths[*target]);
+        }
+    }
+
+    const std::string* doc_text = session ? &session->text : nullptr;
+    if(!doc_text) {
+        return std::nullopt;
+    }
+    lsp::PositionMapper doc_mapper(*doc_text, lsp::PositionEncoding::UTF16);
+    auto offset = doc_mapper.to_offset(position);
+    if(!offset) {
+        return std::nullopt;
+    }
+
+    auto proj_it = workspace.project_index.path_pool.find(path);
+    if(proj_it == workspace.project_index.path_pool.cache.end()) {
+        return std::nullopt;
+    }
+    auto shard_it = workspace.merged_indices.find(proj_it->second);
+    if(shard_it == workspace.merged_indices.end()) {
+        return std::nullopt;
+    }
+
+    auto target = shard_it->second.find_include_definition(*offset);
+    if(!target || *target >= workspace.project_index.path_pool.paths.size()) {
+        return std::nullopt;
+    }
+    return file_location(workspace.project_index.path_pool.path(*target));
+}
+
 std::vector<protocol::Location> Indexer::query_relations(llvm::StringRef path,
                                                          const protocol::Position& position,
                                                          RelationKind kind,
                                                          Session* session) {
+    if(kind.value() == RelationKind::Definition) {
+        if(auto include = find_include_definition(path, position, session)) {
+            return {*include};
+        }
+    }
+
     auto hit = resolve_cursor(path, position, session);
     if(hit.hash == 0)
         return {};
diff --git a/src/server/compiler/indexer.h b/src/server/compiler/indexer.h
@@ -224,6 +224,11 @@ class Indexer {
                              const protocol::Position& position,
                              Session* session);
 
+    /// Resolve an include directive argument at (position), if any.
+    std::optional<protocol::Location> find_include_definition(llvm::StringRef path,
+                                                              const protocol::Position& position,
+                                                              Session* session);
+
     /// Collect relations grouped by target symbol, across all index sources.
     void collect_grouped_relations(
         index::SymbolHash hash,
diff --git a/src/server/worker/stateful_worker.cpp b/src/server/worker/stateful_worker.cpp
@@ -241,7 +241,7 @@ void StatefulWorker::register_handlers() {
                         return result ? to_raw(*result) : kota::codec::RawValue{"null"};
                     });
                 case K::GoToDefinition:
-                    // TODO: Implement go-to-definition
+                    // Indexer handles go-to-definition; worker fallback is intentionally empty.
                     co_return kota::codec::RawValue{"[]"};
                 case K::SemanticTokens:
                     co_return co_await with_ast(params.path, [&](DocumentEntry& doc) {
diff --git a/src/server/workspace/workspace.cpp b/src/server/workspace/workspace.cpp
@@ -5,6 +5,7 @@
 
 #include "support/filesystem.h"
 #include "support/logging.h"
+#include "syntax/lexer.h"
 #include "syntax/scan.h"
 
 #include "kota/codec/json/json.h"
@@ -36,6 +37,24 @@ const static index::Occurrence* lookup_occurrence(const std::vector<index::Occur
     return best;
 }
 
+std::optional<std::uint32_t>
+    lookup_include_definition(llvm::StringRef content,
+                              llvm::ArrayRef<index::IncludeLocation> includes,
+                              std::uint32_t offset) {
+    auto argument = find_directive_argument_at(content, offset);
+    if(!argument) {
+        return std::nullopt;
+    }
+
+    for(auto& include: includes) {
+        if(include.include == static_cast<std::uint32_t>(-1) &&
+           include.line == argument->line) {
+            return include.path_id;
+        }
+    }
+    return std::nullopt;
+}
+
 std::optional<std::pair<index::SymbolHash, protocol::Range>>
     OpenFileIndex::find_occurrence(std::uint32_t offset) const {
     if(!mapper)
@@ -53,6 +72,10 @@ std::optional<std::pair<index::SymbolHash, protocol::Range>>
     };
 }
 
+std::optional<std::uint32_t> OpenFileIndex::find_include_definition(std::uint32_t offset) const {
+    return lookup_include_definition(content, include_locations, offset);
+}
+
 std::optional<std::pair<index::SymbolHash, protocol::Range>>
     MergedIndexShard::find_occurrence(std::uint32_t offset) const {
     auto* m = mapper();
@@ -73,6 +96,10 @@ std::optional<std::pair<index::SymbolHash, protocol::Range>>
     return result;
 }
 
+std::optional<std::uint32_t> MergedIndexShard::find_include_definition(std::uint32_t offset) const {
+    return index.find_include_definition(offset);
+}
+
 llvm::SmallVector<std::uint32_t> Workspace::on_file_saved(std::uint32_t path_id) {
     llvm::SmallVector<std::uint32_t> dirtied;
 
diff --git a/src/server/workspace/workspace.h b/src/server/workspace/workspace.h
@@ -6,6 +6,7 @@
 #include <string>
 #include <unordered_map>
 #include <utility>
+#include <vector>
 
 #include "command/command.h"
 #include "command/toolchain.h"
@@ -56,6 +57,8 @@ struct HeaderFileContext {
 struct OpenFileIndex {
     index::FileIndex file_index;
     index::SymbolTable symbols;
+    std::vector<std::string> include_paths;
+    std::vector<index::IncludeLocation> include_locations;
     std::string content;  ///< Buffer text at index time (for position mapping).
 
