30 Commits

Author SHA1 Message Date
Myriad-Dreamin
1f8168577b wip 2026-04-20 23:10:08 +08:00
Myriad-Dreamin
60cab39f92 dev: make server public 2026-04-20 23:10:08 +08:00
Myriad-Dreamin
67bedd1ae0 dev: index nodes 2026-04-20 23:10:08 +08:00
Myriad-Dreamin
7ab8ad9513 fix: crash 2026-04-20 23:10:08 +08:00
Myriad-Dreamin
c11178fd77 dev: builtin lib 2026-04-20 23:10:08 +08:00
Myriad-Dreamin
150f47c590 hardcode clang 2026-04-20 23:10:08 +08:00
Myriad-Dreamin
b7987fded3 dev: workaround cc1 compilation 2026-01-26 17:45:00 +08:00
Myriad-Dreamin
44a4bd4107 dev: index notify 2026-01-26 17:44:47 +08:00
Myriad-Dreamin
044e4c4b27 feat: implement serialize for optional 2026-01-26 17:44:47 +08:00
Myriad-Dreamin
68eb63ba04 ActiveFile 2026-01-26 17:44:47 +08:00
Myriad-Dreamin
39d3648fdd feat: add method to check indices size 2026-01-26 17:43:16 +08:00
Myriad-Dreamin
6640c05d66 build: cannot build cpptrace 2026-01-26 17:43:16 +08:00
Myriad-Dreamin
69ac764ef2 feat: split copy_headers for external builds 2026-01-26 17:43:16 +08:00
Myriad-Dreamin
04c6ca5337 fix: correct server ref impl 2026-01-26 17:43:16 +08:00
Myriad-Dreamin
2408978d2d fix: correct plugin_paths decl 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
74f75f107f fix: compile error 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
0e2e487bc9 feat: introduce any type alias for llvm::json::Value 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
6be48bccd2 feat: implement dummy workspace/executeCommand 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
4262734d21 dev: simplify protocol def 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
ed0d7db3db dev: simplify protocol def 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
824b305f93 dev: simplify protocol def 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
f89793c66d dev: more lifetime hooks 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
4425fb5244 dev: simplify 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
25a85a3b8e feat: implement apis 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
0f95344abe fix: typo 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
3511915886 dev: remove useless comments 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
6e50451c43 fix: typo 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
99d9363b95 fix: paths 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
7533d4d15e docs: compile and laod plugin sections 2026-01-26 01:01:39 +08:00
Myriad-Dreamin
a118c16e96 feat: add minimal support to clice server plugins 2026-01-26 01:01:39 +08:00
496 changed files with 21226 additions and 34001 deletions

View File

@@ -139,15 +139,3 @@ KeepEmptyLines:
AtEndOfFile: false
AtStartOfBlock: false
AtStartOfFile: false
StatementMacros:
- DECO_CFG_START
- DECO_CFG
- DECO_CFG_END
- DecoKV
- DecoFlag
- DecoComma
- DecoInput
- DecoPack
- DecoKVStyled
- DecoMulti

View File

@@ -1,50 +0,0 @@
---
Checks: >
-*,
bugprone-*,
modernize-*,
performance-*,
readability-*,
-modernize-use-trailing-return-type,
-readability-magic-numbers,
-readability-else-after-return,
-readability-braces-around-statements,
-readability-avoid-const-params-in-decls,
-readability-named-parameter,
-readability-implicit-bool-conversion,
-readability-use-anyofallof,
-bugprone-easily-swappable-parameters,
-bugprone-exception-escape,
-bugprone-narrowing-conversions,
-modernize-use-nodiscard,
WarningsAsErrors: ""
HeaderFilterRegex: "(src|tests)/.*"
CheckOptions:
# Naming conventions matching project style
- key: readability-identifier-naming.ClassCase
value: CamelCase
- key: readability-identifier-naming.StructCase
value: CamelCase
- key: readability-identifier-naming.EnumCase
value: CamelCase
- key: readability-identifier-naming.EnumConstantCase
value: CamelCase
- key: readability-identifier-naming.TemplateParameterCase
value: CamelCase
- key: readability-identifier-naming.TypeAliasCase
value: CamelCase
- key: readability-identifier-naming.FunctionCase
value: lower_case
- key: readability-identifier-naming.MethodCase
value: lower_case
- key: readability-identifier-naming.VariableCase
value: lower_case
- key: readability-identifier-naming.ParameterCase
value: lower_case
- key: readability-identifier-naming.MemberCase
value: lower_case
- key: readability-identifier-naming.NamespaceCase
value: lower_case

View File

@@ -1,268 +0,0 @@
# clice — Project Guide
## Project Overview
clice is a next-generation C++ language server (LSP) built on LLVM/Clang, targeting modern C++ (C++20/23). It uses a multi-process architecture with a master server coordinating stateless and stateful workers.
## Core Correction Patterns — Lessons from Past Interactions
The following patterns were extracted from extensive real-world collaboration. These are recurring mistakes that MUST be avoided. Read them carefully — they represent hard-won lessons, not hypothetical concerns.
### Pattern 1: Misjudging Real-World Priorities
AI tends to optimize whatever metric looks most impressive, rather than what actually matters in the user's real scenario.
**Example**: During performance optimization, the AI proudly reported "hot cache is 4-7x faster!" — but the function in question runs at LSP server startup, which is ALWAYS a cold start. Optimizing hot cache was completely meaningless.
**Rule**: Before optimizing or analyzing anything, first understand the REAL usage scenario. Ask yourself: "When does this code actually run? What does the user actually experience?" Do not chase metrics that look good on paper but are irrelevant in practice.
### Pattern 2: Pushing Without Local Verification
The most common and most damaging pattern. AI proposes a fix, pushes it immediately, CI fails, then another fix, push, fail again — wasting CI cycles and the user's time.
**Rule**: NEVER push code that you haven't verified locally. Before every push:
- Build locally with the same configuration CI uses.
- Run the relevant tests locally and confirm they pass.
- If you cannot reproduce the CI environment locally, say so — do not just "try and see."
- "It compiles" is NOT sufficient. Tests must pass.
### Pattern 3: Superficial Refactoring
When asked to refactor, AI tends to do mechanical code movement (copy functions from A to B) without understanding the deeper design intent (ownership, responsibility boundaries, API cleanliness).
**Example**: When splitting `MasterServer` into `Workspace` and `Session`, the AI moved functions but kept ugly APIs like `f(path_id, sessions_map)` instead of the clean `f(Session&)` that the refactoring was meant to achieve.
**Rule**: When refactoring, understand the WHY. Ask: "What design problem is this refactoring solving?" If you're just moving code around without improving the abstractions, you're not refactoring — you're rearranging deck chairs.
### Pattern 4: Fixing Only the Immediate Instance, Not the Pattern
When given a cleanup instruction, AI applies it to the single file or function currently being discussed, ignoring all other occurrences in the project.
**Example**: User says "remove decorative `===` comment separators." AI removes them from `workspace.cpp` only. User has to say: "The other files too!"
**Rule**: When given a cleanup or style instruction, apply it project-wide. Use `Grep` to find ALL occurrences and fix them all in one pass. Think: "Where else does this pattern appear?"
### Pattern 5: Never Skip, Disable, or Work Around Failing Tests
When stuck on a difficult bug (especially flaky CI, race conditions, platform-specific issues), AI may propose marking tests as `continue-on-error`, skipping them, or adding `expected-failure` annotations to make CI green.
**Rule**: This is ABSOLUTELY FORBIDDEN. If a test fails, fix the root cause. There are ZERO exceptions. Skipping a test to make CI green is not "fixing" — it is hiding a bug. If you ever find yourself thinking "maybe we should just skip this test," stop and reconsider your approach entirely.
### Pattern 6: Excessive Confirmation Seeking vs. Premature Execution
AI oscillates between two extremes: asking "should I do X?" for every trivial decision, or silently executing major changes without confirmation.
**Rule**: Calibrate based on reversibility and impact:
- **Small, reversible changes** (formatting, renaming a local variable, adding a test): just do it.
- **Architecture decisions, API changes, large refactors**: propose the plan first, wait for confirmation.
- **Pushing to remote, creating PRs, modifying CI**: always confirm.
- When the user says "go ahead" or "do it," execute fully without asking again mid-way.
## Code Reuse & Understanding Before Implementation
**This is the single most important rule in this project.** Before writing ANY new code, you MUST thoroughly read and understand the existing codebase first. This project has a rich set of utilities, abstractions, and patterns already in place — duplicating them wastes effort and creates maintenance burden.
Concrete requirements:
1. **Read before you write.** Before implementing a feature or fix, explore the relevant modules in `src/`. Search for existing helpers, utilities, and patterns that solve the same or similar problems. Use `Grep`, `Glob`, and `Agent` tools to investigate thoroughly — do not assume something doesn't exist just because you haven't seen it yet.
2. **Reuse existing infrastructure.** This project already has:
- A `Lexer` class (`src/syntax/lexer.h`) — do not hand-write token scanning logic.
- A `PositionMapper` for source location conversion — do not reimplement offset-to-line/column math.
- `CompilationUnitRef` methods (`decompose_location`, `decompose_range`, `file_path`, `directives`, etc.) — use them instead of raw Clang APIs.
- `SemanticVisitor` for AST traversal — extend it, do not write custom recursive AST walkers.
- `Tester` framework for unit tests with VFS, annotation support, and multi-phase compilation — use it, do not create ad-hoc test setups.
- Utility functions in `src/support/` — check there before writing new helpers.
3. **Follow established patterns.** When adding a new feature (e.g., a new LSP request handler), look at how 2-3 existing features of the same kind are implemented. Match their structure: same file organization, same function signatures, same error handling patterns. If every other feature in `src/feature/` follows a certain pattern, yours should too.
4. **Do not reinvent what the project already has.** If you find yourself writing a helper function that feels generic (string manipulation, path handling, JSON serialization, source range conversion), STOP and search the codebase first. There is a high probability it already exists. Creating duplicates leads to inconsistencies and bugs when one copy gets updated but the other doesn't.
5. **When in doubt, ask.** If you're unsure whether an existing utility covers your use case or whether to extend an existing abstraction vs. create a new one, ask the user rather than guessing.
## Source Layout
- `src/server/` — LSP server core: master server, compiler, indexer, stateful/stateless workers
- `src/feature/` — LSP feature implementations: hover, completion, document links, semantic tokens, etc.
- `src/compile/` — Compilation orchestration: compilation unit, directives, diagnostics
- `src/index/` — Symbol indexing: TUIndex, ProjectIndex, MergedIndex, include graph
- `src/semantic/` — Semantic analysis: symbol kinds, relations, AST visitor, template resolver
- `src/syntax/` — Lexer, scanner, token types, dependency graph
- `src/command/` — CLI parsing, compilation database, toolchain detection
- `src/support/` — Utilities: logging, filesystem, JSON, string helpers
## Build System
- Uses **pixi** for environment management and **CMake + Ninja** for building.
- Two build types: `Debug` and `RelWithDebInfo` (default).
- Build output goes to `build/[type]/`.
- See `/build`, `/test`, `/format` commands for common operations.
## Commit Message Format
Use **conventional commits** — enforced by CI:
```
<type>(<scope>): <short description>
```
- **Types**: `feat`, `fix`, `refactor`, `chore`, `docs`, `ci`, `test`
- **Scopes**: match `src/` subdirectories or feature names, e.g. `completion`, `server`, `index`, `tests`, `document links`
- Keep the subject line under 70 characters.
## Tests
Three types of tests, all must pass before committing:
- **Unit tests** (`tests/unit/`): C++ tests using the project's own test framework. Test names should be at most 4 words.
- **Integration tests** (`tests/integration/`): Python pytest tests that start a real clice server and communicate via LSP.
- **Smoke tests** (`tests/smoke/`): Replay recorded LSP sessions via `tests/replay.py`.
### Integration Test Style
- Keep tests concise. Do NOT write large comment blocks explaining the test layout or expected behavior.
- Use descriptive test function names and short inline comments only where logic is non-obvious.
## Pre-PR Review
Before opening a PR, launch **3 parallel subagents** to review the diff independently:
1. **Correctness reviewer**: Check for logic errors, edge cases, undefined behavior, and off-by-one mistakes.
2. **Style reviewer**: Verify the code follows this project's naming conventions, coding style, and CLAUDE.md rules.
3. **Test reviewer**: Confirm test coverage is adequate — new functionality has tests, edge cases are covered, and no existing tests were broken or weakened.
Each agent should read the full diff (`git diff main...HEAD`) and report issues. Fix all reported issues before opening the PR.
## Pre-commit Checklist
Before committing code, you MUST:
1. **Run `pixi run format`** to format all source files.
2. **Pass all three types of tests:**
- Unit tests: `pixi run unit-test [type]`
- Integration tests: `pixi run integration-test [type]`
- Smoke tests: `pixi run smoke-test [type]`
3. **All test failures must be fixed before committing.** This is a HARD REQUIREMENT with NO exceptions:
- If a test fails, it MUST be fixed before you commit. Do NOT commit with known failures.
- Do NOT skip, disable, or mark tests as expected-failure to work around breakage.
- Do NOT argue "this test was already broken before my changes" — if it fails on your branch, it is YOUR responsibility to fix it before committing. The main branch CI is green; any failure on your branch is caused by your changes, period.
- Do NOT defer fixing to a follow-up PR. Fix it NOW, in this branch, before committing.
---
## C++ Coding Style
### Template & Type Traits
- Do NOT blindly add `std::remove_cvref_t` on every template parameter. Understand C++ template argument deduction rules:
- `template<typename T> void f(T x)``T` is always deduced as a non-reference, non-cv-qualified type. No need for `remove_cvref_t`.
- `template<typename T> void f(T& x)``T` is deduced as the referred-to type (possibly cv-qualified, but never a reference). No need for `remove_cvref_t` to strip references.
- `template<typename T> void f(const T& x)``T` is deduced as a non-const, non-reference type. No need for `remove_cvref_t`.
- `template<typename T> void f(T&& x)`**forwarding reference**: `T` CAN be deduced as an lvalue reference (e.g., `int&`). This is the ONLY case where `std::remove_cvref_t<T>` is needed to get the bare type.
- Class template parameters and return types are also never deduced as references; don't add `remove_cvref_t` on them either.
### Type Traits & Concepts (C++20/23)
- This project targets C++20/23. Use variable templates directly for type traits — do NOT use the old pattern of wrapping a class template static member in a variable template. Prefer:
```cpp
// Good: directly specialize a variable template
template<typename T>
inline constexpr bool is_my_type_v = false;
template<>
inline constexpr bool is_my_type_v<MyType> = true;
```
```cpp
// Bad: unnecessary class template wrapper
template<typename T>
struct is_my_type : std::false_type {};
template<>
struct is_my_type<MyType> : std::true_type {};
template<typename T>
inline constexpr bool is_my_type_v = is_my_type<T>::value;
```
- When defining a concept that checks a type trait, do NOT add `std::remove_cvref_t` unless you specifically intend the concept to see through references/cv-qualifiers. If the concept is meant for a bare type, just use `T` directly — the caller is responsible for passing the right type.
```cpp
// Good
template<typename T>
concept MyTrait = is_my_type_v<T>;
// Bad: unnecessary remove_cvref_t
template<typename T>
concept MyTrait = is_my_type_v<std::remove_cvref_t<T>>;
```
### Naming Conventions
- **Variables, member fields, function names**: `snake_case`. Class member fields do NOT use any special suffix/prefix (no trailing `_`, no `m_` prefix).
- **Class names, template parameter names, enum names**: `PascalCase`. Exception: some class names also use `snake_case` — follow the existing style in the project.
- **Enum values**: `PascalCase`.
### String Literals
- Prefer C++11 raw string literals `R"(...)"` over escaped strings. Avoid `\"`, `\\`, `\n` in string literals when a raw literal is cleaner.
### Error Handling
- **Prefer `if` with init-statements to tightly scope error variables**, but avoid them when they compromise code readability or flatten control flow.
- **Omit redundant conditions:** If the error type provides an `operator bool` or evaluates implicitly (e.g., standard error codes, custom error wrappers), omit the redundant condition check.
- **Avoid forced `else` branches:** If scoping the variable inside the `if` requires you to introduce an `else` block for the success path (especially when returning early on error), declare the variable in the local scope instead to keep the control flow flat.
```cpp
// Good: Omit redundant condition when the type has operator bool
if (auto err = foo()) {
/* handle error */
}
// Bad: Redundant condition check
if (auto err = foo(); err) {
/* handle error */
}
// Good: Use init-statement when a custom condition is required,
// AND the variable isn't needed outside the if-statement
if (auto result = foo(); !result.has_value()) {
/* handle error */
}
// --- Scope and Control Flow Considerations ---
// Bad: Using init-statement forces an 'else' block because 'result'
// goes out of scope, leading to nested/redundant code.
if (auto result = get_data(); !result.has_value()) {
return result.error();
} else {
process(result.value()); // Success path is forced into a nested block
}
// Good: Declare as a regular local variable to allow early exit
// and keep the success path un-nested (flat control flow).
auto result = get_data();
if (!result.has_value()) {
return result.error();
}
process(result.value());
```
### Style
- Prefer `[[maybe_unused]]` over `(void)` for intentionally unused variables or parameters.
### Modern C++ Usage
- Use C++20/23 APIs whenever possible. Do NOT use `<iostream>` facilities (`std::cout`, `std::cin`, `std::cerr`, etc.). Also do NOT use C-style I/O (`printf`, `fprintf`, etc.).
- Prefer `std::ranges` / `std::views` APIs over raw loops and traditional `<algorithm>` calls.
- If the project depends on LLVM, prefer LLVM's efficient data structures (e.g., `llvm::SmallVector`, `llvm::DenseMap`, `llvm::StringMap`, `llvm::StringRef`) over their `std` counterparts when appropriate.
### Parameter Passing Preferences
- For string parameters, prefer `llvm::StringRef` > `std::string_view` > `const std::string&`.
- For array/span parameters, prefer `llvm::ArrayRef` > `std::span` > `const std::vector&`.

View File

@@ -1,15 +0,0 @@
Build the project. Accepts an optional argument for build type: `Debug` or `RelWithDebInfo` (default).
Available build commands:
- CMake configure only: `pixi run cmake-config [type]`
- CMake build only (skip configure): `pixi run cmake-build [type]`
- Full build (configure + build): `pixi run build [type]`
- Build a specific target: `pixi run cmake-build [type]` then `cmake --build build/[type] --target [target]`
Common targets: `clice`, `unit_tests`
Example usage:
- `/build` — full build RelWithDebInfo
- `/build Debug` — full build Debug

View File

@@ -1,5 +0,0 @@
Format all project source files.
Run: `pixi run format`
Formats C++, Python, Lua, JS/TS, Markdown, JSON, TOML, and YAML files.