     /// Cached PositionMapper built from `content`.  Avoids re-scanning line
@@ -67,6 +70,9 @@ struct OpenFileIndex {
     std::optional<std::pair<index::SymbolHash, protocol::Range>>
         find_occurrence(std::uint32_t offset) const;
 
+    /// Find an include definition target path id at byte offset.
+    std::optional<std::uint32_t> find_include_definition(std::uint32_t offset) const;
+
     /// Iterate relations matching `kind`, calling back with pre-converted ranges.
     /// Callback: (const index::Relation&, protocol::Range) -> bool (true = continue).
     template <typename Fn>
@@ -115,6 +121,9 @@ struct MergedIndexShard {
     std::optional<std::pair<index::SymbolHash, protocol::Range>>
         find_occurrence(std::uint32_t offset) const;
 
+    /// Find an include definition target path id at byte offset.
+    std::optional<std::uint32_t> find_include_definition(std::uint32_t offset) const;
+
     /// Iterate relations matching `kind`, calling back with pre-converted ranges.
     /// Callback: (const index::Relation&, protocol::Range) -> bool (true = continue).
     template <typename Fn>
diff --git a/src/syntax/lexer.cpp b/src/syntax/lexer.cpp
@@ -162,4 +162,33 @@ std::optional<LocalSourceRange> find_directive_argument(llvm::StringRef content,
     return std::nullopt;
 }
 
+std::optional<DirectiveArgument> find_directive_argument_at(llvm::StringRef content,
+                                                            std::uint32_t offset,
+                                                            const clang::LangOptions* lang_opts) {
+    if(offset >= content.size()) {
+        return std::nullopt;
+    }
+
+    std::uint32_t line_start = 0;
+    if(offset > 0) {
+        auto pos = content.rfind('\n', offset - 1);
+        if(pos != llvm::StringRef::npos) {
+            line_start = static_cast<std::uint32_t>(pos + 1);
+        }
+    }
+
+    auto range = find_directive_argument(content, line_start, lang_opts);
+    if(!range || !range->contains(offset)) {
+        return std::nullopt;
+    }
+
+    std::uint32_t line = 1;
+    for(char c: content.take_front(line_start)) {
+        if(c == '\n') {
+            ++line;
+        }
+    }
+    return DirectiveArgument{*range, line};
+}
+
 }  // namespace clice
diff --git a/src/syntax/lexer.h b/src/syntax/lexer.h
@@ -82,4 +82,17 @@ std::optional<LocalSourceRange>
                             std::uint32_t offset,
                             const clang::LangOptions* lang_opts = nullptr);
 
+struct DirectiveArgument {
+    LocalSourceRange range;
+    std::uint32_t line = 0;
+};
+
+/// Find the directive argument containing `offset`.
+/// Returns its source range and 1-based line number, or nullopt if `offset`
+/// is not inside a directive argument.
+std::optional<DirectiveArgument>
+    find_directive_argument_at(llvm::StringRef content,
+                               std::uint32_t offset,
+                               const clang::LangOptions* lang_opts = nullptr);
+
 }  // namespace clice
diff --git a/tests/corpus/definition/include_definition.test b/tests/corpus/definition/include_definition.test
@@ -0,0 +1,11 @@
+#[test.h]
+#pragma once
+int from_test;
+
+#[macro.h]
+int from_macro;
+
+#[main.cpp]
+#define HEADER "macro.h"
+#$(outside)include "$(quoted)test.h"
+#include $(macro)HEADER
diff --git a/tests/snapshots/include_definition/snapshot/definition/include_definition.test.snap.yml b/tests/snapshots/include_definition/snapshot/definition/include_definition.test.snap.yml
@@ -0,0 +1,8 @@
+---
+source: include_definition_tests.cpp
+created_at: 2026-06-18
+input_file: definition/include_definition.test
+---
+- { point: "macro", target: "macro.h", range: "0:0-0:0" }
+- { point: "outside" }
+- { point: "quoted", target: "test.h", range: "0:0-0:0" }
diff --git a/tests/unit/index/include_definition_tests.cpp b/tests/unit/index/include_definition_tests.cpp
diff --git a/tests/unit/server/stateful_worker_tests.cpp b/tests/unit/server/stateful_worker_tests.cpp