View File

@@ -1,19 +0,0 @@
Run tests. Accepts an optional argument for build type: `Debug` or `RelWithDebInfo` (default).
Available test commands:
- Unit tests: `pixi run unit-test [type]`
- Integration tests: `pixi run integration-test [type]`
- Smoke tests: `pixi run smoke-test [type]`
- All tests (unit + integration): `pixi run test [type]`
Filtering specific tests:
- Unit tests: `pixi run unit-test [type] --test-filter=SuiteName.CaseName`
- Integration tests: `pixi run pytest tests/integration -k "test_name" --executable=./build/[type]/bin/clice`
- Smoke tests: `pixi run python tests/replay.py tests/smoke/specific.jsonl --clice=./build/[type]/bin/clice`
Example usage:
- `/test` — run all tests (RelWithDebInfo)
- `/test Debug` — run all tests (Debug)

View File

@@ -1,7 +1,2 @@
chat:
auto_reply: false
reviews:
auto_review:
enabled: true
summary:
enabled: false

7
.gitattributes vendored
View File

@@ -1,9 +1,2 @@
# SCM syntax highlighting & preventing 3-way merges
pixi.lock merge=binary linguist-language=YAML linguist-generated=true -diff
# Force LF line endings for test data so that byte offsets from clang
# (which reads from disk) match the content sent by didOpen in tests.
tests/data/** text eol=lf
# Treat trace files as binary to suppress text diffs
tests/smoke/*.jsonl linguist-generated=true binary

View File

@@ -1,44 +0,0 @@
name: benchmark
on:
workflow_dispatch:
jobs:
benchmark:
strategy:
fail-fast: false
matrix:
os: [ubuntu-24.04, macos-15, windows-2025]
runs-on: ${{ matrix.os }}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
uses: actions/checkout@v4
- uses: ./.github/actions/setup-pixi
- name: Build scan_benchmark
run: |
pixi run cmake-config RelWithDebInfo ON "-DCLICE_ENABLE_BENCHMARK=ON"
cmake --build build/RelWithDebInfo --target scan_benchmark
- name: Clone LLVM
run: git clone --depth 1 https://github.com/llvm/llvm-project.git
- name: Generate CDB
run: |
cmake -B llvm-build -G Ninja \
-DCMAKE_EXPORT_COMPILE_COMMANDS=ON \
-DCMAKE_TOOLCHAIN_FILE="$(pwd)/cmake/toolchain.cmake" \
-DLLVM_ENABLE_PROJECTS="clang;clang-tools-extra;lld;lldb;mlir;polly;flang;bolt" \
-DLLVM_ENABLE_RUNTIMES="compiler-rt;libcxx;libcxxabi;libunwind" \
llvm-project/llvm
- name: Run benchmark
run: ./build/RelWithDebInfo/bin/scan_benchmark --runs 20 llvm-build/compile_commands.json
- name: Stop sccache server
if: runner.os == 'Windows'
run: pixi run -- sccache --stop-server || true

View File

@@ -24,11 +24,10 @@ jobs:
args: --lint ./docs
continue-on-error: true
- name: Check diff
run: |
if ! git diff --quiet; then
echo "::error::Formatting changes detected. Please run 'pixi run format' and commit the result."
git --no-pager diff --stat
git --no-pager diff
exit 1
fi
- name: Suggest changes
if: github.event_name == 'pull_request'
uses: reviewdog/action-suggester@v1.24.0
with:
tool_name: "fmt"
fail_level: any
filter_mode: nofilter

View File

@@ -20,11 +20,9 @@ jobs:
run: pixi run build-docs
- name: Deploy to GitHub Pages
uses: peaceiris/actions-gh-pages@v4
if: github.ref == 'refs/heads/main'
uses: peaceiris/actions-gh-pages@v4
with:
personal_token: ${{ secrets.PUBLISH_DOCS }}
external_repository: clice-io/docs
publish_dir: ./docs/.vitepress/dist
destination_dir: clice
keep_files: true
github_token: ${{ secrets.GITHUB_TOKEN }}
publish_dir: docs/.vitepress/dist
cname: clice.io

View File

@@ -19,7 +19,7 @@ jobs:
clice: ${{ steps.filter.outputs.clice }}
vscode: ${{ steps.filter.outputs.vscode }}
cmake: ${{ steps.filter.outputs.cmake }}
xmake: ${{ steps.filter.outputs.xmake }}
steps:
- uses: actions/checkout@v4
- uses: dorny/paths-filter@v3
@@ -46,34 +46,21 @@ jobs:
- 'tests/**'
- 'config/**'
- '.github/workflows/test-cmake.yml'
conventional-commit:
if: ${{ !startsWith(github.ref, 'refs/tags/') }}
runs-on: ubuntu-latest
steps:
- name: Check conventional commit format
env:
IS_PR: ${{ github.event_name == 'pull_request' }}
PR_TITLE: ${{ github.event.pull_request.title }}
COMMIT_MSG: ${{ github.event.head_commit.message }}
run: |
pattern='^(feat|fix|refactor|chore|build|ci|docs|test|perf|style|revert)(\(.+\))?: .+'
if [[ "$IS_PR" == "true" ]]; then
subject="$PR_TITLE"
label="PR title"
else
subject=$(echo "$COMMIT_MSG" | head -n1)
label="Commit message"
fi
if [[ ! "$subject" =~ $pattern ]]; then
echo "::error::$label must follow conventional commit format: type(scope)?: description"
echo " Valid types: feat, fix, refactor, chore, build, ci, docs, test, perf, style, revert"
echo " Got: '$subject'"
exit 1
fi
xmake:
- 'xmake.lua'
- 'src/**'
- 'include/**'
- 'tests/**'
- 'config/**'
- '.github/workflows/test-xmake.yml'
format:
needs: changes
permissions:
contents: read
checks: write
issues: write
pull-requests: write
if: ${{ needs.changes.outputs.format == 'true' }}
uses: ./.github/workflows/check-format.yml
@@ -83,12 +70,11 @@ jobs:
permissions:
contents: write
uses: ./.github/workflows/deploy-docs.yml
secrets: inherit
# clice:
# needs: changes
# if: ${{ needs.changes.outputs.clice == 'true' }}
# uses: ./.github/workflows/publish-clice.yml
clice:
needs: changes
if: ${{ needs.changes.outputs.clice == 'true' }}
uses: ./.github/workflows/publish-clice.yml
vscode:
needs: changes
@@ -100,6 +86,11 @@ jobs:
if: ${{ needs.changes.outputs.cmake == 'true' }}
uses: ./.github/workflows/test-cmake.yml
xmake:
needs: changes
if: ${{ needs.changes.outputs.xmake == 'true' }}
uses: ./.github/workflows/test-xmake.yml
release-clice:
permissions:
contents: write
@@ -117,17 +108,16 @@ jobs:
checks-passed:
if: ${{ always() && !startsWith(github.ref, 'refs/tags/') }}
needs:
- conventional-commit
- format
- deploy
# - clice
- clice
- vscode
- cmake
- xmake
runs-on: ubuntu-latest
steps:
- name: Check results
uses: re-actors/alls-green@release/v1
with:
allowed-skips: conventional-commit,format,deploy,clice,vscode,cmake
allowed-skips: format,deploy,clice,vscode,cmake,xmake
jobs: ${{ toJSON(needs) }}

View File

@@ -29,18 +29,32 @@ jobs:
runs-on: ${{ matrix.os }}
defaults:
run:
shell: bash
steps:
- name: Checkout repository
uses: actions/checkout@v4
- name: Setup xmake
uses: xmake-io/github-action-setup-xmake@v1
with:
xmake-version: 3.0.5
actions-cache-folder: ".xmake-cache"
actions-cache-key: ${{ matrix.os }}
package-cache: true
package-cache-key: ${{ matrix.os }}-pkg-release-v1
build-cache: true
build-cache-key: ${{ matrix.os }}-build-release-v1
- uses: ./.github/actions/setup-pixi
with:
environments: package
- name: Remove ci llvm toolchain on Windows
if: runner.os == 'Windows'
run: |
# @see https://github.com/xmake-io/xmake/issues/7158
xmake lua os.rmdir "C:/Program Files/Microsoft Visual Studio/2022/Enterprise/VC/Tools/Llvm"
xmake lua os.rmdir "C:/Program Files/LLVM"
- name: Package
run: pixi run package
@@ -49,7 +63,7 @@ jobs:
uses: svenstaro/upload-release-action@v2
with:
repo_token: ${{ secrets.GITHUB_TOKEN }}
file: build/RelWithDebInfo/${{ matrix.artifact_name }}
file: build/xpack/clice/${{ matrix.artifact_name }}
asset_name: ${{ matrix.asset_name }}
tag: ${{ github.ref }}
overwrite: true
@@ -59,7 +73,7 @@ jobs:
uses: svenstaro/upload-release-action@v2
with:
repo_token: ${{ secrets.GITHUB_TOKEN }}
file: build/RelWithDebInfo/${{ matrix.symbol_artifact_name }}
file: build/xpack/clice/${{ matrix.symbol_artifact_name }}
asset_name: ${{ matrix.symbol_asset_name }}
tag: ${{ github.ref }}
overwrite: true

View File

@@ -3,22 +3,18 @@ name: cmake
on:
workflow_call:
env:
CCACHE_DIR: ${{ github.workspace }}/.cache/ccache
SCCACHE_DIR: ${{ github.workspace }}/.cache/sccache
CCACHE_BASEDIR: ${{ github.workspace }}
SCCACHE_BASEDIRS: ${{ github.workspace }}
CCACHE_COMPILERCHECK: content
CCACHE_MAXSIZE: 2G
SCCACHE_CACHE_SIZE: 2G
jobs:
build:
strategy:
fail-fast: false
matrix:
os: [windows-2025, ubuntu-24.04, macos-15]
build_type: [Debug, RelWithDebInfo]
include:
- os: windows-2025
build_type: RelWithDebInfo
- os: ubuntu-24.04
build_type: Debug
- os: macos-15
build_type: Debug
runs-on: ${{ matrix.os }}
steps:
- name: Checkout repository
@@ -26,44 +22,8 @@ jobs:
- uses: ./.github/actions/setup-pixi
- name: Restore compiler cache
uses: actions/cache@v4
with:
path: ${{ runner.os == 'Windows' && '.cache/sccache' || '.cache/ccache' }}
key: ${{ runner.os }}-${{ matrix.build_type }}-ccache-${{ github.sha }}
restore-keys: |
${{ runner.os }}-${{ matrix.build_type }}-ccache-
- name: Zero cache stats
run: |
if [ "$RUNNER_OS" = "Windows" ]; then
pixi run -- sccache --stop-server || true
pixi run -- sccache --zero-stats || true
else
pixi run -- ccache --zero-stats || true
fi
shell: bash
- name: Build
run: pixi run build ${{ matrix.build_type }} ON
- name: Unit Test
run: pixi run unit-test ${{ matrix.build_type }}
- name: Integration Test
run: pixi run integration-test ${{ matrix.build_type }}
- name: Smoke Test
if: success() || failure()
run: pixi run smoke-test ${{ matrix.build_type }}
- name: Print cache stats and stop server
if: always()
run: |
if [ "$RUNNER_OS" = "Windows" ]; then
pixi run -- sccache --show-stats
pixi run -- sccache --stop-server || true
else
pixi run -- ccache --show-stats
fi
shell: bash
- name: Test
run: pixi run test ${{ matrix.build_type }}

42
.github/workflows/test-xmake.yml vendored Normal file
View File

@@ -0,0 +1,42 @@
name: xmake
on:
workflow_call:
jobs:
build:
strategy:
fail-fast: false
matrix:
os: [windows-2025, ubuntu-24.04, macos-15]
build_type: [debug, releasedbg]
exclude:
- os: windows-2025
build_type: debug
runs-on: ${{ matrix.os }}
steps:
- name: Checkout repository
uses: actions/checkout@v4
- name: Setup xmake
uses: xmake-io/github-action-setup-xmake@v1
with:
xmake-version: 3.0.5
actions-cache-folder: ".xmake-cache"
actions-cache-key: ${{ matrix.os }}
package-cache: true
package-cache-key: ${{ matrix.os }}-pixi
build-cache: true
build-cache-key: ${{ matrix.os }}-${{ matrix.build_type }}
- uses: ./.github/actions/setup-pixi
- name: Build
run: pixi run xmake ${{ matrix.build_type }}
- name: Test
run: pixi run xmake-test
- name: Remove llvm package (Linux)
if: runner.os == 'Linux'
run: xmake require --uninstall clice-llvm

13
.gitignore vendored
View File

@@ -35,7 +35,7 @@
*build*/
temp/
.cache/
.xmake/
.llvm*/
.clice/
compile_commands.json
@@ -56,20 +56,13 @@ __pycache__/
tests/unit/Local/
# IDEs & Editors
/.vscode/*
!/.vscode/launch.json
!/.vscode/tasks.json
/.vscode/
.vs/
.idea/
.claude
.clangd
# pixi environments
.env
.pixi/*
!.pixi/config.toml
.codex/
.claude/*
!.claude/CLAUDE.md
!.claude/commands/
openspec/

83
.vscode/launch.json vendored
View File

@@ -1,83 +0,0 @@
{
"version": "0.2.0",
"configurations": [
{
"type": "lldb",
"request": "launch",
"name": "Debug clice",
"program": "${workspaceFolder}/build/Debug/bin/clice",
"args": ["--mode=socket", "--port=50051"],
"cwd": "${workspaceFolder}"
},
{
"type": "lldb",
"request": "launch",
"name": "Debug clice (socket, RelWithDebInfo)",
"program": "${workspaceFolder}/build/RelWithDebInfo/bin/clice",
"args": ["--mode=socket", "--port=50051"],
"cwd": "${workspaceFolder}"
},
{
"name": "VSCode Extension (pipe)",
"type": "extensionHost",
"request": "launch",
"args": [
"--disable-extension=llvm-vs-code-extensions.vscode-clangd",
"--disable-extension=ms-vscode.cpptools",
"--disable-extension=ms-vscode.cpptools-extension-pack",
"--extensionDevelopmentPath=${workspaceFolder}/editors/vscode"
],
"env": {
"CLICE_MODE": "pipe"
},
"outFiles": ["${workspaceFolder}/editors/vscode/dist/**/*.js"],
"preLaunchTask": "npm: watch vscode ext"
},
{
"name": "VSCode Extension (socket)",
"type": "extensionHost",
"request": "launch",
"args": [
"--disable-extension=llvm-vs-code-extensions.vscode-clangd",
"--disable-extension=ms-vscode.cpptools",
"--disable-extension=ms-vscode.cpptools-extension-pack",
"--extensionDevelopmentPath=${workspaceFolder}/editors/vscode"
],
"env": {
"CLICE_MODE": "socket"
},
"outFiles": ["${workspaceFolder}/editors/vscode/dist/**/*.js"],
"preLaunchTask": "npm: watch vscode ext"
},
{
"type": "lldb",
"request": "launch",
"name": "Unit Test",
"program": "${workspaceFolder}/build/Debug/bin/unit_tests",
"args": ["--test-dir=./tests/data", "--test-filter=${input:filter}"],
"cwd": "${workspaceFolder}"
},
{
"type": "lldb",
"request": "launch",
"name": "Release Unit Test",
"program": "${workspaceFolder}/build/RelWithDebInfo/bin/unit_tests",
"args": ["--test-dir=./tests/data", "--test-filter=${input:filter}"],
"cwd": "${workspaceFolder}"
}
],
"compounds": [
{
"name": "clice + VSCode Extension (socket)",
"configurations": ["Debug clice", "VSCode Extension (socket)"],
"stopAll": true
}
],
"inputs": [
{
"id": "filter",
"type": "promptString",
"description": "Unit Test Filter"
}
]
}

42
.vscode/tasks.json vendored
View File

@@ -1,42 +0,0 @@
{
"version": "2.0.0",
"tasks": [
{
"label": "npm: install vscode deps",
"type": "shell",
"command": "pnpm",
"args": ["install"],
"options": {
"cwd": "${workspaceFolder}/editors/vscode"
},
"problemMatcher": []
},
{
"label": "npm: watch vscode ext",
"type": "shell",
"command": "pnpm",
"args": ["run", "watch"],
"options": {
"cwd": "${workspaceFolder}/editors/vscode"
},
"dependsOn": "npm: install vscode deps",
"problemMatcher": "$ts-webpack-watch",
"isBackground": true,
"presentation": {
"reveal": "never",
"group": "watchers"
}
},
{
"label": "npm: package vscode ext",
"type": "shell",
"command": "pnpm",
"args": ["run", "package"],
"options": {
"cwd": "${workspaceFolder}/editors/vscode"
},
"dependsOn": "npm: install vscode deps",
"problemMatcher": []
}
]
}

View File

@@ -13,45 +13,14 @@ set(CMAKE_LIBRARY_OUTPUT_DIRECTORY "${PROJECT_BINARY_DIR}/lib")
set(CMAKE_RUNTIME_OUTPUT_DIRECTORY "${PROJECT_BINARY_DIR}/bin")
option(CLICE_ENABLE_LTO "Enable ThinLTO for all targets" OFF)
option(CLICE_USE_LIBCXX "Use libc++ instead of libstdc++" OFF)
option(CLICE_OFFLINE_BUILD "Disable network downloads during configuration" OFF)
option(CLICE_ENABLE_TEST "Build unit tests" OFF)
option(CLICE_CI_ENVIRONMENT "Enable CI-specific configuration" OFF)
option(CLICE_ENABLE_BENCHMARK "Build benchmarks" OFF)
option(CLICE_RELEASE "Enable release packaging (LTO + strip + pack)" OFF)
# Global flags that apply to all targets (including FetchContent dependencies).
if(NOT MSVC)
add_compile_options(-ffunction-sections -fdata-sections)
endif()
if(APPLE)
# https://conda-forge.org/docs/maintainer/knowledge_base/#newer-c-features-with-old-sdk
add_compile_definitions(_LIBCPP_DISABLE_AVAILABILITY=1)
endif()
if(MSVC OR (CMAKE_CXX_COMPILER_ID MATCHES "Clang" AND
CMAKE_CXX_COMPILER_FRONTEND_VARIANT STREQUAL "MSVC"))
string(APPEND CMAKE_EXE_LINKER_FLAGS " -Wl,/OPT:REF")
string(APPEND CMAKE_SHARED_LINKER_FLAGS " -Wl,/OPT:REF")
elseif(APPLE)
string(APPEND CMAKE_EXE_LINKER_FLAGS " -Wl,-dead_strip")
string(APPEND CMAKE_SHARED_LINKER_FLAGS " -Wl,-dead_strip")
else()
string(APPEND CMAKE_EXE_LINKER_FLAGS " -static-libstdc++ -static-libgcc -Wl,--gc-sections")
string(APPEND CMAKE_SHARED_LINKER_FLAGS " -Wl,--gc-sections")
endif()
# Make sure all third libraries are affected by ABI related options
if(CLICE_USE_LIBCXX)
string(APPEND CMAKE_CXX_FLAGS " -stdlib=libc++")
string(APPEND CMAKE_EXE_LINKER_FLAGS " -stdlib=libc++")
string(APPEND CMAKE_SHARED_LINKER_FLAGS " -stdlib=libc++")
endif()
if(CLICE_RELEASE)
set(CLICE_ENABLE_LTO ON)
endif()
if(CLICE_ENABLE_LTO)
string(APPEND CMAKE_C_FLAGS " -flto=thin")
string(APPEND CMAKE_CXX_FLAGS " -flto=thin")
@@ -60,12 +29,17 @@ if(CLICE_ENABLE_LTO)
string(APPEND CMAKE_MODULE_LINKER_FLAGS " -flto=thin")
endif()
if(CMAKE_BUILD_TYPE STREQUAL "Debug")
add_compile_options(-fsanitize=address)
if(CMAKE_CXX_COMPILER_FRONTEND_VARIANT STREQUAL "MSVC")
# clang-cl (MSVC frontend): manually link ASan runtime since clang-cl
# doesn't handle -fsanitize=address linking automatically.
if(NOT WIN32)
string(APPEND CMAKE_EXE_LINKER_FLAGS " -fsanitize=address")
string(APPEND CMAKE_SHARED_LINKER_FLAGS " -fsanitize=address")
endif()
if(MSVC AND CMAKE_CXX_COMPILER_ID MATCHES "Clang")
execute_process(
COMMAND ${CMAKE_CXX_COMPILER} --print-resource-dir
OUTPUT_VARIABLE CLANG_RESOURCE_DIR
@@ -75,47 +49,27 @@ if(CMAKE_BUILD_TYPE STREQUAL "Debug")
link_directories(${ASAN_LIB_PATH})
set(ASAN_LINK_FLAGS "")
list(APPEND ASAN_LINK_FLAGS "clang_rt.asan_dynamic-x86_64.lib")
list(APPEND ASAN_LINK_FLAGS "/wholearchive:clang_rt.asan_dynamic_runtime_thunk-x86_64.lib")
foreach(flag ${ASAN_LINK_FLAGS})
string(APPEND CMAKE_EXE_LINKER_FLAGS " ${flag}")
string(APPEND CMAKE_SHARED_LINKER_FLAGS " ${flag}")
string(APPEND CMAKE_MODULE_LINKER_FLAGS " ${flag}")
set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} ${flag}")
set(CMAKE_SHARED_LINKER_FLAGS "${CMAKE_SHARED_LINKER_FLAGS} ${flag}")
set(CMAKE_MODULE_LINKER_FLAGS "${CMAKE_MODULE_LINKER_FLAGS} ${flag}")
endforeach()
else()
string(APPEND CMAKE_EXE_LINKER_FLAGS " -fsanitize=address")
string(APPEND CMAKE_SHARED_LINKER_FLAGS " -fsanitize=address")
endif()
endif()
if(WIN32)
# Disable Identical COMDAT Folding in Debug to avoid ASan ODR false positives.
string(APPEND CMAKE_EXE_LINKER_FLAGS " -Wl,/OPT:NOICF")
string(APPEND CMAKE_SHARED_LINKER_FLAGS " -Wl,/OPT:NOICF")
endif()
if(APPLE)
# https://conda-forge.org/docs/maintainer/knowledge_base/#newer-c-features-with-old-sdk
string(APPEND CMAKE_CXX_FLAGS " -D_LIBCPP_DISABLE_AVAILABILITY=1")
endif()
include("${PROJECT_SOURCE_DIR}/cmake/package.cmake")
# Project-specific options (not applied to third-party deps).
add_library(clice_options INTERFACE)
if(MSVC OR (CMAKE_CXX_COMPILER_ID MATCHES "Clang" AND
CMAKE_CXX_COMPILER_FRONTEND_VARIANT STREQUAL "MSVC"))
target_compile_options(clice_options INTERFACE /GR- /EHs-c- /Zc:preprocessor)
else()
target_compile_options(clice_options INTERFACE
-fno-rtti
-fno-exceptions
-Wno-deprecated-declarations
$<$<COMPILE_LANG_AND_ID:CXX,Clang,AppleClang>:-Wno-undefined-inline>
)
endif()
if(WIN32)
target_link_libraries(clice_options INTERFACE version ntdll)
endif()
if(CLICE_ENABLE_TEST)
target_compile_definitions(clice_options INTERFACE CLICE_ENABLE_TEST=1)
endif()
@@ -124,47 +78,123 @@ if(CLICE_CI_ENVIRONMENT)
target_compile_definitions(clice_options INTERFACE CLICE_CI_ENVIRONMENT=1)
endif()
set(FBS_SCHEMA_FILE "${PROJECT_SOURCE_DIR}/src/index/schema.fbs")
set(GENERATED_HEADER "${PROJECT_BINARY_DIR}/generated/schema_generated.h")
if(WIN32)
target_link_libraries(clice_options INTERFACE version ntdll)
endif()
if(WIN32)
set(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreadedDLL")
target_link_options(clice_options INTERFACE
-fuse-ld=lld-link
-Wl,/OPT:REF
#,/OPT:NOICF
)
elseif(APPLE)
target_link_options(clice_options INTERFACE
-fuse-ld=lld
-Wl,-dead_strip
)
elseif(CMAKE_SYSTEM_NAME STREQUAL "Linux")
target_link_options(clice_options INTERFACE
-fuse-ld=lld
-static-libstdc++ -static-libgcc
-Wl,--gc-sections
)
endif()
if(MSVC OR (CMAKE_CXX_COMPILER_ID MATCHES "Clang" AND
CMAKE_CXX_COMPILER_FRONTEND_VARIANT STREQUAL "MSVC"))
target_compile_options(clice_options INTERFACE
/GR-
/EHsc-
/Zc:preprocessor
)
else()
target_compile_options(clice_options INTERFACE
-fno-rtti
-fno-exceptions
-Wno-deprecated-declarations
-Wno-undefined-inline
-ffunction-sections
-fdata-sections
)
endif()
set(FBS_SCHEMA_FILE "${CMAKE_CURRENT_SOURCE_DIR}/include/Index/schema.fbs")
set(GENERATED_HEADER "${CMAKE_CURRENT_BINARY_DIR}/generated/schema_generated.h")
add_custom_command(
OUTPUT "${GENERATED_HEADER}"
COMMAND $<TARGET_FILE:flatc> --cpp -o "${PROJECT_BINARY_DIR}/generated" "${FBS_SCHEMA_FILE}"
DEPENDS "${FBS_SCHEMA_FILE}"
OUTPUT ${GENERATED_HEADER}
COMMAND $<TARGET_FILE:flatc> --cpp -o ${CMAKE_CURRENT_BINARY_DIR}/generated ${FBS_SCHEMA_FILE}
DEPENDS ${FBS_SCHEMA_FILE}
COMMENT "Generating C++ header from ${FBS_SCHEMA_FILE}"
)
add_custom_target(generate_flatbuffers_schema DEPENDS "${GENERATED_HEADER}")
file(GLOB_RECURSE CLICE_CORE_SOURCES CONFIGURE_DEPENDS "${PROJECT_SOURCE_DIR}/src/*.cpp")
add_library(clice-core STATIC ${CLICE_CORE_SOURCES})
add_library(clice::core ALIAS clice-core)
add_dependencies(clice-core generate_flatbuffers_schema)
target_include_directories(clice-core PUBLIC
"${PROJECT_SOURCE_DIR}/src"
"${PROJECT_BINARY_DIR}/generated"
add_custom_target(
generate_flatbuffers_schema
DEPENDS ${GENERATED_HEADER}
)
target_link_libraries(clice-core PUBLIC
set(CONFIG_SOURCE_FILE "${CMAKE_CURRENT_SOURCE_DIR}/config/clang-tidy-config.h")
set(CONFIG_GENERATED_FILE "${CMAKE_CURRENT_BINARY_DIR}/generated/clang-tidy-config.h")
add_custom_command(
OUTPUT ${CONFIG_GENERATED_FILE}
COMMAND ${CMAKE_COMMAND} -E copy_if_different ${CONFIG_SOURCE_FILE} ${CONFIG_GENERATED_FILE}
DEPENDS ${CONFIG_SOURCE_FILE}
COMMENT "Generating C++ header from ${CONFIG_SOURCE_FILE}"
)
add_custom_target(
generate_config
DEPENDS ${CONFIG_GENERATED_FILE}
)
add_library(clice_core_deps INTERFACE)
target_include_directories(clice_core_deps INTERFACE
"${PROJECT_SOURCE_DIR}/include"
"${CMAKE_CURRENT_BINARY_DIR}/generated"
)
target_link_libraries(clice_core_deps INTERFACE
clice_options
llvm-libs
libuv::libuv
spdlog::spdlog
tomlplusplus::tomlplusplus
roaring::roaring
flatbuffers
eventide::ipc::lsp
eventide::serde::toml
simdjson::simdjson
llvm-libs
)
add_library(clice_builtin_api INTERFACE)
target_link_libraries(clice_builtin_api INTERFACE clice_core_deps)
include("${PROJECT_SOURCE_DIR}/cmake/builtin-libraries.cmake")
clice_include_builtin_library_modules()
file(GLOB_RECURSE CLICE_SOURCES CONFIGURE_DEPENDS
"${PROJECT_SOURCE_DIR}/src/AST/*.cpp"
"${PROJECT_SOURCE_DIR}/src/Async/*.cpp"
"${PROJECT_SOURCE_DIR}/src/Basic/*.cpp"
"${PROJECT_SOURCE_DIR}/src/Compiler/*.cpp"
"${PROJECT_SOURCE_DIR}/src/Index/*.cpp"
"${PROJECT_SOURCE_DIR}/src/Feature/*.cpp"
"${PROJECT_SOURCE_DIR}/src/Server/*.cpp"
"${PROJECT_SOURCE_DIR}/src/Support/*.cpp"
)
add_library(clice-core STATIC "${CLICE_SOURCES}")
add_dependencies(clice-core generate_flatbuffers_schema generate_config)
target_link_libraries(clice-core PUBLIC clice_core_deps)
clice_finalize_builtin_libraries(TARGET clice-core)
add_executable(clice "${PROJECT_SOURCE_DIR}/src/clice.cc")
target_link_libraries(clice PRIVATE clice::core eventide::deco)
target_link_libraries(clice PRIVATE clice-core)
install(TARGETS clice RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR})
add_custom_target(copy_clang_resource ALL
COMMAND ${CMAKE_COMMAND} -E copy_directory
"${LLVM_INSTALL_PATH}/lib/clang"
"${PROJECT_BINARY_DIR}/lib/clang"
COMMENT "Copying clang resource directory"
message(STATUS "Copying resource directory for development build")
file(
COPY "${LLVM_INSTALL_PATH}/lib/clang"
DESTINATION "${PROJECT_BINARY_DIR}/lib"
)
install(
DIRECTORY "${LLVM_INSTALL_PATH}/lib/clang"
@@ -173,35 +203,11 @@ install(
if(CLICE_ENABLE_TEST)
file(GLOB_RECURSE CLICE_TEST_SOURCES CONFIGURE_DEPENDS
"${PROJECT_SOURCE_DIR}/tests/unit/*/*_tests.cpp"
)
set(CLICE_TEST_SUPPORT_SOURCES
"${PROJECT_SOURCE_DIR}/tests/unit/test/annotation.cpp"
"${PROJECT_SOURCE_DIR}/tests/unit/test/tester.cpp"
)
"${PROJECT_SOURCE_DIR}/tests/unit/*/*.cpp")
add_executable(unit_tests
"${CLICE_TEST_SOURCES}"
"${PROJECT_SOURCE_DIR}/tests/unit/unit_tests.cc"
${CLICE_TEST_SOURCES}
${CLICE_TEST_SUPPORT_SOURCES}
)
target_include_directories(unit_tests PRIVATE
"${PROJECT_SOURCE_DIR}/src"
"${PROJECT_SOURCE_DIR}/tests/unit"
)
target_link_libraries(unit_tests PRIVATE clice::core eventide::zest eventide::deco)
endif()
if(CLICE_ENABLE_BENCHMARK)
add_executable(scan_benchmark
"${PROJECT_SOURCE_DIR}/benchmarks/scan_benchmark.cpp"
)
target_include_directories(scan_benchmark PRIVATE
"${PROJECT_SOURCE_DIR}/src"
)
target_link_libraries(scan_benchmark PRIVATE clice::core eventide::deco)
endif()
if(CLICE_RELEASE)
include("${PROJECT_SOURCE_DIR}/cmake/release.cmake")
target_include_directories(unit_tests PUBLIC "${PROJECT_SOURCE_DIR}")
target_link_libraries(unit_tests PRIVATE clice-core cpptrace::cpptrace)
endif()

View File

@@ -5,7 +5,7 @@
![C++ Standard](https://img.shields.io/badge/C++-23-blue.svg)
[![GitHub license](https://img.shields.io/github/license/clice-io/clice)](https://github.com/clice-io/clice/blob/main/LICENSE)
[![Actions status](https://github.com/clice-io/clice/workflows/CI/badge.svg)](https://github.com/clice-io/clice/actions)
[![Documentation](https://img.shields.io/badge/view-documentation-blue)](https://docs.clice.io/clice/)
[![Documentation](https://img.shields.io/badge/view-documentation-blue)](https://clice.io)
[![Ask DeepWiki](https://deepwiki.com/badge.svg)](https://deepwiki.com/clice-io/clice)
[![Discord](https://img.shields.io/badge/Discord-%235865F2.svg?logo=discord&logoColor=white)](https://discord.gg/PA3UxW2VA3)
@@ -35,4 +35,4 @@ Download the latest `clice` binary from the [releases page](https://github.com/c
## Documentation
To learn more about building, installing, and configuring clice, or to dive deep into its features and architecture, please visit our official documentation at [**clice.io**](https://docs.clice.io/clice/).
To learn more about building, installing, and configuring clice, or to dive deep into its features and architecture, please visit our official documentation at [**clice.io**](https://clice.io/).

View File

@@ -1,413 +0,0 @@
/// Benchmark for scan_dependency_graph on a real compilation database.
///
/// Usage:
/// scan_benchmark [OPTIONS] <compile_commands.json>
///
/// Example:
/// ./build/RelWithDebInfo/bin/scan_benchmark \
/// /home/ykiko/C++/clice/.llvm/build-debug/compile_commands.json
///
/// ./build/RelWithDebInfo/bin/scan_benchmark --log-level info --export graph.json \
/// /home/ykiko/C++/clice/.llvm/build-debug/compile_commands.json
#include <algorithm>
#include <chrono>
#include <cstdlib>
#include <fstream>
#include <map>
#include <numeric>
#include <print>
#include <set>
#include <thread>
#include "command/command.h"
#include "eventide/deco/deco.h"
#include "eventide/serde/json/serializer.h"
#include "support/filesystem.h"
#include "support/logging.h"
#include "support/path_pool.h"
#include "syntax/dependency_graph.h"
#include "llvm/Support/FileSystem.h"
namespace et = eventide;
using namespace clice;
struct BenchmarkOptions {
DecoKV(names = {"--log-level"}; help = "Log level: trace, debug, info, warn, error, off";
required = false;)
<std::string> log_level = "off";
DecoKV(names = {"--export"}; help = "Export dependency graph as JSON to this path";
required = false;)
<std::string> export_path;
DecoKV(names = {"--runs"}; help = "Number of cold start iterations"; required = false;)
<int> runs = 20;
DecoFlag(names = {"-h", "--help"}; help = "Show help message"; required = false;)
help;
DecoInput(meta_var = "CDB"; help = "Path to compile_commands.json"; required = false;)
<std::string> cdb_path;
};
struct FileNode {
std::string path;
std::string module_name;
std::vector<std::string> includes;
};
struct GraphExport {
std::vector<FileNode> files;
};
void export_graph_json(const PathPool& path_pool,
const DependencyGraph& graph,
llvm::StringRef output_path) {
// Build reverse module map: path_id -> module_name.
llvm::DenseMap<std::uint32_t, llvm::StringRef> path_to_module;
for(auto& [name, path_ids]: graph.modules()) {
for(auto path_id: path_ids) {
path_to_module[path_id] = name;
}
}
GraphExport export_data;
for(std::uint32_t id = 0; id < path_pool.paths.size(); id++) {
auto inc_ids = graph.get_all_includes(id);
if(inc_ids.empty()) {
continue;
}
FileNode node;
node.path = path_pool.paths[id].str();
auto mod_it = path_to_module.find(id);
if(mod_it != path_to_module.end()) {
node.module_name = mod_it->second.str();
}
for(auto flagged_id: inc_ids) {
auto raw_id = flagged_id & DependencyGraph::PATH_ID_MASK;
node.includes.push_back(path_pool.paths[raw_id].str());
}
export_data.files.push_back(std::move(node));
}
auto json = et::serde::json::to_json(export_data);
if(!json) {
std::println(stderr, "Failed to serialize dependency graph");
return;
}
std::ofstream out(output_path.str());
if(!out) {
std::println(stderr, "Failed to open output file: {}", output_path);
return;
}
out << *json;
std::println("Graph exported to {} ({} files)", output_path, export_data.files.size());
}
void print_report(const ScanReport& report) {
std::println("===============================================================");
std::println(" Dependency Scan Report");
std::println("===============================================================");
// Timing.
std::println("");
std::println(" Time: {}ms", report.elapsed_ms);
std::println(" Waves: {}", report.waves);
// File counts.
std::println("");
std::println(" Files");
std::println(" Source files (from CDB): {}", report.source_files);
std::println(" Header files (discovered): {}", report.header_files);
std::println(" Total: {}", report.total_files);
std::println(" Modules: {}", report.modules);
// Include edges.
std::println("");
std::println(" Include Edges");
std::println(" Total: {}", report.total_edges);
std::println(" Unconditional: {}", report.unconditional_edges);
std::println(" Conditional: {} (inside #if/#ifdef)", report.conditional_edges);
// Resolution accuracy.
std::println("");
std::println(" Resolution");
std::println(" #include directives: {}", report.includes_found);
std::println(" Resolved: {}", report.includes_resolved);
auto unresolved_count = report.includes_found - report.includes_resolved;
std::println(" Unresolved: {}", unresolved_count);
if(report.includes_found > 0) {
double rate = 100.0 * static_cast<double>(report.includes_resolved) /
static_cast<double>(report.includes_found);
std::println(" Accuracy: {:.1f}%", rate);
}
// Wall-clock phase breakdown.
std::println("");
std::println(" Phase Breakdown (wall-clock)");
std::println(" Config extraction: {}ms (prewarm={}ms, loop={}ms)",
report.config_ms,
report.prewarm_ms,
report.config_loop_ms);
std::println(" Dir cache pre-pop: {}ms (overlapped with Phase 1)", report.dir_cache_ms);
std::println(" Phase 1 (read+scan, parallel): {}ms", report.phase1_ms);
std::println(" Phase 2 (include resolve): {}ms", report.phase2_ms);
std::println(" Phase 3 (graph build): {}ms", report.phase3_ms);
// Per-wave breakdown.
if(!report.wave_stats.empty()) {
std::println("");
std::println(" Per-Wave Breakdown");
std::println(" {:>5s} {:>8s} {:>8s} {:>8s} {:>8s} {:>8s} {:>10s} {:>10s}",
"Wave",
"Files",
"P1(ms)",
"P2(ms)",
"Next",
"Prefetch",
"DirList",
"DirHits");
for(std::size_t i = 0; i < report.wave_stats.size(); i++) {
auto& ws = report.wave_stats[i];
std::println(" {:>5} {:>8} {:>8} {:>8} {:>8} {:>8} {:>10} {:>10}",
i,
ws.files,
ws.phase1_ms,
ws.phase2_ms,
ws.next_files,
ws.prefetch_count,
ws.dir_listings,
ws.dir_hits);
}
}
// Phase 2 breakdown.
if(report.p2_resolve_us > 0) {
auto other_us = report.phase2_ms * 1000 - report.p2_resolve_us;
std::println("");
std::println(" Phase 2 Breakdown (single-threaded)");
std::println(" resolve_include: {:.1f}ms", report.p2_resolve_us / 1000.0);
std::println(" Other (cache lookup, intern, graph): {:.1f}ms", other_us / 1000.0);
}
// Cumulative I/O statistics.
std::println("");
std::println(" I/O Statistics (cumulative across threads)");
std::println(" File read: {:.1f}ms (sum of all threads)", report.read_us / 1000.0);
std::println(" Lexer scan: {:.1f}ms (sum of all threads)", report.scan_us / 1000.0);
std::println(" Filesystem: {:.1f}ms ({} readdir calls, {} dir cache hits)",
report.fs_us / 1000.0,
report.dir_listings,
report.dir_hits);
std::println(" File lookups: {}", report.fs_lookups);
std::println(" Include cache hits: {}", report.include_cache_hits);
std::println(" Scan result cache hits: {}", report.scan_cache_hits);
if(report.dir_listings + report.dir_hits > 0) {
double hit_rate = 100.0 * static_cast<double>(report.dir_hits) /
static_cast<double>(report.dir_listings + report.dir_hits);
std::println(" Dir cache hit rate: {:.1f}%", hit_rate);
}
std::println("");
std::println("===============================================================");
}
int main(int argc, const char** argv) {
auto args = deco::util::argvify(argc, argv);
auto result = deco::cli::parse<BenchmarkOptions>(args);
if(!result.has_value()) {
std::println(stderr, "Error: {}", result.error().message);
return 1;
}
auto& opts = result->options;
if(opts.help.value_or(false) || !opts.cdb_path.has_value()) {
std::ostringstream oss;
deco::cli::write_usage_for<BenchmarkOptions>(oss, "scan_benchmark [OPTIONS] <cdb>");
std::print("{}", oss.str());
return opts.help.value_or(false) ? 0 : 1;
}
// Configure logging.
auto level = spdlog::level::from_str(*opts.log_level);
clice::logging::options.level = level;
clice::logging::stderr_logger("scan_benchmark", clice::logging::options);
// resource_dir() is self-initializing (lazy static) — no setup needed.
auto& cdb_path = *opts.cdb_path;
auto hw_threads = std::thread::hardware_concurrency();
auto runs = *opts.runs;
if(runs <= 0) {
std::println(stderr, "Error: --runs must be positive (got {})", runs);
return 1;
}
// Set UV_THREADPOOL_SIZE if not already set.
// Use at least libuv's default (4) so low-core CI runners don't regress.
if(!std::getenv("UV_THREADPOOL_SIZE")) {
auto pool_size = std::max(hw_threads, 4u);
static std::string env = "UV_THREADPOOL_SIZE=" + std::to_string(pool_size);
putenv(env.data());
}
std::println("Hardware threads: {}", hw_threads);
std::println("UV_THREADPOOL_SIZE: {}", std::getenv("UV_THREADPOOL_SIZE"));
std::println("Log level: {}", *opts.log_level);
std::println("CDB: {}", cdb_path);
std::println("");
// Load compilation database.
auto t0 = std::chrono::steady_clock::now();
CompilationDatabase cdb;
auto count = cdb.load(cdb_path);
auto t1 = std::chrono::steady_clock::now();
auto load_ms = std::chrono::duration_cast<std::chrono::milliseconds>(t1 - t0).count();
std::println("CDB loaded: {} entries in {}ms", count, load_ms);
{
std::set<const CompilationInfo*> unique_contexts;
std::set<const CanonicalCommand*> unique_canonicals;
std::map<const CanonicalCommand*, int> canonical_hist;
for(auto& entry: cdb.get_entries()) {
unique_contexts.insert(entry.info.ptr);
unique_canonicals.insert(entry.info->canonical.ptr);
canonical_hist[entry.info->canonical.ptr]++;
}
double dedup_ratio =
unique_contexts.empty() ? 0.0 : static_cast<double>(count) / unique_contexts.size();
std::println(
"Context dedup: {} files -> {} unique contexts ({:.1f}x), {} unique canonicals",
count,
unique_contexts.size(),
dedup_ratio,
unique_canonicals.size());
// If canonical dedup is poor, dump diagnostics.
if(unique_canonicals.size() > 200) {
// Sort canonicals by frequency (descending).
std::vector<std::pair<int, const CanonicalCommand*>> sorted;
for(auto& [ptr, cnt]: canonical_hist)
sorted.push_back({cnt, ptr});
std::ranges::sort(sorted,
std::greater{},
&std::pair<int, const CanonicalCommand*>::first);
// Show top-5 canonical commands.
for(int i = 0; i < std::min(5, (int)sorted.size()); i++) {
auto [cnt, cmd] = sorted[i];
std::println(" canonical[{}] ({} files, {} args):", i, cnt, cmd->arguments.size());
for(auto arg: cmd->arguments)
std::println(" {}", arg);
}
// Show a singleton canonical (count==1) to see what per-file arg leaks in.
for(auto& [cnt, cmd]: sorted) {
if(cnt == 1) {
std::println(" singleton canonical ({} args):", cmd->arguments.size());
for(auto arg: cmd->arguments)
std::println(" {}", arg);
break;
}
}
// Find two canonicals that differ by only a few args.
if(sorted.size() >= 2) {
auto* a = sorted[0].second;
auto* b = sorted[1].second;
std::println(" --- Canonical diff (top-1 vs top-2) ---");
auto max_len = std::max(a->arguments.size(), b->arguments.size());
for(std::size_t i = 0; i < max_len; i++) {
llvm::StringRef av = i < a->arguments.size() ? a->arguments[i] : "<missing>";
llvm::StringRef bv = i < b->arguments.size() ? b->arguments[i] : "<missing>";
if(av != bv)
std::println(" DIFF[{}]: '{}' vs '{}'", i, av, bv);
else
std::println(" SAME[{}]: '{}'", i, av);
}
}
}
}
std::println("\nRunning {} cold start scan(s)...\n", runs);
PathPool path_pool;
DependencyGraph graph;
std::vector<std::int64_t> elapsed_times;
std::vector<std::int64_t> config_times;
std::vector<std::int64_t> phase1_times;
std::vector<std::int64_t> phase2_times;
elapsed_times.reserve(runs);
config_times.reserve(runs);
phase1_times.reserve(runs);
phase2_times.reserve(runs);
for(int i = 0; i < runs; i++) {
// True cold start: rebuild CDB (clears toolchain & config caches),
// reset PathPool and DependencyGraph.
cdb = CompilationDatabase{};
cdb.load(cdb_path);
path_pool = PathPool{};
graph = DependencyGraph{};
auto report = scan_dependency_graph(cdb, path_pool, graph);
elapsed_times.push_back(report.elapsed_ms);
config_times.push_back(report.config_ms);
phase1_times.push_back(report.phase1_ms);
phase2_times.push_back(report.phase2_ms);
std::println("[run {:2}] {}ms | config={}ms phase1={}ms phase2={}ms | files={}",
i + 1,
report.elapsed_ms,
report.config_ms,
report.phase1_ms,
report.phase2_ms,
report.total_files);
// Print detailed report for the first run only.
if(i == 0) {
std::println("");
print_report(report);
}
}
// Summary statistics.
if(runs > 1) {
auto stats = [](std::vector<std::int64_t>& v) {
std::ranges::sort(v);
auto sum = std::accumulate(v.begin(), v.end(), std::int64_t{0});
return std::tuple{v.front(), sum / static_cast<std::int64_t>(v.size()), v.back()};
};
auto [e_min, e_avg, e_max] = stats(elapsed_times);
auto [c_min, c_avg, c_max] = stats(config_times);
auto [p1_min, p1_avg, p1_max] = stats(phase1_times);
auto [p2_min, p2_avg, p2_max] = stats(phase2_times);
std::println("\n Summary ({} runs) min avg max", runs);
std::println(" Total: {:>7} {:>6} {:>6}", e_min, e_avg, e_max);
std::println(" Config extraction: {:>7} {:>6} {:>6}", c_min, c_avg, c_max);
std::println(" Phase 1 (read+scan):{:>7} {:>6} {:>6}", p1_min, p1_avg, p1_max);
std::println(" Phase 2 (resolve): {:>7} {:>6} {:>6}", p2_min, p2_avg, p2_max);
}
// Export dependency graph as JSON if requested.
if(opts.export_path.has_value()) {
export_graph_json(path_pool, graph, *opts.export_path);
}
return 0;
}

View File

@@ -1,17 +0,0 @@
if(OUTPUT MATCHES "\\.tar\\.gz$")
execute_process(
COMMAND ${CMAKE_COMMAND} -E tar czf "${OUTPUT}" .
WORKING_DIRECTORY "${WORK_DIR}"
COMMAND_ERROR_IS_FATAL ANY
)
elseif(OUTPUT MATCHES "\\.zip$")
execute_process(
COMMAND ${CMAKE_COMMAND} -E tar cf "${OUTPUT}" --format=zip .
WORKING_DIRECTORY "${WORK_DIR}"
COMMAND_ERROR_IS_FATAL ANY
)
else()
message(FATAL_ERROR "Unsupported archive format: ${OUTPUT}")
endif()
message(STATUS "Created: ${OUTPUT}")

View File

@@ -0,0 +1,158 @@
include_guard(GLOBAL)
set(
CLICE_BUILTIN_LIBRARY_MODULES
"${CLICE_BUILTIN_LIBRARY_MODULES}"
CACHE STRING
"Semicolon-separated list of CMake modules that register extra builtin clice libraries"
)
function(clice_add_builtin_library)
set(options)
set(oneValueArgs NAME ENTRYPOINT)
set(multiValueArgs SOURCES INCLUDE_DIRECTORIES LINK_LIBRARIES COMPILE_DEFINITIONS COMPILE_OPTIONS)
cmake_parse_arguments(CBL "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
if(CBL_UNPARSED_ARGUMENTS)
message(
FATAL_ERROR
"clice_add_builtin_library got unexpected arguments: ${CBL_UNPARSED_ARGUMENTS}"
)
endif()
if(NOT CBL_NAME)
message(FATAL_ERROR "clice_add_builtin_library requires NAME")
endif()
if(NOT CBL_SOURCES)
message(FATAL_ERROR "clice_add_builtin_library(${CBL_NAME}) requires SOURCES")
endif()
if(NOT CBL_ENTRYPOINT)
message(FATAL_ERROR "clice_add_builtin_library(${CBL_NAME}) requires ENTRYPOINT")
endif()
if(NOT TARGET clice_builtin_api)
message(
FATAL_ERROR
"clice_builtin_api must be defined before calling clice_add_builtin_library(${CBL_NAME})"
)
endif()
set(target "clice_builtin_${CBL_NAME}")
if(TARGET "${target}")
message(FATAL_ERROR "builtin library target '${target}' already exists")
endif()
add_library("${target}" OBJECT)
target_sources("${target}" PRIVATE ${CBL_SOURCES})
target_link_libraries("${target}" PUBLIC clice_builtin_api)
add_dependencies("${target}" generate_flatbuffers_schema generate_config)
if(CBL_INCLUDE_DIRECTORIES)
target_include_directories("${target}" PRIVATE ${CBL_INCLUDE_DIRECTORIES})
endif()
if(CBL_LINK_LIBRARIES)
target_link_libraries("${target}" PUBLIC ${CBL_LINK_LIBRARIES})
endif()
if(CBL_COMPILE_DEFINITIONS)
target_compile_definitions("${target}" PRIVATE ${CBL_COMPILE_DEFINITIONS})
endif()
if(CBL_COMPILE_OPTIONS)
target_compile_options("${target}" PRIVATE ${CBL_COMPILE_OPTIONS})
endif()
set_property(GLOBAL APPEND PROPERTY CLICE_BUILTIN_LIBRARY_TARGETS "${target}")
set_property(GLOBAL APPEND PROPERTY CLICE_BUILTIN_LIBRARY_ENTRYPOINTS "${CBL_ENTRYPOINT}")
endfunction()
function(clice_include_builtin_library_modules)
foreach(module_path IN LISTS CLICE_BUILTIN_LIBRARY_MODULES)
cmake_path(
ABSOLUTE_PATH module_path
BASE_DIRECTORY "${PROJECT_SOURCE_DIR}"
NORMALIZE
OUTPUT_VARIABLE module_abs_path
)
if(NOT EXISTS "${module_abs_path}")
message(
FATAL_ERROR
"builtin library module '${module_path}' does not exist: ${module_abs_path}"
)
endif()
include("${module_abs_path}")
endforeach()
endfunction()
function(clice_finalize_builtin_libraries)
set(options)
set(oneValueArgs TARGET REGISTRATION_SOURCE)
cmake_parse_arguments(CBL "${options}" "${oneValueArgs}" "" ${ARGN})
if(CBL_UNPARSED_ARGUMENTS)
message(
FATAL_ERROR
"clice_finalize_builtin_libraries got unexpected arguments: ${CBL_UNPARSED_ARGUMENTS}"
)
endif()
if(NOT CBL_TARGET)
message(FATAL_ERROR "clice_finalize_builtin_libraries requires TARGET")
endif()
if(NOT TARGET "${CBL_TARGET}")
message(FATAL_ERROR "target '${CBL_TARGET}' does not exist")
endif()
if(NOT CBL_REGISTRATION_SOURCE)
set(CBL_REGISTRATION_SOURCE "${CMAKE_CURRENT_BINARY_DIR}/generated/builtin-libraries.cpp")
endif()
get_property(builtin_targets GLOBAL PROPERTY CLICE_BUILTIN_LIBRARY_TARGETS)
get_property(builtin_entrypoints GLOBAL PROPERTY CLICE_BUILTIN_LIBRARY_ENTRYPOINTS)
if(NOT builtin_targets)
set(builtin_targets "")
endif()
if(NOT builtin_entrypoints)
set(builtin_entrypoints "")
endif()
set(registration_source_content "#include \"Server/Plugin.h\"\n\nnamespace clice {\n\n")
foreach(entrypoint IN LISTS builtin_entrypoints)
string(APPEND registration_source_content "::clice::PluginInfo ${entrypoint}();\n")
endforeach()
string(APPEND registration_source_content "\nvoid register_builtin_server_plugins(ServerPluginBuilder& builder) {\n")
foreach(entrypoint IN LISTS builtin_entrypoints)
string(
APPEND
registration_source_content
" ${entrypoint}().register_server_callbacks(builder);\n"
)
endforeach()
string(APPEND registration_source_content "}\n\n} // namespace clice\n")
get_filename_component(registration_source_dir "${CBL_REGISTRATION_SOURCE}" DIRECTORY)
file(MAKE_DIRECTORY "${registration_source_dir}")
file(CONFIGURE OUTPUT "${CBL_REGISTRATION_SOURCE}" CONTENT "${registration_source_content}" @ONLY)
target_sources("${CBL_TARGET}" PRIVATE "${CBL_REGISTRATION_SOURCE}")
if(builtin_targets)
foreach(builtin_target IN LISTS builtin_targets)
target_sources("${CBL_TARGET}" PRIVATE $<TARGET_OBJECTS:${builtin_target}>)
endforeach()
target_link_libraries("${CBL_TARGET}" PRIVATE ${builtin_targets})
endif()
endfunction()

View File

@@ -101,15 +101,8 @@ function(setup_llvm LLVM_VERSION)
clangToolingSyntax
)
else()
file(GLOB LLVM_LIBRARIES CONFIGURE_DEPENDS "${LLVM_INSTALL_PATH}/lib/${CMAKE_STATIC_LIBRARY_PREFIX}LLVM[a-zA-Z]*${CMAKE_STATIC_LIBRARY_SUFFIX}")
file(GLOB CLANG_LIBRARIES CONFIGURE_DEPENDS "${LLVM_INSTALL_PATH}/lib/${CMAKE_STATIC_LIBRARY_PREFIX}clang[a-zA-Z]*${CMAKE_STATIC_LIBRARY_SUFFIX}")
# TODO: find a better way to find out whether zlib and zstd are needed
# Currently link if present in the LLVM lib directory
file(GLOB OTHER_REQUIRED_LIBS CONFIGURE_DEPENDS
"${LLVM_INSTALL_PATH}/lib/${CMAKE_STATIC_LIBRARY_PREFIX}z${CMAKE_STATIC_LIBRARY_SUFFIX}"
"${LLVM_INSTALL_PATH}/lib/${CMAKE_STATIC_LIBRARY_PREFIX}zstd${CMAKE_STATIC_LIBRARY_SUFFIX}"
)
target_link_libraries(llvm-libs INTERFACE ${LLVM_LIBRARIES} ${CLANG_LIBRARIES} ${OTHER_REQUIRED_LIBS})
file(GLOB LLVM_LIBRARIES CONFIGURE_DEPENDS "${LLVM_INSTALL_PATH}/lib/*${CMAKE_STATIC_LIBRARY_SUFFIX}")
target_link_libraries(llvm-libs INTERFACE ${LLVM_LIBRARIES})
target_compile_definitions(llvm-libs INTERFACE CLANG_BUILD_STATIC=1)
endif()
endfunction()

View File

@@ -7,22 +7,50 @@ setup_llvm("21.1.4+r1")
include(FetchContent)
set(FETCHCONTENT_UPDATES_DISCONNECTED ON)
if(WIN32)
set(NULL_DEVICE NUL)
else()
set(NULL_DEVICE /dev/null)
endif()
# libuv
FetchContent_Declare(
libuv
GIT_REPOSITORY https://github.com/libuv/libuv.git
GIT_TAG v1.x
GIT_SHALLOW TRUE
)
if(NOT WIN32 AND CMAKE_BUILD_TYPE STREQUAL "Debug")
set(ASAN ON CACHE BOOL "Enable AddressSanitizer for libuv" FORCE)
endif()
set(LIBUV_BUILD_SHARED OFF CACHE BOOL "" FORCE)
set(LIBUV_BUILD_TESTS OFF CACHE BOOL "" FORCE)
set(BUILD_SHARED_LIBS OFF CACHE BOOL "" FORCE)
# spdlog
FetchContent_Declare(
spdlog
GIT_REPOSITORY https://github.com/gabime/spdlog.git
GIT_TAG v1.15.3
GIT_SHALLOW TRUE
GIT_TAG v1.15.3
GIT_SHALLOW TRUE
)
# tomlplusplus
FetchContent_Declare(
tomlplusplus
GIT_REPOSITORY https://github.com/marzer/tomlplusplus.git
GIT_TAG v3.4.0
GIT_SHALLOW TRUE
)
set(SPDLOG_USE_STD_FORMAT ON CACHE BOOL "" FORCE)
set(SPDLOG_NO_EXCEPTIONS ON CACHE BOOL "" FORCE)
# croaring
FetchContent_Declare(
croaring
GIT_REPOSITORY https://github.com/RoaringBitmap/CRoaring.git
GIT_TAG v4.4.2
GIT_SHALLOW TRUE
GIT_TAG v4.4.2
GIT_SHALLOW TRUE
)
set(ENABLE_ROARING_TESTS OFF CACHE INTERNAL "" FORCE)
set(ENABLE_ROARING_MICROBENCHMARKS OFF CACHE INTERNAL "" FORCE)
@@ -31,26 +59,43 @@ set(ENABLE_ROARING_MICROBENCHMARKS OFF CACHE INTERNAL "" FORCE)
FetchContent_Declare(
flatbuffers
GIT_REPOSITORY https://github.com/google/flatbuffers.git
GIT_TAG v25.9.23
GIT_SHALLOW TRUE
GIT_TAG v25.9.23
GIT_SHALLOW TRUE
)
set(FLATBUFFERS_BUILD_GRPC OFF CACHE BOOL "" FORCE)
set(FLATBUFFERS_BUILD_TESTS OFF CACHE BOOL "" FORCE)
set(FLATBUFFERS_BUILD_FLATHASH OFF CACHE BOOL "" FORCE)
FetchContent_Declare(
eventide
GIT_REPOSITORY https://github.com/clice-io/eventide
GIT_TAG main
GIT_SHALLOW TRUE
FetchContent_MakeAvailable(libuv spdlog tomlplusplus croaring flatbuffers)
if(CLICE_ENABLE_TEST)
# cpptrace
FetchContent_Declare(
cpptrace
GIT_REPOSITORY https://github.com/jeremy-rifkin/cpptrace.git
GIT_TAG v1.0.4
GIT_SHALLOW TRUE
)
set(CPPTRACE_DISABLE_CXX_20_MODULES ON CACHE BOOL "" FORCE)
FetchContent_MakeAvailable(cpptrace)
endif()
if(WIN32)
target_compile_definitions(uv_a PRIVATE _CRT_SECURE_NO_WARNINGS)
endif()
if(NOT MSVC AND TARGET uv_a)
target_compile_options(uv_a PRIVATE
"-Wno-unused-function"
"-Wno-unused-variable"
"-Wno-unused-but-set-variable"
"-Wno-deprecated-declarations"
"-Wno-missing-braces"
)
endif()
target_compile_definitions(spdlog PUBLIC
SPDLOG_USE_STD_FORMAT=1
SPDLOG_NO_EXCEPTIONS=1
)
set(ETD_ENABLE_ZEST ON)
set(ETD_ENABLE_TEST OFF)
set(ETD_SERDE_ENABLE_SIMDJSON ON)
set(ETD_SERDE_ENABLE_YYJSON ON)
set(ETD_SERDE_ENABLE_TOML ON)
set(ETD_ENABLE_EXCEPTIONS OFF)
set(ETD_ENABLE_RTTI OFF)
FetchContent_MakeAvailable(eventide spdlog croaring flatbuffers)

View File

@@ -1,78 +0,0 @@
include_guard()
set(CLICE_PACK_DIR "${PROJECT_BINARY_DIR}/pack")
set(CLICE_SYMBOL_DIR "${PROJECT_BINARY_DIR}/pack-symbol")
if(WIN32)
set(CLICE_ARCHIVE_EXT ".zip")
set(CLICE_SYMBOL_NAME "clice.pdb")
else()
set(CLICE_ARCHIVE_EXT ".tar.gz")
if(APPLE)
set(CLICE_SYMBOL_NAME "clice.dSYM")
else()
set(CLICE_SYMBOL_NAME "clice.debug")
endif()
endif()
if(WIN32)
add_custom_target(clice-strip ALL
COMMAND ${CMAKE_COMMAND} -E make_directory "${CLICE_SYMBOL_DIR}"
COMMAND ${CMAKE_COMMAND} -E copy_if_different
"$<TARGET_PDB_FILE:clice>"
"${CLICE_SYMBOL_DIR}/${CLICE_SYMBOL_NAME}"
DEPENDS clice
COMMENT "Collecting PDB for clice"
)
elseif(APPLE)
add_custom_target(clice-strip ALL
COMMAND ${CMAKE_COMMAND} -E make_directory "${CLICE_SYMBOL_DIR}"
COMMAND dsymutil "$<TARGET_FILE:clice>" -o "${CLICE_SYMBOL_DIR}/${CLICE_SYMBOL_NAME}"
COMMAND strip -x "$<TARGET_FILE:clice>"
DEPENDS clice
COMMENT "Extracting dSYM and stripping clice"
)
else()
add_custom_target(clice-strip ALL
COMMAND ${CMAKE_COMMAND} -E make_directory "${CLICE_SYMBOL_DIR}"
COMMAND ${CMAKE_OBJCOPY} --only-keep-debug "$<TARGET_FILE:clice>" "${CLICE_SYMBOL_DIR}/${CLICE_SYMBOL_NAME}"
COMMAND ${CMAKE_STRIP} --strip-debug --strip-unneeded "$<TARGET_FILE:clice>"
COMMAND ${CMAKE_OBJCOPY} --add-gnu-debuglink="${CLICE_SYMBOL_DIR}/${CLICE_SYMBOL_NAME}" "$<TARGET_FILE:clice>"
DEPENDS clice
COMMENT "Extracting debug symbols and stripping clice"
)
endif()
add_custom_target(clice-pack ALL
DEPENDS clice-strip copy_clang_resource
COMMAND ${CMAKE_COMMAND} -E rm -rf "${CLICE_PACK_DIR}"
COMMAND ${CMAKE_COMMAND} -E make_directory "${CLICE_PACK_DIR}/clice/bin"
COMMAND ${CMAKE_COMMAND} -E copy "$<TARGET_FILE:clice>" "${CLICE_PACK_DIR}/clice/bin/"
COMMAND ${CMAKE_COMMAND} -E copy_directory "${LLVM_INSTALL_PATH}/lib/clang" "${CLICE_PACK_DIR}/clice/lib/clang"
COMMAND ${CMAKE_COMMAND} -E copy "${PROJECT_SOURCE_DIR}/docs/clice.toml" "${CLICE_PACK_DIR}/clice/"
COMMAND ${CMAKE_COMMAND}
-DOUTPUT="${PROJECT_BINARY_DIR}/clice${CLICE_ARCHIVE_EXT}"
-DWORK_DIR="${CLICE_PACK_DIR}"
-P "${PROJECT_SOURCE_DIR}/cmake/archive.cmake"
COMMENT "Packaging clice distribution"
)
if(APPLE)
set(CLICE_COPY_SYMBOL_CMD ${CMAKE_COMMAND} -E copy_directory
"${CLICE_SYMBOL_DIR}/${CLICE_SYMBOL_NAME}" "${CLICE_SYMBOL_DIR}/pack/${CLICE_SYMBOL_NAME}")
else()
set(CLICE_COPY_SYMBOL_CMD ${CMAKE_COMMAND} -E copy
"${CLICE_SYMBOL_DIR}/${CLICE_SYMBOL_NAME}" "${CLICE_SYMBOL_DIR}/pack/")
endif()
add_custom_target(clice-pack-symbol ALL
DEPENDS clice-strip
COMMAND ${CMAKE_COMMAND} -E rm -rf "${CLICE_SYMBOL_DIR}/pack"
COMMAND ${CMAKE_COMMAND} -E make_directory "${CLICE_SYMBOL_DIR}/pack"
COMMAND ${CLICE_COPY_SYMBOL_CMD}
COMMAND ${CMAKE_COMMAND}
-DOUTPUT="${PROJECT_BINARY_DIR}/clice-symbol${CLICE_ARCHIVE_EXT}"
-DWORK_DIR="${CLICE_SYMBOL_DIR}/pack"
-P "${PROJECT_SOURCE_DIR}/cmake/archive.cmake"
COMMENT "Packaging clice debug symbols"
)

View File

@@ -1,49 +1,44 @@
cmake_minimum_required(VERSION 3.30)
set(CMAKE_C_COMPILER clang CACHE STRING "")
set(CMAKE_CXX_COMPILER clang++ CACHE STRING "")
if(WIN32)
set(CMAKE_C_COMPILER clang-cl CACHE STRING "C compiler")
set(CMAKE_CXX_COMPILER clang-cl CACHE STRING "C++ compiler")
set(AR_PROGRAM_NAME "llvm-lib")
set(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreadedDLL" CACHE STRING "MSVC runtime")
find_program(LLVM_AR_PATH "llvm-ar")
find_program(LLVM_LLD_LINK_PATH "lld-link")
if(LLVM_LLD_LINK_PATH)
set(CMAKE_LINKER "${LLVM_LLD_LINK_PATH}" CACHE FILEPATH "Linker")
endif()
else()
set(CMAKE_C_COMPILER clang CACHE STRING "C compiler")
set(CMAKE_CXX_COMPILER clang++ CACHE STRING "C++ compiler")
set(AR_PROGRAM_NAME "llvm-ar")
set(CMAKE_EXE_LINKER_FLAGS "-fuse-ld=lld" CACHE STRING "Executable linker flags")
set(CMAKE_SHARED_LINKER_FLAGS "-fuse-ld=lld" CACHE STRING "Shared library linker flags")
set(CMAKE_MODULE_LINKER_FLAGS "-fuse-ld=lld" CACHE STRING "Module linker flags")
endif()
find_program(LLVM_AR_PATH ${AR_PROGRAM_NAME})
if(LLVM_AR_PATH)
set(CMAKE_AR "${LLVM_AR_PATH}" CACHE FILEPATH "")
set(CMAKE_C_COMPILER_AR "${LLVM_AR_PATH}" CACHE FILEPATH "")
set(CMAKE_CXX_COMPILER_AR "${LLVM_AR_PATH}" CACHE FILEPATH "")
set(CMAKE_AR "${LLVM_AR_PATH}" CACHE FILEPATH "Archiver")
set(CMAKE_C_COMPILER_AR "${LLVM_AR_PATH}" CACHE FILEPATH "C archiver")
set(CMAKE_CXX_COMPILER_AR "${LLVM_AR_PATH}" CACHE FILEPATH "C++ archiver")
endif()
find_program(LLVM_RANLIB_PATH "llvm-ranlib")
if(LLVM_RANLIB_PATH)
set(CMAKE_RANLIB "${LLVM_RANLIB_PATH}" CACHE FILEPATH "")
set(CMAKE_C_COMPILER_RANLIB "${LLVM_RANLIB_PATH}" CACHE FILEPATH "")
set(CMAKE_CXX_COMPILER_RANLIB "${LLVM_RANLIB_PATH}" CACHE FILEPATH "")
set(CMAKE_RANLIB "${LLVM_RANLIB_PATH}" CACHE FILEPATH "Ranlib")
set(CMAKE_C_COMPILER_RANLIB "${LLVM_RANLIB_PATH}" CACHE FILEPATH "C ranlib")
set(CMAKE_CXX_COMPILER_RANLIB "${LLVM_RANLIB_PATH}" CACHE FILEPATH "C++ ranlib")
endif()
find_program(LLVM_NM_PATH "llvm-nm")
if(LLVM_NM_PATH)
set(CMAKE_NM "${LLVM_NM_PATH}" CACHE FILEPATH "")
set(CMAKE_NM "${LLVM_NM_PATH}" CACHE FILEPATH "Symbol lister")
endif()
find_program(LLVM_RC_PATH "llvm-rc")
if(LLVM_RC_PATH)
set(CMAKE_RC_COMPILER "${LLVM_RC_PATH}" CACHE FILEPATH "")
endif()
if(WIN32)
find_program(SCCACHE_PATH "sccache")
if(SCCACHE_PATH)
set(CMAKE_C_COMPILER_LAUNCHER "${SCCACHE_PATH}" CACHE FILEPATH "")
set(CMAKE_CXX_COMPILER_LAUNCHER "${SCCACHE_PATH}" CACHE FILEPATH "")
endif()
set(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreadedDLL" CACHE STRING "")
set(CMAKE_EXE_LINKER_FLAGS_INIT "-fuse-ld=lld-link")
set(CMAKE_SHARED_LINKER_FLAGS_INIT "-fuse-ld=lld-link")
set(CMAKE_MODULE_LINKER_FLAGS_INIT "-fuse-ld=lld-link")
else()
find_program(CCACHE_PATH "ccache")
if(CCACHE_PATH)
set(CMAKE_C_COMPILER_LAUNCHER "${CCACHE_PATH}" CACHE FILEPATH "")
set(CMAKE_CXX_COMPILER_LAUNCHER "${CCACHE_PATH}" CACHE FILEPATH "")
endif()
set(CMAKE_EXE_LINKER_FLAGS_INIT "-fuse-ld=lld")
set(CMAKE_SHARED_LINKER_FLAGS_INIT "-fuse-ld=lld")
set(CMAKE_MODULE_LINKER_FLAGS_INIT "-fuse-ld=lld")
set(CMAKE_RC_COMPILER "${LLVM_RC_PATH}" CACHE FILEPATH "Resource compiler")
endif()

View File

@@ -6,7 +6,7 @@ export default defineConfig({
title: "clice",
description: "a powerful and modern C++ language server",
cleanUrls: true,
base: "/clice/",
base: "/",
rewrites: {
"en/:rest*": ":rest*",
},

View File

@@ -20,7 +20,7 @@ max_active_file = 8
cache_dir = "${workspace}/.clice/cache"
# Directory for storing index files.
index_dir = "${workspace}/.clice/index"
logging_dir = "${workspace}/.clice/logs"
logging_dir = "${workspace}/.clice/logging"
# Compile commands files or directories to search for compile_commands.json files.
compile_commands_paths = ["${workspace}/build"]

View File

@@ -1,183 +0,0 @@
# Server Architecture
clice uses a **multi-process architecture** where a single **Master Server** coordinates multiple **Worker** processes. This design isolates Clang AST operations (which are memory-heavy and may crash) from the main LSP event loop.
## Overview
```
┌──────────────┐ JSON/LSP ┌────────────────┐ Bincode/IPC ┌──────────────────┐
│ LSP Client │ ◄──────────► │ Master Server │ ◄─────────────► │ Stateful Workers │
│ (Editor) │ (stdio) │ │ (stdio) │ (AST cache) │
└──────────────┘ │ - Lifecycle │ └──────────────────┘
│ - Documents │
│ - CDB │ Bincode/IPC ┌──────────────────┐
│ - Build drain │ ◄─────────────► │ Stateless Workers│
│ - Indexing │ (stdio) │ (one-shot tasks)│
└────────────────┘ └──────────────────┘
```
## Master Server
The master server (`src/server/master_server.cpp`) is the central coordinator. It runs a single-threaded async event loop and never touches Clang directly. Its responsibilities:
### LSP Lifecycle
The server progresses through these states:
1. **Uninitialized** — waiting for `initialize` request
2. **Initialized** — capabilities exchanged, waiting for `initialized` notification
3. **Ready** — workers spawned, workspace loaded, accepting requests
4. **ShuttingDown**`shutdown` received, draining work
5. **Exited**`exit` received, stopping the event loop
On `initialized`, the master:
- Loads configuration from `clice.toml` (or uses defaults)
- Starts the worker pool (spawns stateful + stateless processes)
- Loads `compile_commands.json` and builds an include graph
- Starts the background indexer coroutine (if enabled)
### Document Management
Each open document is tracked in a `DocumentState` with:
- Current `version` and `text` (kept in sync via `didOpen`/`didChange`)
- A `generation` counter to detect stale compile results
- Build state flags (`build_running`, `build_requested`, `drain_scheduled`)
When a document is opened or changed:
1. The include graph is re-scanned (via dependency directives)
2. The compile unit is registered/updated in the `CompileGraph`
3. A debounced build is scheduled
### Build Drain
The `run_build_drain` coroutine implements debounced compilation:
1. Wait for the debounce timer (default 200ms) to expire
2. Ensure PCH/PCM dependencies are ready via `CompileGraph`
3. Send a `compile` request to the assigned stateful worker
4. Publish diagnostics from the result (or clear them on failure)
5. If more edits arrived during compilation (`build_requested`), loop back to step 2
This ensures rapid typing doesn't trigger a compile per keystroke.
### Request Routing
Feature requests are split between two worker types:
**Stateful workers** (affinity-routed by file path):
- `textDocument/hover`
- `textDocument/semanticTokens/full`
- `textDocument/inlayHint`
- `textDocument/foldingRange`
- `textDocument/documentSymbol`
- `textDocument/documentLink`
- `textDocument/codeAction`
- `textDocument/definition`
**Stateless workers** (round-robin):
- `textDocument/completion`
- `textDocument/signatureHelp`
All feature responses use `RawValue` passthrough — the worker serializes the LSP result to JSON, and the master forwards the raw JSON bytes to the client without deserializing. This avoids bincode↔JSON conversion overhead and serde annotation conflicts.
## Worker Pool
The worker pool (`src/server/worker_pool.cpp`) manages spawning and communicating with worker processes. Each worker is a child process of the same `clice` binary, launched with `--mode stateful-worker` or `--mode stateless-worker`.
### Communication
Workers communicate with the master via **stdio pipes** using a **bincode** serialization format (via `eventide::ipc::BincodePeer`). This is more compact and faster than JSON for internal IPC, while the master handles JSON for the external LSP protocol.
### Stateful Worker Routing
Stateful workers use **affinity routing**: each file is consistently assigned to the same worker so that the worker retains the cached AST. Assignment uses a **least-loaded** strategy for new files, with **LRU tracking** to manage ownership.
When a worker exceeds its document capacity (currently hardcoded at 16 documents), it evicts the least-recently-used document and notifies the master via an `evicted` notification.
### Stateless Worker Routing
Stateless workers use simple **round-robin** dispatch. Each request includes the full source text and compilation arguments, so any worker can handle it independently.
## Stateful Worker
The stateful worker (`src/server/stateful_worker.cpp`) caches compiled ASTs in memory. Key behavior:
- **Compile**: Parses source code into a `CompilationUnit`, caches the AST, and returns diagnostics as a `RawValue` (JSON bytes)
- **Feature queries**: Look up the cached AST and invoke the corresponding `feature::*` function (hover, semantic tokens, etc.), serializing the result to JSON
- **Document updates**: Received as notifications — the worker updates the stored text and marks the document as `dirty`, causing feature queries to return `null` until recompilation
- **Eviction**: LRU-based; evicts the oldest document when capacity is exceeded, notifying the master
- **Concurrency**: Each document has a per-document `et::mutex` (strand) to serialize compilation and feature queries. Heavy work (compilation, feature extraction) runs on a thread pool via `et::queue`.
## Stateless Worker
The stateless worker (`src/server/stateless_worker.cpp`) handles one-shot requests that don't benefit from cached ASTs:
- **Completion**: Creates a fresh compilation with `CompilationKind::Completion` and invokes `feature::code_complete`
- **Signature help**: Similar to completion, using `feature::signature_help`
- **Build PCH**: Compiles a precompiled header to a temporary file
- **Build PCM**: Compiles a C++20 module interface to a temporary file
- **Index**: Compiles a file for indexing (TUIndex generation — currently a stub)
All requests are dispatched to a thread pool via `et::queue`.
## Compile Graph
The compile graph (`src/server/compile_graph.cpp`) tracks compilation unit dependencies as a DAG. It handles:
- **Registration**: Each file registers its included dependencies
- **Cascade invalidation**: When a file changes, all transitive dependents are marked dirty and their ongoing compilations are cancelled
- **Dependency compilation**: Before compiling a file, `compile_deps` ensures all dependencies (PCH, PCMs) are built first
- **Cancellation**: Uses `et::cancellation_source` to abort in-flight compilations when files are invalidated
## Configuration
The server reads configuration from `clice.toml` (or `.clice/config.toml`) in the workspace root. If no config file exists, sensible defaults are computed from system resources:
| Setting | Default | Description |
| ------------------------ | --------------------- | ------------------------------------------- |
| `stateful_worker_count` | CPU cores / 4 | Number of stateful worker processes |
| `stateless_worker_count` | CPU cores / 4 | Number of stateless worker processes |
| `worker_memory_limit` | 4 GB | Memory limit per stateful worker |
| `compile_commands_path` | auto-detect | Path to `compile_commands.json` |
| `cache_dir` | `<workspace>/.clice/` | Cache directory for PCH/PCM files |
| `debounce_ms` | 200 | Debounce interval for recompilation |
| `enable_indexing` | true | Enable background indexing |
| `idle_timeout_ms` | 3000 | Idle time before background indexing starts |
String values support `${workspace}` substitution.
## IPC Protocol
The master and workers communicate using custom RPC messages defined in `src/server/protocol.h`. Each message type has a `RequestTraits` or `NotificationTraits` specialization that defines the method name and result type.
### Stateful Worker Messages
| Method | Direction | Purpose |
| ----------------------------- | ------------ | ------------------------------------- |
| `clice/worker/compile` | Request | Compile source and return diagnostics |
| `clice/worker/hover` | Request | Get hover info at position |
| `clice/worker/semanticTokens` | Request | Get semantic tokens for file |
| `clice/worker/inlayHints` | Request | Get inlay hints for range |
| `clice/worker/foldingRange` | Request | Get folding ranges |
| `clice/worker/documentSymbol` | Request | Get document symbols |
| `clice/worker/documentLink` | Request | Get document links |
| `clice/worker/codeAction` | Request | Get code actions for range |
| `clice/worker/goToDefinition` | Request | Go to definition at position |
| `clice/worker/documentUpdate` | Notification | Update document text (marks dirty) |
| `clice/worker/evict` | Notification | Master → Worker: evict a document |
| `clice/worker/evicted` | Notification | Worker → Master: document was evicted |
### Stateless Worker Messages
| Method | Direction | Purpose |
| ---------------------------- | --------- | ---------------------------- |
| `clice/worker/completion` | Request | Code completion at position |
| `clice/worker/signatureHelp` | Request | Signature help at position |
| `clice/worker/buildPCH` | Request | Build precompiled header |
| `clice/worker/buildPCM` | Request | Build C++20 module interface |
| `clice/worker/index` | Request | Index a translation unit |

View File

@@ -32,6 +32,18 @@ pixi run integration-test Debug
> [!TIP]
> If you want to develop directly with `cmake`, `ninja`, `clang++`, etc., run `pixi shell` to enter a shell with all env vars configured.
### XMake
We also support building with XMake:
```shell
# config & build (default releasedbg)
pixi run xmake
# unit & integration
pixi run xmake-test
```
## Manual Build
If you plan to build manually, first ensure your toolchain matches the versions defined in `pixi.toml`.
@@ -57,6 +69,24 @@ Optional build options:
| CLICE_ENABLE_TEST | OFF | Build clice unit tests |
| CLICE_USE_LIBCXX | OFF | Build clice with libc++ (adds `-std=libc++`); if enabled, ensure the LLVM libs are also built with libc++ |
| CLICE_CI_ENVIRONMENT | OFF | Enable the `CLICE_CI_ENVIRONMENT` macro; some tests only run in CI |
| CLICE_BUILTIN_LIBRARY_MODULES | "" | Semicolon-separated list of CMake modules that register builtin libraries compiled into `clice`; see [Builtin Libraries](./builtin-library.md) |
### XMake
Build clice with:
```bash
xmake f -c --mode=releasedbg --toolchain=clang
xmake build --all
```
Optional build options:
| Option | Default | Effect |
| ------------- | ------- | ---------------------------------------- |
| --llvm | "" | Build clice with LLVM from a custom path |
| --enable_test | false | Build clice unit tests |
| --ci | false | Enable `CLICE_CI_ENVIRONMENT` |
## About LLVM
@@ -64,7 +94,7 @@ clice calls Clang APIs to parse C++ code, so it must link against LLVM/Clang. Be
Two ways to satisfy this dependency:
1. We publish prebuilt binaries of the LLVM version we use at [clice-llvm](https://github.com/clice-io/clice-llvm/releases) for CI and release builds. During builds, cmake downloads these LLVM libs by default.
1. We publish prebuilt binaries of the LLVM version we use at [clice-llvm](https://github.com/clice-io/clice-llvm/releases) for CI and release builds. During builds, cmake and xmake download these LLVM libs by default.
> [!IMPORTANT]
>

View File

@@ -0,0 +1,101 @@
# Builtin Libraries
Builtin libraries are compiled directly into the `clice` binary instead of being loaded later with `--plugin-path`.
This is useful when:
- your plugin sources live outside the `clice` source tree
- you want the builtin to be part of the default executable
- you need extra include directories, compile definitions, or link dependencies during the main build
## CMake Entry Point
`clice` now exposes a small helper module at [cmake/builtin-libraries.cmake](/cmake/builtin-libraries.cmake).
Extra builtin libraries are registered through the cache variable `CLICE_BUILTIN_LIBRARY_MODULES`.
Each value in `CLICE_BUILTIN_LIBRARY_MODULES` must be a CMake file. During configure, `clice` includes those files, and each file calls `clice_add_builtin_library(...)`.
## Minimal Module
Create a CMake file in your external project, for example `/path/to/my-plugin/clice-builtin.cmake`:
```cmake
clice_add_builtin_library(
NAME my_plugin
SOURCES
"${CMAKE_CURRENT_LIST_DIR}/src/MyPlugin.cpp"
INCLUDE_DIRECTORIES
"${CMAKE_CURRENT_LIST_DIR}/include"
ENTRYPOINT
clice_get_my_plugin_server_plugin_info
)
```
Then configure `clice` with:
```shell
cmake -B build -G Ninja \
-DCLICE_BUILTIN_LIBRARY_MODULES="/path/to/my-plugin/clice-builtin.cmake"
```
To load multiple modules, pass a semicolon-separated CMake list:
```shell
cmake -B build -G Ninja \
-DCLICE_BUILTIN_LIBRARY_MODULES="/path/to/a.cmake;/path/to/b.cmake"
```
## `clice_add_builtin_library`
The helper accepts the following arguments:
| Argument | Required | Description |
| --- | --- | --- |
| `NAME` | Yes | Logical name used to create an internal object target |
| `SOURCES` | Yes | Source files compiled into `clice`; absolute paths and out-of-tree paths are supported |
| `ENTRYPOINT` | Yes | Unique function name in namespace `clice` that returns `::clice::PluginInfo` |
| `INCLUDE_DIRECTORIES` | No | Extra include directories for this builtin only |
| `LINK_LIBRARIES` | No | Extra libraries or targets needed by this builtin |
| `COMPILE_DEFINITIONS` | No | Extra compile definitions for this builtin |
| `COMPILE_OPTIONS` | No | Extra compile options for this builtin |
## Entrypoint Requirements
Builtin libraries share a single final executable, so each builtin must use its own unique entrypoint function inside namespace `clice`.
Dynamic plugins use:
```cpp
clice_get_server_plugin_info()
```
Builtin libraries should use a unique name such as:
```cpp
clice_get_my_plugin_server_plugin_info()
```
For example:
```cpp
#include "Server/Plugin.h"
namespace clice {
::clice::PluginInfo clice_get_my_plugin_server_plugin_info() {
return {
CLICE_PLUGIN_API_VERSION,
"MyPlugin",
"v0.0.1",
CLICE_PLUGIN_DEF_HASH,
[](clice::ServerPluginBuilder& builder) {
// register callbacks here
},
};
}
} // namespace clice
```
`clice` generates the static registration glue automatically, so once the module is included, no additional edits to `src/clice.cc` are required.

View File

@@ -0,0 +1,78 @@
You can implement a clice server plugin to extend clice's functionality.
## Use case
When you use `clice` as LSP backend of LLM agents, e.g. claude code, you can add plugin to provide some extra features.
## Writing a plugin
When a plugin is loaded by the server, it will call `clice_get_server_plugin_info` to obtain information about this plugin and about how to register its customization points.
This function needs to be implemented by the plugin, see the example below:
```c++
extern "C" ::clice::PluginInfo LLVM_ATTRIBUTE_WEAK
clice_get_server_plugin_info() {
return {
CLICE_PLUGIN_API_VERSION, "MyPlugin", "v0.1", CLICE_PLUGIN_DEF_HASH,
[](ServerPluginBuilder builder) { ... }
};
}
```
See [PluginProtocol.h](/include/Server/PluginProtocol.h) for more details.
## Compiling a plugin
The plugin must be compiled with the same dependencies and compiler options as clice, otherwise it will cause undefined behavior. [config/llvm-manifest.json](/config/llvm-manifest.json) defines the build information used by clice.
## Loading plugins
For security reasons, clice does not allow loading plugins through configuration files, but must specify the plugin path through command line options.
When `clice` starts, it will load all plugins specified in the command line. You can specify the plugin path through the `--plugin-path` option.
```shell
$ clice --plugin-path /path/to/my-plugin.so
```
## Getting content of `CLICE_PLUGIN_DEF_HASH`
There are two values to return in the `clice_get_server_plugin_info` function.
- `CLICE_PLUGIN_API_VERSION` is used to ensure compability of the `clice_get_server_plugin_info` function between the plugin and the server.
- `CLICE_PLUGIN_DEF_HASH` is used to ensure the consistency of the C++ declarations between the plugin and the server.
To debug the content of `CLICE_PLUGIN_DEF_HASH`, you can run following command:
```shell
$ git clone https://github.com/clice-io/clice.git
$ cd clice
$ git checkout `clice --version --git-describe`
$ python scripts/plugin-def.py content
```
You will get a C source code file, content of which is like this:
```cpp
#if 0
// begin of config/llvm-manifest.json
[
{
"version": "21.1.4+r1",
"filename": "arm64-macos-clang-debug-asan.tar.xz",
"sha256": "7da4b7d63edefecaf11773e7e701c575140d1a07329bbbb038673b6ee4516ff5",
"lto": false,
"asan": true,
"platform": "macosx",
"build_type": "Debug"
},
...
]
...
#endif
```
## Builtin libraries
If you want to compile a plugin directly into the `clice` binary instead of loading it dynamically, see [Builtin Libraries](./builtin-library.md).

View File

@@ -18,6 +18,13 @@ We use pytest to run integration tests. Please refer to `pyproject.toml` to inst
$ pytest -s --log-cli-level=INFO tests/integration --executable=./build/bin/clice
```
If you use xmake as your build system, you can run the tests directly with xmake:
```shell
$ xmake run --verbose unit_tests
$ xmake test --verbose integration_tests/default
```
## Debug
If you want to attach a debugger to clice for debugging, it is recommended to first start clice in socket mode independently, and then connect the client to it.

View File

@@ -54,73 +54,14 @@ bazel run @hedron_compile_commands//:refresh_all
### Visual Studio
Visual Studio (2019 16.1+) can generate a compilation database via CMake integration. Open your project as a CMake project, then configure the generation in `CMakeSettings.json`:
```json
{
"configurations": [
{
"name": "x64-Debug",
"generator": "Ninja",
"buildRoot": "${projectDir}\\build",
"cmakeCommandArgs": "-DCMAKE_EXPORT_COMPILE_COMMANDS=ON"
}
]
}
```
Alternatively, for MSBuild-based projects (`.vcxproj`), you can use [compiledb-vs](https://github.com/pjbroad/compiledb-vs) or [catter](https://github.com/clice-io/catter) to generate the compilation database.
TODO:
### Makefile
For Makefile-based projects, use [bear](https://github.com/rizsotto/Bear) to intercept compilation commands:
```bash
bear -- make
```
This will generate a `compile_commands.json` in the current directory. Note that `bear` requires a clean build to capture all commands — run `make clean` before `bear -- make` if needed.
Alternatively, if you use GNU Make, you can use [compiledb](https://github.com/nicktimko/compiledb):
```bash
compiledb make
```
### Meson
Meson generates a compilation database automatically during setup:
```bash
meson setup build
```
The `compile_commands.json` will be in the `build` directory.
TODO:
### Xmake
Use one of the following approaches to generate a compilation database.
#### Command Line
Run the following command to manually generate a compilation database:
```bash
xmake project -k compile_commands --lsp=clangd build
```
> Compilation database generated manually doesn't automatically update itself. Re-generate if changes are made to the project.
#### VSCode Extension
The Xmake official VSCode extension automatically generates the compilation database when `xmake.lua` is updated. However, it generates the database to the `.vscode` directory by default. Add this setting in `settings.json`:
```json
"xmake.compileCommandsDirectory": "build"
```
to explicitly ask the extension to generate the compilation database in `build`.
### Others
For any other build system, you can use [catter](https://github.com/clice-io/catter) to generate a compilation database. It captures compilation commands through a fake compiler approach and is designed to work reliably with any build system that invokes a compiler executable.
For any other build system, you can try using [bear](https://github.com/rizsotto/Bear) or [scan-build](https://github.com/rizsotto/scan-build) to intercept compilation commands and obtain the compilation database (no guarantee of success). We plan to write a **new tool** in the future that captures compilation commands through a fake compiler approach.

View File

@@ -32,6 +32,18 @@ pixi run integration-test Debug
> [!TIP]
> 如果你想直接使用 `cmake`, `ninja`, `clang++` 等命令进行开发,请运行 `pixi shell` 进入已配置好环境变量的终端
### XMake
我们同样支持使用 XMake 构建:
```shell
# config & build (default releasedbg)
pixi run xmake
# unit & integration
pixi run xmake-test
```
## Manual Build
如果你打算手动构建,请务必先确认你的工具链满足 pixi.toml 中定义的版本要求。
@@ -57,6 +69,24 @@ cmake -B build -G Ninja \
| CLICE_ENABLE_TEST | OFF | 是否构建 clice 的单元测试 |
| CLICE_USE_LIBCXX | OFF | 是否使用 libc++ 来构建 clice添加 `-std=libc++`),如果开启,请确保 LLVM 库也是使用 libc++ 编译的 |
| CLICE_CI_ENVIRONMENT | OFF | 是否打开 `CLICE_CI_ENVIRONMENT` 这个宏,有些测试在 CI 环境才会执行 |
| CLICE_BUILTIN_LIBRARY_MODULES | "" | 以分号分隔的 CMake 模块列表,用于注册会被编译进 `clice` 的 builtin library详见 [Builtin Libraries](./builtin-library.md) |
### XMake
使用如下命令即可构建 clice
```bash
xmake f -c --mode=releasedbg --toolchain=clang
xmake build --all
```
可选的构建选项:
| 选项 | 默认值 | 效果 |
| ------------- | ------ | ------------------------------------ |
| --llvm | "" | 使用自定义路径的 LLVM 库来构建 clice |
| --enable_test | false | 是否构建 clice 的单元测试 |
| --ci | false | 是否打开 `CLICE_CI_ENVIRONMENT` |
## About LLVM
@@ -64,7 +94,7 @@ clice 调用 Clang API 来解析 C++ 代码,因此必须链接 LLVM/Clang 库
主要有两种方式解决这个依赖问题:
1. 我们在 [clice-llvm](https://github.com/clice-io/clice-llvm/releases) 上会发布使用的 LLVM 版本的预编译二进制,用于 CI 或者 release 构建。在构建时 cmake 默认会从此处下载 LLVM 库然后使用。
1. 我们在 [clice-llvm](https://github.com/clice-io/clice-llvm/releases) 上会发布使用的 LLVM 版本的预编译二进制,用于 CI 或者 release 构建。在构建时 cmake 和 xmake 默认会从此处下载 LLVM 库然后使用。
> [!IMPORTANT]
>

View File

@@ -0,0 +1,101 @@
# Builtin Libraries
Builtin library 会被直接编译进 `clice` 可执行文件,而不是在运行时通过 `--plugin-path` 动态加载。
这种方式适合以下场景:
- 插件源码位于 `clice` 源码树之外
- 希望该 builtin 默认随 `clice` 可执行文件一起发布
- 需要为该 builtin 单独补充 include 路径、编译宏或链接依赖
## CMake 入口
`clice` 现在提供了一个辅助模块:[cmake/builtin-libraries.cmake](/cmake/builtin-libraries.cmake)。
额外的 builtin library 通过缓存变量 `CLICE_BUILTIN_LIBRARY_MODULES` 注册。
`CLICE_BUILTIN_LIBRARY_MODULES` 中的每一项都必须是一个 CMake 文件。配置阶段 `clice``include()` 这些文件,而每个文件都需要调用 `clice_add_builtin_library(...)`
## 最小示例
你可以在外部项目中创建一个 CMake 文件,例如 `/path/to/my-plugin/clice-builtin.cmake`
```cmake
clice_add_builtin_library(
NAME my_plugin
SOURCES
"${CMAKE_CURRENT_LIST_DIR}/src/MyPlugin.cpp"
INCLUDE_DIRECTORIES
"${CMAKE_CURRENT_LIST_DIR}/include"
ENTRYPOINT
clice_get_my_plugin_server_plugin_info
)
```
然后在配置 `clice` 时传入:
```shell
cmake -B build -G Ninja \
-DCLICE_BUILTIN_LIBRARY_MODULES="/path/to/my-plugin/clice-builtin.cmake"
```
如果要加载多个模块,可以传入以分号分隔的 CMake 列表:
```shell
cmake -B build -G Ninja \
-DCLICE_BUILTIN_LIBRARY_MODULES="/path/to/a.cmake;/path/to/b.cmake"
```
## `clice_add_builtin_library`
该辅助函数支持以下参数:
| 参数 | 必填 | 说明 |
| --- | --- | --- |
| `NAME` | 是 | 逻辑名称,会用于创建内部 object target |
| `SOURCES` | 是 | 要编译进 `clice` 的源文件,支持绝对路径和源码树外路径 |
| `ENTRYPOINT` | 是 | `clice` 命名空间中的唯一函数名,返回值类型为 `::clice::PluginInfo` |
| `INCLUDE_DIRECTORIES` | 否 | 仅对当前 builtin 生效的额外头文件目录 |
| `LINK_LIBRARIES` | 否 | 当前 builtin 额外需要的库或 target |
| `COMPILE_DEFINITIONS` | 否 | 当前 builtin 额外需要的编译宏 |
| `COMPILE_OPTIONS` | 否 | 当前 builtin 额外需要的编译选项 |
## Entrypoint 要求
所有 builtin library 最终都会被链接进同一个可执行文件,因此每个 builtin 都必须在 `clice` 命名空间中使用唯一的入口函数名。
动态插件通常使用:
```cpp
clice_get_server_plugin_info()
```
builtin library 应该改用类似下面的唯一名字:
```cpp
clice_get_my_plugin_server_plugin_info()
```
例如:
```cpp
#include "Server/Plugin.h"
namespace clice {
::clice::PluginInfo clice_get_my_plugin_server_plugin_info() {
return {
CLICE_PLUGIN_API_VERSION,
"MyPlugin",
"v0.0.1",
CLICE_PLUGIN_DEF_HASH,
[](clice::ServerPluginBuilder& builder) {
// 在这里注册回调
},
};
}
} // namespace clice
```
`clice` 会自动生成静态注册代码,因此只要模块被包含进来,就不需要再手动修改 `src/clice.cc`

View File

@@ -30,7 +30,7 @@ pixi run publish-vscode
1. `pixi shell -e node`
2.`editors/vscode` 下运行 `pnpm run watch`(增量构建)
3. VSCode 中使用Run Extension/Launch Extension”调试配置或执行 `code --extensionDevelopmentPath=$(pwd)/editors/vscode`
3. VSCode 中使用Run Extension/Launch Extension” 调试配置,或执行 `code --extensionDevelopmentPath=$(pwd)/editors/vscode`
常用脚本(在 `pixi shell -e node` 下):

View File

@@ -0,0 +1,78 @@
你可以在 clice 中实现一个 server plugin 来扩展 clice 的功能。
## 用例
当你使用 `clice` 作为 LLM 代理的 LSP 后端时,比如 claude code你可以添加插件来提供一些额外功能。
## 编写插件
当一个插件被服务器加载时,它会调用 `clice_get_server_plugin_info` 来获取关于这个插件的信息以及如何注册它的定制点。
这个函数需要由插件实现,请参考下面的示例:
```cpp
extern "C" ::clice::PluginInfo LLVM_ATTRIBUTE_WEAK
clice_get_server_plugin_info() {
return {
CLICE_PLUGIN_API_VERSION, "MyPlugin", "v0.1", CLICE_PLUGIN_DEF_HASH,
[](ServerPluginBuilder builder) { ... }
};
}
```
请参考 [PluginProtocol.h](/include/Server/PluginProtocol.h) 了解更多细节。
## 编译插件
插件必须使用与 clice 一致的依赖和编译器选项来编译,否则会导致 undefined behavior。[config/llvm-manifest.json](/config/llvm-manifest.json) 中定义了 clice 使用的构建信息。
## 加载插件
为了安全考虑clice 不允许通过配置文件来加载插件,而必须通过命令行选项来指定插件的路径。
`clice` 启动时,它会加载所有在命令行中指定的插件。你可以通过 `--plugin-path` 选项来指定插件的路径。
```shell
$ clice --plugin-path /path/to/my-plugin.so
```
## 获取 `CLICE_PLUGIN_DEF_HASH` 的内容
`clice_get_server_plugin_info` 函数中需要返回两个值。
- `CLICE_PLUGIN_API_VERSION` 用于确保插件和服务器之间的 `clice_get_server_plugin_info` 函数的一致性。
- `CLICE_PLUGIN_DEF_HASH` 用于确保插件和服务器之间的 C++ 声明的一致性。
要调试 `CLICE_PLUGIN_DEF_HASH` 的内容,你可以运行以下命令:
```shell
$ git clone https://github.com/clice-io/clice.git
$ cd clice
$ git checkout `clice --version --git-describe`
$ python scripts/plugin-def.py content > /tmp/plugin-proto.h
```
你将会得到一个 C 源码格式的文件,内容大致如下:
```cpp
#if 0
// begin of config/llvm-manifest.json
[
{
"version": "21.1.4+r1",
"filename": "arm64-macos-clang-debug-asan.tar.xz",
"sha256": "7da4b7d63edefecaf11773e7e701c575140d1a07329bbbb038673b6ee4516ff5",
"lto": false,
"asan": true,
"platform": "macosx",
"build_type": "Debug"
},
...
]
...
#endif
```
## Builtin libraries
如果你希望把插件直接编译进 `clice` 可执行文件,而不是在运行时动态加载,请参考 [Builtin Libraries](./builtin-library.md)。

View File

@@ -18,6 +18,13 @@ $ ./build/bin/unit_tests --test-dir="./tests/data"
$ pytest -s --log-cli-level=INFO tests/integration --executable=./build/bin/clice
```
如果你使用 xmake 作为构建系统,可以直接通过 xmake 运行测试:
```shell
$ xmake run --verbose unit_tests
$ xmake test --verbose integration_tests/default
```
## Debug
如果想在 clice 上附加调试器并进行调试,推荐先单独以 socket 模式启动 clice然后再将客户端连接到 clice 上

View File

@@ -54,73 +54,14 @@ bazel run @hedron_compile_commands//:refresh_all
### Visual Studio
Visual Studio2019 16.1+)可以通过 CMake 集成来生成编译数据库。将项目作为 CMake 项目打开,然后在 `CMakeSettings.json` 中配置:
```json
{
"configurations": [
{
"name": "x64-Debug",
"generator": "Ninja",
"buildRoot": "${projectDir}\\build",
"cmakeCommandArgs": "-DCMAKE_EXPORT_COMPILE_COMMANDS=ON"
}
]
}
```
对于基于 MSBuild 的项目(`.vcxproj`),可以使用 [compiledb-vs](https://github.com/pjbroad/compiledb-vs) 或 [catter](https://github.com/clice-io/catter) 来生成编译数据库。
TODO:
### Makefile
对于基于 Makefile 的项目,使用 [bear](https://github.com/rizsotto/Bear) 来拦截编译命令:
```bash
bear -- make
```
这会在当前目录生成 `compile_commands.json`。注意 `bear` 需要干净的构建来捕获所有命令——如果需要的话,在运行 `bear -- make` 之前先执行 `make clean`
另外,如果使用 GNU Make也可以使用 [compiledb](https://github.com/nicktimko/compiledb)
```bash
compiledb make
```
### Meson
Meson 在配置阶段会自动生成编译数据库:
```bash
meson setup build
```
`compile_commands.json` 会生成在 `build` 目录下。
TODO:
### Xmake
用下列任意方法生成编译数据库。
#### 命令行手动生成
在命令行中执行以下命令:
```bash
xmake project -k compile_commands --lsp=clangd build
```
> 通过这种方法生成的编译数据库无法自动更新,需要在项目编译配置更改时手动重新生成。
#### VSCode 扩展
Xmake 提供的官方 VSCode 扩展会在 `xmake.lua` 更新时自动生成编译数据库。然而默认情况下,它将编译数据库生成到了 `.vscode` 文件夹。在 `settings.json` 中添加以下配置:
```json
"xmake.compileCommandsDirectory": "build"
```
以将编译数据库的生成目录调整到 `build`,供 clice 使用。
### Others
对于任意其它的构建系统,可以使用 [catter](https://github.com/clice-io/catter) 来生成编译数据库。它通过伪装编译器的方式来捕获编译命令,能够可靠地与任何调用编译器可执行文件的构建系统配合工作
对于任意其它的构建系统,可以尝试使用 [bear](https://github.com/rizsotto/Bear) 或者 [scan-build](https://github.com/rizsotto/scan-build) 来拦截编译命令并获取到编译数据库(不保证成功)。我们计划在未来编写一个**新的工具**,通过假编译器的方式来实现编译命令的捕获

View File

@@ -0,0 +1,9 @@
{
// See http://go.microsoft.com/fwlink/?LinkId=827846
// for the documentation about the extensions.json format
"recommendations": [
"dbaeumer.vscode-eslint",
"amodio.tsl-problem-matcher",
"ms-vscode.extension-test-runner"
]
}

17
editors/vscode/.vscode/launch.json vendored Normal file
View File

@@ -0,0 +1,17 @@
// A launch configuration that compiles the extension and then opens it inside a new window
// Use IntelliSense to learn about possible attributes.
// Hover to view descriptions of existing attributes.
// For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
{
"version": "0.2.0",
"configurations": [
{
"name": "Run Extension",
"type": "extensionHost",
"request": "launch",
"args": ["--extensionDevelopmentPath=${workspaceFolder}"],
"outFiles": ["${workspaceFolder}/dist/**/*.js"],
"preLaunchTask": "${defaultBuildTask}"
}
]
}

13
editors/vscode/.vscode/settings.json vendored Normal file
View File

@@ -0,0 +1,13 @@
// Place your settings in this file to overwrite default and user settings.
{
"files.exclude": {
"out": false, // set this to true to hide the "out" folder with the compiled JS files
"dist": false // set this to true to hide the "dist" folder with the compiled JS files
},
"search.exclude": {
"out": true, // set this to false to include "out" folder in search results
"dist": true // set this to false to include "dist" folder in search results
},
// Turn off tsc task auto detection since we have the necessary tasks as npm scripts
"typescript.tsc.autoDetect": "off"
}

37
editors/vscode/.vscode/tasks.json vendored Normal file
View File

@@ -0,0 +1,37 @@
// See https://go.microsoft.com/fwlink/?LinkId=733558
// for the documentation about the tasks.json format
{
"version": "2.0.0",
"tasks": [
{
"type": "npm",
"script": "watch",
"problemMatcher": "$ts-webpack-watch",
"isBackground": true,
"presentation": {
"reveal": "never",
"group": "watchers"
},
"group": {
"kind": "build",
"isDefault": true
}
},
{
"type": "npm",
"script": "watch-tests",
"problemMatcher": "$tsc-watch",
"isBackground": true,
"presentation": {
"reveal": "never",
"group": "watchers"
},
"group": "build"
},
{
"label": "tasks: watch-tests",
"dependsOn": ["npm: watch", "npm: watch-tests"],
"problemMatcher": []
}
]
}

View File

@@ -62,12 +62,7 @@
}
}
},
"commands": [
{
"command": "clice.restart",
"title": "Clice: Restart Language Server"
}
],
"commands": [],
"semanticTokenTypes": [
{
"id": "character",

View File

@@ -10,7 +10,7 @@ interface Setting {
export function getSetting(): Setting | undefined {
const setting = vscode.workspace.getConfiguration("clice");
const executable = setting.get<string>("executable");
const mode = process.env.CLICE_MODE || setting.get<string>("mode");
const mode = setting.get<string>("mode");
if (mode !== "pipe" && mode !== "socket") {
vscode.window.showErrorMessage(`Unexpected mode: ${mode}`);

View File

@@ -1,8 +1,7 @@
#pragma once
#include <concepts>
#include "compile/compilation_unit.h"
#include "AST/SourceCode.h"
#include "Compiler/CompilationUnit.h"
#include "clang/AST/RecursiveASTVisitor.h"

View File

@@ -0,0 +1,63 @@
#pragma once
#include "Support/Enum.h"
namespace clice {
struct RelationKind : refl::Enum<RelationKind, true, uint32_t> {
enum Kind : uint32_t {
Invalid,
Declaration,
Definition,
Reference,
WeakReference,
// Write Relation.
Read,
Write,
Interface,
Implementation,
/// When target is a type definition of source, source is possible type or constructor.
TypeDefinition,
/// When target is a base class of source.
Base,
/// When target is a derived class of source.
Derived,
/// When target is a constructor of source.
Constructor,
/// When target is a destructor of source.
Destructor,
// When target is a caller of source.
Caller,
// When target is a callee of source.
Callee,
};
using Enum::Enum;
constexpr bool isDeclOrDef() {
return is_one_of(Declaration, Definition);
}
constexpr bool isReference() {
return is_one_of(Reference, WeakReference);
}
constexpr bool isBetweenSymbol() {
return is_one_of(Interface,
Implementation,
TypeDefinition,
Base,
Derived,
Constructor,
Destructor);
}
constexpr bool isCall() {
return is_one_of(Caller, Callee);
}
};
} // namespace clice

View File

@@ -17,12 +17,9 @@ namespace clice {
/// completion, you cannot get go-to-definition, etc. To avoid this, we just use
/// some heuristics to simplify the dependent names as normal type/expression.
/// For example, `std::vector<T>::value_type` can be simplified as `T`.
///
/// Thread safety: NOT thread-safe. Each compilation unit should have its own resolver.
/// The `resolved` cache persists across multiple resolve() calls on the same unit.
class TemplateResolver {
public:
explicit TemplateResolver(clang::Sema& sema) : sema(sema) {}
TemplateResolver(clang::Sema& sema) : sema(sema) {}
clang::QualType resolve(clang::QualType type);
@@ -30,7 +27,7 @@ public:
void resolve(clang::UnresolvedLookupExpr* expr);
// TODO: Use a clearer approach for resolving UnresolvedLookupExpr.
// TODO: use a relative clear way to resolve `UnresolvedLookupExpr`.
void resolve(clang::UnresolvedUsingType* type);
@@ -53,7 +50,7 @@ public:
if(identifier) {
return lookup(template_name.getQualifier(), identifier);
} else {
/// TODO: Operators don't have an IdentifierInfo; need DeclarationName-based lookup.
/// FIXME: Operators does't have a name.
return {};
}
}
@@ -63,7 +60,7 @@ public:
}
lookup_result lookup(const clang::UnresolvedLookupExpr* expr) {
/// TODO: Only returns the first TemplateDecl; should handle overloaded lookups.
/// FIXME:
for(auto decl: expr->decls()) {
if(auto TD = llvm::dyn_cast<clang::TemplateDecl>(decl)) {
return lookup_result(TD);
@@ -77,8 +74,8 @@ public:
return {};
}
/// TODO: Implement dependent member expression lookup (e.g. `x.template foo<T>()`).
lookup_result lookup(const clang::CXXDependentScopeMemberExpr* expr) {
/// TODO:
lookup_result lookup(clang::CXXDependentScopeMemberExpr* expr) {
return {};
}
@@ -86,19 +83,16 @@ public:
return lookup(decl->getQualifier(), decl->getDeclName());
}
lookup_result lookup(const clang::UnresolvedUsingTypenameDecl* decl) {
lookup_result resolve(const clang::UnresolvedUsingTypenameDecl* decl) {
return lookup(decl->getQualifier(), decl->getDeclName());
}
#ifndef NDEBUG
static inline bool debug = false;
#endif
private:
clang::Sema& sema;
/// Cache of resolved dependent types, keyed by AST node pointer.
/// Shared across resolve() calls within the same TU for performance.
/// This is safe because a given AST node (DependentNameType*, etc.) has a
/// unique identity within the TU — the same pointer always refers to the same
/// syntactic occurrence. Different syntactic occurrences of the "same" type
/// have different AST node pointers.
llvm::DenseMap<const void*, clang::QualType> resolved;
};

View File

@@ -2,7 +2,7 @@
#include <stack>
#include "syntax/token.h"
#include "SourceCode.h"
#include "llvm/ADT/SmallVector.h"
#include "clang/AST/ASTTypeTraits.h"

View File

@@ -1,10 +1,10 @@
#pragma once
#include "semantic/ast_utility.h"
#include "semantic/filtered_ast_visitor.h"
#include "semantic/relation_kind.h"
#include "semantic/resolver.h"
#include "semantic/symbol_kind.h"
#include "FilterASTVisitor.h"
#include "RelationKind.h"
#include "Resolver.h"
#include "SymbolKind.h"
#include "Utility.h"
namespace clice {

193
include/AST/SourceCode.h Normal file
View File

@@ -0,0 +1,193 @@
#pragma once
#include <tuple>
#include "clang/Basic/SourceLocation.h"
#include "clang/Lex/Token.h"
namespace std {
template <>
struct tuple_size<clang::SourceRange> : std::integral_constant<std::size_t, 2> {};
template <>
struct tuple_element<0, clang::SourceRange> {
using type = clang::SourceLocation;
};
template <>
struct tuple_element<1, clang::SourceRange> {
using type = clang::SourceLocation;
};
} // namespace std
namespace clang {
/// Through ADL, make `clang::SourceRange` could be destructured.
template <std::size_t I>
clang::SourceLocation get(clang::SourceRange range) {
if constexpr(I == 0) {
return range.getBegin();
} else {
return range.getEnd();
}
}
class Lexer;
} // namespace clang
namespace clice {
struct LocalSourceRange {
/// The begin position offset to the source file.
uint32_t begin = static_cast<uint32_t>(-1);
/// The end position offset to the source file.
uint32_t end = static_cast<uint32_t>(-1);
constexpr bool operator==(const LocalSourceRange& other) const = default;
constexpr auto length() {
return end - begin;
}
constexpr bool contains(uint32_t offset) const {
return offset >= begin && offset <= end;
}
constexpr bool intersects(const LocalSourceRange& other) const {
return begin <= other.end && end >= other.begin;
}
constexpr bool valid() const {
return begin != -1 && end != -1;
}
};
using TokenKind = clang::tok::TokenKind;
struct Token {
/// Whether this token is at the start of line.
bool is_at_start_of_line = false;
/// Whether this token is a preprocessor directive.
bool is_pp_keyword = false;
/// The kind of this token.
TokenKind kind;
/// The source range of this token.
LocalSourceRange range;
bool valid() {
return range.valid();
}
llvm::StringRef name() const {
return clang::tok::getTokenName(kind);
}
llvm::StringRef text(llvm::StringRef content) const {
assert(range.valid() && "Invalid source range");
return content.substr(range.begin, range.end - range.begin);
}
bool is_eod() const {
return kind == clang::tok::eod;
}
bool is_eof() const {
return kind == clang::tok::eof;
}
bool is_identifier() const {
return kind == clang::tok::raw_identifier;
}
bool is_directive_hash() const {
return is_at_start_of_line && kind == clang::tok::hash;
}
/// The tokens after the include directive are regarded as
/// a whole token, whose kind is `header_name`. For example
/// `<iostream>` and `"test.h"` are both header name.
bool is_header_name() const {
return kind == clang::tok::header_name;
}
};
class Lexer {
public:
Lexer(llvm::StringRef content,
bool ignore_comments = true,
const clang::LangOptions* lang_opts = nullptr,
bool ignore_end_of_directive = true);
Lexer(const Lexer&) = delete;
Lexer(Lexer&&) = delete;
Lexer& operator=(const Lexer&) = delete;
Lexer& operator=(Lexer&&) = delete;
~Lexer();
void lex(Token& token);
/// Get the token before this token without moving the lexer.
Token last();
/// Get the token after this token without moving the lexer.
Token next();
/// Advance the lexer and return the next token.
Token advance();
/// Advance the lexer if the next token kind is the param.
std::optional<Token> advance_if(llvm::function_ref<bool(const Token&)> callback);
std::optional<Token> advance_if(llvm::StringRef spelling) {
return advance_if([&](const Token& token) {
return token.is_identifier() && token.text(content) == spelling;
});
}
std::optional<Token> advance_if(TokenKind kind) {
return advance_if([&](const Token& token) { return token.kind == kind; });
}
/// Advance the lexer until meet the specific kind token.
Token advance_until(TokenKind kind);
private:
/// If this is set to false, the lexer will emit tok::eod at the end
/// of directive.
bool ignore_end_of_directive = true;
/// Whether we are lexing the preprocessor directive.
bool parse_pp_keyword = false;
/// Whether we are lexing the header name.
bool parse_header_name = false;
bool module_declaration_context = true;
/// The cache of last token.
Token last_token;
/// The cache of current token.
Token current_token;
/// The cache of next token.
std::optional<Token> next_token;
/// The lexed content.
llvm::StringRef content;
std::unique_ptr<clang::Lexer> lexer;
};
} // namespace clice

View File

@@ -1,9 +1,8 @@
#pragma once
#include <cstdint>
#include "Support/Enum.h"
#include "clang/AST/Decl.h"
#include "clang/Basic/TokenKinds.h"
namespace clice {
@@ -15,8 +14,8 @@ namespace clice {
/// consistently across our responses to the client and in the index. Users who prefer to stick to
/// standard LSP kinds can map our `SymbolKind` to the corresponding LSP kinds through
/// configuration.
struct SymbolKind {
enum Kind : std::uint8_t {
struct SymbolKind : refl::Enum<SymbolKind, false, uint8_t> {
enum Kind : uint8_t {
Comment = 0, ///< C/C++ comments.
Number, ///< C/C++ number literal.
Character, ///< C/C++ character literal.
@@ -51,121 +50,37 @@ struct SymbolKind {
Invalid,
};
constexpr SymbolKind() = default;
using Enum::Enum;
constexpr SymbolKind(Kind kind) : kind_value(kind) {}
constexpr explicit SymbolKind(std::uint8_t kind) : kind_value(static_cast<Kind>(kind)) {}
constexpr operator Kind() const {
return kind_value;
}
constexpr std::uint8_t value_of() const {
return static_cast<std::uint8_t>(kind_value);
}
constexpr std::uint8_t value() const {
return value_of();
}
constexpr inline static auto InvalidEnum = Kind::Invalid;
static SymbolKind from(const clang::Decl* decl);
static SymbolKind from(const clang::tok::TokenKind kind);
private:
Kind kind_value = Invalid;
};
struct SymbolModifiers {
enum Kind : std::uint32_t {
struct SymbolModifiers : refl::Enum<SymbolModifiers, true, uint32_t> {
enum Kind {
/// Represents that the symbol is a declaration(e.g. function declaration).
Declaration = 0,
/// Represents that the symbol is a definition(e.g. function definition).
Definition = 1,
Definition,
/// Represents that the symbol is const modified(e.g. `const` variable).
Const = 2,
Const,
/// Represents that the symbol is overloaded(e.g. overloaded functions and operators).
Overloaded = 3,
Overloaded,
/// Represents that the symbol is a part of type(e.g. `*` in `int*`).
Typed = 4,
Typed,
/// Represents that the symbol is a template(e.g. class template or function template).
Templated = 5,
/// Represents that the symbol is deprecated.
Deprecated = 6,
/// Represents that the symbol is deduced.
Deduced = 7,
/// Represents that the symbol is readonly.
Readonly = 8,
/// Represents that the symbol is static.
Static = 9,
/// Represents that the symbol is abstract.
Abstract = 10,
/// Represents that the symbol is virtual.
Virtual = 11,
/// Represents that the symbol is a dependent name.
DependentName = 12,
/// Represents that the symbol comes from the default library.
DefaultLibrary = 13,
/// Represents that the symbol is used through a mutable reference.
UsedAsMutableReference = 14,
/// Represents that the symbol is used through a mutable pointer.
UsedAsMutablePointer = 15,
/// Represents that the symbol is a constructor or destructor.
ConstructorOrDestructor = 16,
/// Represents that the symbol is user-defined.
UserDefined = 17,
/// Represents that the symbol is function-scoped.
FunctionScope = 18,
/// Represents that the symbol is class-scoped.
ClassScope = 19,
/// Represents that the symbol is file-scoped.
FileScope = 20,
/// Represents that the symbol is global-scoped.
GlobalScope = 21,
Templated,
};
constexpr static std::uint32_t to_mask(Kind kind) {
return std::uint32_t(1) << static_cast<std::uint32_t>(kind);
}
constexpr SymbolModifiers() = default;
constexpr SymbolModifiers(Kind kind) : value(to_mask(kind)) {}
constexpr explicit SymbolModifiers(std::uint32_t bits) : value(bits) {}
constexpr operator std::uint32_t() const {
return value;
}
constexpr bool contains(Kind kind) const {
return (value & to_mask(kind)) != 0;
}
private:
std::uint32_t value = 0;
using Enum::Enum;
};
} // namespace clice

10
include/Async/Async.h Normal file
View File

@@ -0,0 +1,10 @@
#pragma once
#include "Event.h"
#include "FileSystem.h"
#include "Gather.h"
#include "Lock.h"
#include "Network.h"
#include "Sleep.h"
#include "ThreadPool.h"
#include "libuv.h"

96
include/Async/Awaiter.h Normal file
View File

@@ -0,0 +1,96 @@
#pragma once
#include <expected>
#include "Task.h"
#include "libuv.h"
namespace clice::async::awaiter {
template <typename Request>
struct uv_base;
template <typename Request>
requires (is_uv_handle_v<Request>)
struct uv_base<Request> {
/// For libuv handles, `uv_close` must be called to release resources.
/// However, `uv_close` can only be async operation. When close is actually called,
/// the promise object may have already been destroyed, leading to undefined behavior
/// such as use-after-free. To avoid this situation, we allocate memory separately
/// for the handle and destroy it in the callback function.
Request& request;
uv_base() : request(*static_cast<Request*>(std::malloc(sizeof(Request)))) {}
~uv_base() {
uv_close(reinterpret_cast<uv_handle_t*>(&request),
[](uv_handle_t* handle) { std::free(handle); });
}
};
template <typename Request>
requires (is_uv_req_v<Request>)
struct uv_base<Request> {
/// For libuv requests, they don't need to be closed. We can lay them on the promise
/// object directly.
Request request;
};
/// The CRTP base class for the awaiter of libuv async operations. The Derived should
/// implement the `start` and `cleanup` functions.
template <typename Derived, typename Request, typename Ret, typename... Extras>
struct uv : uv_base<Request> {
int error = 0;
promise_base* continuation;
bool await_ready() const noexcept {
return false;
}
/// The callback function to handle the async operation. This should always called
/// in the main thread.
static void callback(Request* request, Extras... extras) {
auto& self = *static_cast<Derived*>(request->data);
/// The derived should implement the cleanup function to release resources or set
/// the error code if the async operation fails.
self.cleanup(extras...);
/// Then we resume the coroutine. It may destroy the current task,
/// If the task is cancelled and disposable.
self.continuation->resume();
}
template <typename Promise>
std::coroutine_handle<> await_suspend(std::coroutine_handle<Promise> waiting) noexcept {
continuation = &waiting.promise();
this->request.data = static_cast<Derived*>(this);
auto& self = *static_cast<Derived*>(this);
/// Start the async operation.
error = self.start(callback);
/// If the async operation fails, resume the coroutine immediately.
if(error < 0) {
return continuation->resume_handle();
}
/// Otherwise, return the coroutine handle to resume later.
return std::noop_coroutine();
}
std::expected<Ret, std::error_code> await_resume() noexcept {
if(error < 0) {
return std::unexpected(std::error_code(error, category()));
}
if constexpr(!std::is_void_v<Ret>) {
return static_cast<Derived*>(this)->result();
} else {
return std::expected<void, std::error_code>();
}
}
};
} // namespace clice::async::awaiter

58
include/Async/Event.h Normal file
View File

@@ -0,0 +1,58 @@
#pragma once
#include "Task.h"
#include "llvm/ADT/ArrayRef.h"
namespace clice::async {
namespace awaiter {
struct event {
bool ready;
llvm::SmallVectorImpl<promise_base*>& awaiters;
bool await_ready() const noexcept {
return ready;
}
template <typename Promise>
void await_suspend(std::coroutine_handle<Promise> handle) const noexcept {
awaiters.emplace_back(&handle.promise());
}
void await_resume() const noexcept {}
};
} // namespace awaiter
class Event {
public:
Event() = default;
void set() {
ready = true;
for(auto* awaiter: awaiters) {
awaiter->schedule();
}
}
void unset() {
ready = false;
}
void clear() {
ready = false;
awaiters.clear();
}
auto operator co_await() {
return awaiter::event{ready, awaiters};
}
private:
bool ready = false;
llvm::SmallVector<promise_base*, 4> awaiters;
};
} // namespace clice::async

View File

@@ -0,0 +1,95 @@
#pragma once
#include <chrono>
#include <cstddef>
#include "Awaiter.h"
#include "Task.h"
#include "libuv.h"
#include "Support/Enum.h"
#include "Support/JSON.h"
#include "llvm/ADT/FunctionExtras.h"
#include "llvm/ADT/StringRef.h"
namespace clice::async {
namespace fs {
struct Mode : refl::Enum<Mode, true> {
enum Kind {
/// Open the file for reading.
Read = 0,
/// Open the file for writing.
Write,
/// Open the file for reading and writing.
ReadWrite,
/// If the file does not exist, create it.
Create,
/// If the file exists, append the data to the end of the file.
Append,
/// If the file exists, truncate the file to zero length.
Truncate,
/// If the file exists, fail the open.
Exclusive,
};
using Enum::Enum;
};
struct handle {
public:
handle(uv_file file) : file(file) {}
handle(const handle&) = delete;
handle(handle&& other) noexcept : file(other.file) {
other.file = -1;
}
~handle();
handle& operator=(const handle&) = delete;
handle& operator=(handle&& other) noexcept = delete;
int value() const {
return file;
}
private:
uv_file file;
};
/// Open the file asynchronously.
Result<handle> open(std::string path, Mode mode);
/// Read the file asynchronously, make sure the buffer is valid until the task is done.
Result<ssize_t> read(const handle& handle, char* buffer, std::size_t size);
Result<std::string> read(std::string path, Mode mode = Mode::Read);
/// Write the file asynchronously, make sure the buffer is valid until the task is done.
Result<void> write(const handle& handle, char* buffer, std::size_t size);
Result<void> write(std::string path,
char* buffer,
std::size_t size,
Mode mode = Mode(Mode::Write, Mode::Create, Mode::Truncate));
struct Stats {
std::chrono::milliseconds mtime;
size_t size;
};
Result<Stats> stat(std::string path);
} // namespace fs
} // namespace clice::async

143
include/Async/Gather.h Normal file
View File

@@ -0,0 +1,143 @@
#pragma once
#include <thread>
#include <tuple>
#include "Event.h"
#include "Task.h"
namespace clice::async {
struct none {};
template <typename Task, typename V = typename std::remove_cvref_t<Task>::value_type>
using task_value_t = std::conditional_t<std::is_void_v<V>, none, V>;
template <typename... Tasks>
auto gather(Tasks&&... tasks) -> Task<std::tuple<task_value_t<Tasks>...>> {
constexpr static std::size_t count = sizeof...(Tasks);
Event event;
std::size_t finished = 0;
auto run_task = [&](auto& task) -> Task<task_value_t<decltype(task)>> {
using V = typename std::remove_cvref_t<decltype(task)>::value_type;
if constexpr(std::is_void_v<V>) {
co_await task;
/// Check if all tasks are finished. If so, set the event to
/// resume the gather handle.
finished += 1;
if(finished == count) {
event.set();
}
co_return none{};
} else {
auto result = co_await task;
finished += 1;
if(finished == count) {
event.set();
}
co_return std::move(result);
}
};
auto schedule_task = [&](auto& task) {
auto core = run_task(task);
core.schedule();
return core;
};
std::tuple all = {schedule_task(tasks)...};
/// Wait for all tasks to finish.
co_await event;
/// Return the results of all tasks.
co_return [&]<std::size_t... Is>(std::index_sequence<Is...>) {
return std::make_tuple(std::get<Is>(all).result()...);
}(std::make_index_sequence<count>{});
}
/// Run the tasks in parallel and return the results.
template <typename... Tasks>
auto run(Tasks&&... tasks) {
auto core = gather(std::forward<Tasks>(tasks)...);
core.schedule();
async::run();
assert(core.done() && "run: not done");
return core.result();
}
template <ranges::input_range Range, typename Coroutine>
requires requires(Coroutine coroutine, ranges::range_value_t<Range> value) {
{ coroutine(value) } -> std::same_as<Task<bool>>;
}
Task<bool> gather(Range&& range,
Coroutine&& coroutine,
std::size_t concurrency = std::thread::hardware_concurrency()) {
std::vector<Task<>> tasks;
tasks.reserve(concurrency);
auto iter = ranges::begin(range);
auto end = ranges::end(range);
Event event;
std::size_t finished = 0;
bool cancelled = false;
auto run_task = [&](auto& value) -> async::Task<> {
/// Execute the first task.
auto task = coroutine(value);
/// If any task fails, cancel all tasks and return false.
if(auto result = co_await task; !result) {
for(auto& task: tasks) {
task.cancel();
task.dispose();
}
cancelled = true;
event.set();
co_return;
}
finished += 1;
/// Check if still have tasks to run. If so, run the next task.
while(iter != end) {
auto task = coroutine(*iter);
iter++;
finished -= 1;
if(auto result = co_await task; !result) {
for(auto& task: tasks) {
task.cancel();
task.dispose();
}
cancelled = true;
event.set();
co_return;
}
finished += 1;
}
/// Check if all tasks are finished. If so, set the event to
/// resume the gather handle.
if(finished == tasks.size()) {
event.set();
}
};
/// Fill tasks.
while(iter != end && tasks.size() < concurrency) {
tasks.emplace_back(run_task(*iter));
tasks.back().schedule();
iter++;
}
co_await event;
co_return !cancelled;
}
} // namespace clice::async

79
include/Async/Lock.h Normal file
View File

@@ -0,0 +1,79 @@
#pragma once
#include "Event.h"
namespace clice::async {
namespace awaiter {
struct lock {
llvm::SmallVectorImpl<promise_base*>& awaiters;
bool await_ready() const noexcept {
return false;
}
template <typename Promise>
void await_suspend(std::coroutine_handle<Promise> handle) const noexcept {
awaiters.emplace_back(&handle.promise());
}
void await_resume() const noexcept {}
};
} // namespace awaiter
class Lock {
friend class guard;
public:
Lock() = default;
class Guard {
public:
Guard(Lock* lock) : lock(lock) {}
Guard(Guard&& other) : lock(other.lock) {
other.lock = nullptr;
}
~Guard() {
if(!lock) {
return;
}
lock->locked = false;
if(!lock->awaiters.empty()) {
lock->awaiters.front()->schedule();
lock->awaiters.erase(lock->awaiters.begin());
}
}
private:
Lock* lock;
};
/// Try to get the lock. If the lock is locked, the current coroutine will be
/// suspended and wait for the lock to be released.
Task<Guard> try_lock() {
/// Note that this task also may be canceled, we make sure
/// even cancel, it can resume one task(through destructor).
Guard guard(this);
if(locked) {
co_await awaiter::lock{awaiters};
}
locked = true;
/// Use `std::move` to make sure it will not resume
/// the awaiter here.
co_return std::move(guard);
}
private:
bool locked = false;
llvm::SmallVector<promise_base*, 4> awaiters;
};
} // namespace clice::async

23
include/Async/Network.h Normal file
View File

@@ -0,0 +1,23 @@
#pragma once
#include "Task.h"
#include "libuv.h"
#include "Support/JSON.h"
#include "llvm/ADT/FunctionExtras.h"
#include "llvm/ADT/StringRef.h"
namespace clice::async::net {
using Callback = llvm::unique_function<Task<void>(json::Value)>;
/// Listen on stdin/stdout, callback is called when there is a LSP message available.
void listen(Callback callback);
/// Listen on the given host and port, callback is called when there is a LSP message available.
void listen(const char* host, unsigned int port, Callback callback);
/// Write a JSON value to the client.
Task<> write(json::Value value);
} // namespace clice::async::net

37
include/Async/Sleep.h Normal file
View File

@@ -0,0 +1,37 @@
#pragma once
#include <chrono>
#include "Awaiter.h"
namespace clice::async {
namespace awaiter {
struct sleep : uv<sleep, uv_timer_t, void> {
std::chrono::milliseconds duration;
int start(auto callback) {
int err = uv_timer_init(async::loop, &request);
if(err < 0) {
return err;
}
return uv_timer_start(&request, callback, duration.count(), 0);
}
void cleanup() {
error = uv_timer_stop(&request);
}
};
} // namespace awaiter
inline auto sleep(std::chrono::milliseconds duration) {
return awaiter::sleep{{}, duration};
}
inline auto sleep(std::size_t milliseconds) {
return sleep(std::chrono::milliseconds(milliseconds));
}
}; // namespace clice::async

334
include/Async/Task.h Normal file
View File

@@ -0,0 +1,334 @@
#pragma once
#include <cassert>
#include <coroutine>
#include <cstdint>
#include <cstdlib>
#include <optional>
#include <source_location>
#include "Support/Format.h"
namespace clice::async {
template <typename T>
class Task;
struct promise_base {
enum Flags : uint8_t {
Empty = 0,
/// The task is cancelled.
Cancelled = 1,
/// The coroutine handle will be destroyed when the task is done or cancelled.
Disposable = 1 << 1,
/// The coroutine is done or is cancelled and resumed, means it will never
/// scheduled again.
Finished = 1 << 2,
};
uint8_t flags;
void* data;
/// The coroutine handle that is waiting for the task to complete.
/// If this is a top-level coroutine, it is empty.
promise_base* continuation = nullptr;
promise_base* next = nullptr;
std::source_location location;
template <typename Promise>
void set(std::coroutine_handle<Promise> handle) {
flags = Empty;
data = handle.address();
}
auto handle() const noexcept {
return std::coroutine_handle<>::from_address(data);
}
void schedule();
bool done() const noexcept {
return handle().done();
}
void destroy() {
handle().destroy();
}
void cancel() {
auto p = this;
while(p) {
p->flags |= Flags::Cancelled;
p = p->next;
}
}
bool cancelled() const noexcept {
return flags & Flags::Cancelled;
}
void dispose() {
flags |= Flags::Disposable;
}
bool disposable() const noexcept {
return flags & Flags::Disposable;
}
void finish() {
flags |= Flags::Finished;
}
bool finished() {
return flags & Flags::Finished;
}
std::coroutine_handle<> resume_handle() {
if(cancelled()) {
/// If the task is cancelled and disposable, destroy the coroutine handle.
auto p = this;
while(p && p->cancelled()) {
auto con = p->continuation;
if(p->disposable()) {
p->destroy();
} else {
p->finish();
}
p = con;
}
return std::noop_coroutine();
} else {
/// Otherwise, resume the coroutine handle.
return handle();
}
}
void resume() {
resume_handle().resume();
}
};
namespace awaiter {
/// The awaiter for the final suspend point of `Task`.
struct final {
promise_base* continuation;
bool await_ready() noexcept {
return false;
}
template <typename Promise>
std::coroutine_handle<> await_suspend(std::coroutine_handle<Promise> current) noexcept {
std::coroutine_handle<> handle = std::noop_coroutine();
/// In the final suspend point, this coroutine is already done.
/// So try to resume the waiting coroutine if it exists.
if(continuation) {
continuation->next = nullptr;
handle = continuation->resume_handle();
}
/// Mark current coroutine as finished.
current.promise().finish();
if(current.promise().disposable()) {
/// If this task is disposable, destroy the coroutine handle.
current.destroy();
}
return handle;
}
void await_resume() noexcept {}
};
/// The awaiter for the `Task` type.
template <typename T, typename P>
struct task {
std::coroutine_handle<P> handle;
bool await_ready() noexcept {
return false;
}
template <typename Promise>
auto await_suspend(std::coroutine_handle<Promise> waiting) noexcept {
/// Store the waiting coroutine in the promise for later scheduling.
/// It will be scheduled in the final suspend point.
assert(!handle.promise().continuation && "await_suspend: already waiting");
handle.promise().continuation = &waiting.promise();
waiting.promise().next = &handle.promise();
/// If this `Task` is awaited from another coroutine, we should schedule
/// the this task first.
return handle.promise().resume_handle();
}
T await_resume() noexcept {
if constexpr(!std::is_void_v<T>) {
assert(handle.promise().value.has_value() && "await_resume: value not set");
return std::move(*handle.promise().value);
}
}
};
} // namespace awaiter
template <typename T = void>
class Task {
public:
template <typename V>
struct promise_result {
std::optional<V> value;
template <typename U>
void return_value(U&& val) noexcept {
assert(!value.has_value() && "return_value: value already set");
value.emplace(std::forward<U>(val));
}
};
// WORKAROUND: GCC bug - full specialization in non-namespace scope not supported
// see: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=85282
template <std::same_as<void> V>
struct promise_result<V> {
void return_void() noexcept {}
};
struct promise_type : promise_base, promise_result<T> {
promise_type(std::source_location location = std::source_location::current()) {
set(handle());
this->location = location;
};
auto get_return_object() {
return Task<T>(handle());
}
auto initial_suspend() {
return std::suspend_always();
}
auto final_suspend() noexcept {
return awaiter::final{continuation};
}
void unhandled_exception() {
std::abort();
}
auto handle() {
return std::coroutine_handle<promise_type>::from_promise(*this);
}
};
using coroutine_handle = std::coroutine_handle<promise_type>;
using value_type = T;
public:
Task() = default;
Task(coroutine_handle handle) : core(handle) {}
Task(const Task&) = delete;
Task(Task&& other) noexcept : core(other.core) {
other.core = nullptr;
}
Task& operator=(const Task&) = delete;
Task& operator=(Task&& other) noexcept {
if(core) {
core.destroy();
}
core = other.core;
other.core = nullptr;
return *this;
}
~Task() {
if(core) {
core.destroy();
}
}
public:
coroutine_handle handle() const noexcept {
return core;
}
coroutine_handle release() noexcept {
auto handle = core;
core = nullptr;
return handle;
}
bool empty() const noexcept {
return !core;
}
bool done() const noexcept {
return core.done();
}
void schedule() {
core.promise().schedule();
}
/// Cancel the task, the suspend point after the current one will be skipped.
void cancel() {
core.promise().cancel();
}
bool cancelled() {
return core.promise().cancelled();
}
/// Dispose the task, it will be destroyed when finished or cancelled.
void dispose() {
core.promise().dispose();
core = nullptr;
}
bool finished() {
return core.promise().finished();
}
T result() {
if constexpr(!std::is_void_v<T>) {
return std::move(core.promise().value.value());
}
}
auto operator co_await() const noexcept {
return awaiter::task<T, promise_type>{core};
}
void stacktrace() {
promise_base* handle = core;
while(handle) {
std::println("{}:{}:{}",
handle->location.file_name(),
handle->location.line(),
handle->location.function_name());
handle = handle->continuation;
}
}
private:
coroutine_handle core;
};
} // namespace clice::async

View File

@@ -0,0 +1,60 @@
#pragma once
#include "Awaiter.h"
namespace clice::async {
namespace awaiter {
template <typename Ret>
struct value {
std::optional<Ret> value;
};
template <>
struct value<void> {};
template <typename Work, typename Ret>
struct thread_pool : value<Ret>, uv<thread_pool<Work, Ret>, uv_work_t, Ret, int> {
Work work;
/// `uv_work_t` has two callback functions, `work_cb` is executed in the thread pool,
/// and `after_work_cb` is executed in the main thread.
static void work_cb(uv_work_t* work) {
auto& awaiter = uv_cast<thread_pool>(work);
if constexpr(!std::is_void_v<Ret>) {
awaiter.value.emplace(awaiter.work());
} else {
awaiter.work();
}
}
int start(auto callback) {
return uv_queue_work(async::loop, &this->request, work_cb, callback);
}
void cleanup(int status) {
this->error = status;
}
Ret result() {
if constexpr(!std::is_void_v<Ret>) {
return std::move(*this->value);
}
}
};
} // namespace awaiter
template <typename Work, typename Ret = std::invoke_result_t<Work>>
async::Task<Ret> submit(Work&& work) {
using W = std::remove_cvref_t<Work>;
auto result = co_await awaiter::thread_pool<W, Ret>{{}, {}, std::forward<Work>(work)};
if(!result) {
/// Thread pool task should never fails.
std::abort();
}
co_return std::move(*result);
}
} // namespace clice::async

79
include/Async/libuv.h Normal file
View File

@@ -0,0 +1,79 @@
#pragma once
#ifdef _WIN32
#define NOMINMAX
#endif
#include "uv.h"
#ifdef _WIN32
#undef THIS
#endif
#include <cassert>
#include <expected>
#include <system_error>
#include <type_traits>
#include "Support/Logging.h"
#include "Support/TypeTraits.h"
namespace clice::async {
/// The default event loop.
extern uv_loop_t* loop;
template <typename T, typename U>
T& uv_cast(U* u) {
assert(u && u->data && "uv_cast: invalid uv handle");
return *static_cast<std::remove_cvref_t<T>*>(u->data);
}
#define UV_TYPE_ITER(_, name) || std::is_same_v<T, uv_##name##_t>
/// Check if the type `T` is a libuv handle.
template <typename T>
constexpr bool is_uv_handle_v = false UV_HANDLE_TYPE_MAP(UV_TYPE_ITER);
/// Check if the type `T` is a libuv request.
template <typename T>
constexpr bool is_uv_req_v = false UV_REQ_TYPE_MAP(UV_TYPE_ITER);
template <typename T>
constexpr bool is_uv_stream_v = std::is_same_v<T, uv_stream_t> || std::is_same_v<T, uv_tcp_t> ||
std::is_same_v<T, uv_pipe_t> || std::is_same_v<T, uv_tty_t>;
#undef UV_TYPE_ITER
template <typename T, typename U>
T* uv_cast(U& u) {
if constexpr(std::is_same_v<T, uv_handle_t>) {
static_assert(is_uv_handle_v<std::remove_cvref_t<U>>, "uv_cast: invalid uv handle");
} else if constexpr(std::is_same_v<T, uv_req_t>) {
static_assert(is_uv_req_v<std::remove_cvref_t<U>>, "uv_cast: invalid uv request");
} else if constexpr(std::is_same_v<T, uv_stream_t>) {
static_assert(is_uv_stream_v<std::remove_cvref_t<U>>, "uv_cast: invalid uv stream");
} else {
static_assert(dependent_false<U>, "uv_cast: invalid type");
}
return reinterpret_cast<T*>(&u);
}
void uv_check_result(const int result,
const std::source_location location = std::source_location::current());
template <typename T>
class Task;
template <typename T>
using Result = Task<std::expected<T, std::error_code>>;
const std::error_category& category();
void init();
void run();
void stop();
} // namespace clice::async

125
include/Compiler/Command.h Normal file
View File

@@ -0,0 +1,125 @@
#pragma once
#include <expected>
#include "Toolchain.h"
#include "Support/Enum.h"
#include "Support/Format.h"
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/DenseSet.h"
#include "llvm/ADT/StringMap.h"
#include "llvm/Support/Allocator.h"
namespace clice {
struct CommandOptions {
/// Ignore unknown commands arguments.
bool ignore_unknown = true;
/// Inject resource directory to the command.
bool resource_dir = false;
/// Query the compiler driver for additional information, such as system includes and target.
bool query_toolchain = false;
/// Suppress the warning log if failed to query driver info.
/// Set true in unittests to avoid cluttering test output.
bool suppress_logging = false;
/// The commands that you want to remove from original commands list.
llvm::ArrayRef<std::string> remove;
/// The commands that you want to add to original commands list.
llvm::ArrayRef<std::string> append;
};
enum class UpdateKind : std::uint8_t {
Unchanged,
Inserted,
Deleted,
};
struct UpdateInfo {
/// The kind of update.
UpdateKind kind;
/// The updated file.
std::uint32_t path_id;
/// The compilation context of this file command, which could
/// be used to identity the same file with different compilation
/// contexts.
const void* context;
};
struct CompilationContext {
/// The working directory of compilation.
llvm::StringRef directory;
/// The compilation arguments.
std::vector<const char*> arguments;
};
std::string print_argv(llvm::ArrayRef<const char*> args);
class CompilationDatabase {
public:
CompilationDatabase();
CompilationDatabase(const CompilationDatabase&) = delete;
CompilationDatabase(CompilationDatabase&& other);
CompilationDatabase& operator=(const CompilationDatabase&) = delete;
CompilationDatabase& operator=(CompilationDatabase&& other);
~CompilationDatabase();
public:
/// Read the compilation database on the give file and return the
/// incremental update infos.
std::vector<UpdateInfo> load_compile_database(llvm::StringRef file);
/// Lookup the compilation context of specific file. If the context
/// param is provided, we will return the compilation context corresponding
/// to the handle. Otherwise we just return the first one(if the file have)
/// multiple compilation contexts.
CompilationContext lookup(llvm::StringRef file,
const CommandOptions& options = {},
const void* context = nullptr);
/// TODO: list all compilation context of the file, this is useful to show
/// all contexts and let user choose one.
/// std::vector<CompilationContext> fetch_all(llvm::StringRef file);
/// Get an the option for specific argument.
static std::optional<std::uint32_t> get_option_id(llvm::StringRef argument);
/// FIXME: bad interface design ...
std::vector<llvm::StringRef> files();
/// FIXME: remove this api?
auto save_string(llvm::StringRef string) -> llvm::StringRef;
#ifdef CLICE_ENABLE_TEST
void add_command(llvm::StringRef directory,
llvm::StringRef file,
llvm::ArrayRef<const char*> arguments);
void add_command(llvm::StringRef directory, llvm::StringRef file, llvm::StringRef command);
/// FIXME: remove this
/// Update commands from json file and return all updated file.
std::expected<std::vector<UpdateInfo>, std::string> load_commands(llvm::StringRef json_content,
llvm::StringRef workspace);
#endif
private:
struct Impl;
std::unique_ptr<Impl> self;
};
} // namespace clice

View File

@@ -1,19 +1,9 @@
#pragma once
#include <atomic>
#include <cassert>
#include <cstdint>
#include <memory>
#include <string>
#include <tuple>
#include <utility>
#include <vector>
#include "compile/compilation_unit.h"
#include "support/filesystem.h"
#include "llvm/ADT/StringMap.h"
#include "llvm/ADT/StringRef.h"
#include "CompilationUnit.h"
#include "Module.h"
#include "Preamble.h"
#include "Support/FileSystem.h"
namespace clang {
@@ -23,46 +13,6 @@ class CodeCompleteConsumer;
namespace clice {
struct PCHInfo {
/// The path of the output PCH file.
std::string path;
/// The building time of this PCH.
std::int64_t mtime;
/// The content used to build this PCH.
std::string preamble;
/// All files involved in building this PCH.
std::vector<std::string> deps;
/// The command arguments used to build this PCH.
std::vector<const char*> arguments;
};
struct ModuleInfo {
/// Whether this module is an interface unit.
/// i.e. has export module declaration.
bool isInterfaceUnit = false;
/// Module name.
std::string name;
/// Dependent modules of this module.
std::vector<std::string> mods;
};
struct PCMInfo : ModuleInfo {
/// PCM file path.
std::string path;
/// Source file path.
std::string srcPath;
/// Files involved in building this PCM(not include module).
std::vector<std::string> deps;
};
struct CompilationParams {
/// The kind of this compilation.
CompilationKind kind;
@@ -75,13 +25,15 @@ struct CompilationParams {
std::string directory;
bool arguments_from_database = false;
/// Responsible for storing the arguments.
std::vector<const char*> arguments;
llvm::IntrusiveRefCntPtr<vfs::FileSystem> vfs = new ThreadSafeFS();
/// Information about reuse PCH.
std::pair<std::string, std::uint32_t> pch;
std::pair<std::string, uint32_t> pch;
/// Information about reuse PCM(name, path).
llvm::StringMap<std::string> pcms;

View File

@@ -1,33 +1,17 @@
#pragma once
#include <chrono>
#include <cstdint>
#include <string>
#include <utility>
#include <vector>
#include "compile/diagnostic.h"
#include "compile/directive.h"
#include "semantic/resolver.h"
#include "syntax/token.h"
#include "Directive.h"
#include "AST/Resolver.h"
#include "AST/SourceCode.h"
#include "AST/SymbolID.h"
#include "Compiler/Diagnostic.h"
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/DenseSet.h"
#include "clang/Tooling/Syntax/Tokens.h"
namespace clice {
namespace index {
// Temporary in-header definition during migration.
struct SymbolID {
std::uint64_t hash;
std::string name;
};
} // namespace index
enum class CompilationKind : std::uint8_t {
/// From preprocessing the source file. Therefore directives
/// are available but AST nodes are not.

View File

@@ -1,12 +1,9 @@
#pragma once
#include <cstdint>
#include <optional>
#include <string>
#include "syntax/token.h"
#include "llvm/ADT/StringRef.h"
#include "AST/SourceCode.h"
namespace clice {

View File

@@ -1,10 +1,6 @@
#pragma once
#include <cstdint>
#include <string>
#include <vector>
#include "syntax/token.h"
#include "AST/SourceCode.h"
#include "llvm/ADT/DenseMap.h"
#include "clang/Lex/MacroInfo.h"
@@ -20,8 +16,11 @@ struct Include {
/// The file id of included file.
clang::FileID fid;
/// Location of the `include` keyword.
/// Location of the `include`.
clang::SourceLocation location;
/// The range of filename(includes `""` or `<>`).
clang::SourceRange filename_range;
};
/// Information about `__has_include` directive.
@@ -37,7 +36,7 @@ struct HasInclude {
/// Information about `#if`, `#ifdef`, `#ifndef`, `#elif`,
/// `#elifdef`, `#else`, `#endif` directive.
struct Condition {
enum class BranchKind : std::uint8_t {
enum class BranchKind : uint8_t {
If = 0,
Elif,
Ifdef,
@@ -50,7 +49,7 @@ struct Condition {
using enum BranchKind;
enum class ConditionValue : std::uint8_t {
enum class ConditionValue : uint8_t {
True = 0,
False,
Skipped,
@@ -74,7 +73,7 @@ struct Condition {
/// Information about macro definition, reference and undef.
struct MacroRef {
enum class Kind : std::uint8_t {
enum class Kind : uint8_t {
Def = 0,
Ref,
Undef,
@@ -94,7 +93,7 @@ struct MacroRef {
/// Information about `#pragma` directive.
struct Pragma {
enum class Kind : std::uint8_t {
enum class Kind : uint8_t {
Region,
EndRegion,
@@ -129,39 +128,6 @@ struct Import {
std::vector<clang::SourceLocation> name_locations;
};
/// Information about `#embed` directive.
struct Embed {
/// The file name in the embed directive, not including quotes or angle brackets.
llvm::StringRef file_name;
/// The actual file that may be embedded by this embed directive.
clang::OptionalFileEntryRef file;
/// Whether the file name is angled.
bool is_angled;
/// Location of the `#` token.
clang::SourceLocation loc;
/// TODO: Currently we do not store parameters of the embed directive.
/// See clang::LexEmbedParametersResult for details.
};
/// Information about `__has_embed` directive.
struct HasEmbed {
/// The file name in the embed directive, not including quotes or angle brackets.
llvm::StringRef file_name;
/// The actual file that may be embedded by this embed directive.
clang::OptionalFileEntryRef file;
/// Whether the file name is angled.
bool is_angled;
/// Location of the `__has_embed` token.
clang::SourceLocation loc;
};
struct Directive {
std::vector<Include> includes;
std::vector<HasInclude> has_includes;
@@ -169,8 +135,6 @@ struct Directive {
std::vector<MacroRef> macros;
std::vector<Pragma> pragmas;
std::vector<Import> imports;
std::vector<Embed> embeds;
std::vector<HasEmbed> has_embeds;
};
} // namespace clice

46
include/Compiler/Module.h Normal file
View File

@@ -0,0 +1,46 @@
#pragma once
#include <expected>
#include <string>
#include <vector>
#include "Support/Struct.h"
namespace clice {
class CompilationUnit;
struct CompilationParams;
struct ModuleInfo {
/// Whether this module is an interface unit.
/// i.e. has export module declaration.
bool isInterfaceUnit = false;
/// Module name.
std::string name;
/// Dependent modules of this module.
std::vector<std::string> mods;
};
inherited_struct(PCMInfo, ModuleInfo) {
/// PCM file path.
std::string path;
/// Source file path.
std::string srcPath;
/// Files involved in building this PCM(not include module).
std::vector<std::string> deps;
};
/// If input file is module interface unit, return its module name.
/// Otherwise, return an empty string.
std::string scanModuleName(CompilationParams& params);
/// Run the preprocessor to scan the given module unit to
/// collect its module name and dependencies.
std::expected<ModuleInfo, std::string> scanModule(CompilationParams& params);
} // namespace clice

View File

@@ -0,0 +1,40 @@
#pragma once
#include <expected>
#include <string>
#include <vector>
#include "llvm/ADT/StringRef.h"
namespace clice {
class CompilationUnit;
struct CompilationParams;
struct PCHInfo {
/// The path of the output PCH file.
std::string path;
/// The building time of this PCH.
std::int64_t mtime;
/// The content used to build this PCH.
std::string preamble;
/// All files involved in building this PCH.
std::vector<std::string> deps;
/// The command arguments used to build this PCH.
std::vector<const char*> arguments;
};
/// Compute the preamble bound of given content. We just
/// run lex until we find first not directive.
std::uint32_t compute_preamble_bound(llvm::StringRef content);
/// Same as above, but return a group of bounds for chained PCH
/// building.
std::vector<uint32_t> compute_preamble_bounds(llvm::StringRef content);
} // namespace clice

34
include/Compiler/Scan.h Normal file
View File

@@ -0,0 +1,34 @@
#pragma once
#include <string>
#include <vector>
#include "AST/SourceCode.h"
#include "llvm/ADT/StringRef.h"
namespace clice {
struct Inclusion {
/// Whether this file is braced angles.
bool angled;
/// The line of this inclusion(zero based).
/// std::uint32_t line;
/// The included file.
llvm::StringRef file;
};
struct ScanResult {
/// The module file of this file(may be empty).
std::vector<Token> module_name;
/// The includes of file.
std::vector<Inclusion> includes;
};
/// Scan the file and return necessary info.
ScanResult scan(llvm::StringRef content);
} // namespace clice

View File

@@ -1,7 +1,5 @@
#pragma once
#include <vector>
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/StringRef.h"

View File

@@ -1,3 +1 @@
#pragma once
int a = 1;

View File

@@ -0,0 +1,107 @@
#pragma once
#include <cstdint>
#include <vector>
#include "AST/SourceCode.h"
#include "llvm/ADT/StringRef.h"
namespace clice {
struct CompilationParams;
namespace config {
struct CodeCompletionOption {
/// Insert placeholder for keywords? function call parameters? template arguments?
bool enable_keyword_snippet = false;
/// Also apply for lambda ...
bool enable_function_arguments_snippet = false;
bool enable_template_arguments_snippet = false;
bool insert_paren_in_function_call = false;
/// TODO: Add more detailed option, see
/// https://github.com/llvm/llvm-project/issues/63565
bool bundle_overloads = true;
/// The limits of code completion, 0 is non limit.
std::uint32_t limit = 0;
};
}; // namespace config
namespace feature {
enum class CompletionItemKind {
None = 0,
Text,
Method,
Function,
Constructor,
Field,
Variable,
Class,
Interface,
Module,
Property,
Unit,
Value,
Enum,
Keyword,
Snippet,
Color,
File,
Reference,
Folder,
EnumMember,
Constant,
Struct,
Event,
Operator,
TypeParameter
};
/// Represents a single code completion item to be presented to the user.
struct CompletionItem {
/// The primary label displayed in the completion list.
std::string label;
/// Additional details, like a function signature, shown next to the label.
std::string detail;
/// A short description of the item, typically its type or namespace.
std::string description;
/// Full documentation for the item, shown on selection or hover.
std::string document;
/// The kind of item (function, class, etc.), used for an icon.
CompletionItemKind kind;
/// A score for ranking this item against others. Higher is better.
float score;
/// Whether this item is deprecated (often rendered with a strikethrough).
bool deprecated;
/// The text edit to be applied when this item is accepted.
struct Edit {
/// The new text to insert, which may be a snippet.
std::string text;
/// The source range to be replaced by the new text.
LocalSourceRange range;
} edit;
};
using CodeCompletionResult = std::vector<CompletionItem>;
std::vector<CompletionItem> code_complete(CompilationParams& params,
const config::CodeCompletionOption& option);
} // namespace feature
} // namespace clice

View File

@@ -1,3 +1 @@
#pragma once
int b = 2;

View File

@@ -0,0 +1,19 @@
#pragma once
#include "Compiler/Diagnostic.h"
#include "Server/Convert.h"
#include "Support/JSON.h"
namespace clice {
class CompilationUnitRef;
}
namespace clice::feature {
/// FIXME: This is not correct way, we don't want to couple
/// `Feature with Protocol`? Return an array of LSP diagnostic.
json::Value diagnostics(PositionEncodingKind kind, PathMapping mapping, CompilationUnitRef unit);
} // namespace clice::feature

View File

@@ -0,0 +1,3 @@
#pragma once
namespace clice::proto {}

View File

@@ -0,0 +1,26 @@
#pragma once
#include <vector>
#include "AST/SourceCode.h"
#include "Index/Shared.h"
namespace clice::feature {
struct DocumentLink {
/// The range of the whole link.
LocalSourceRange range;
/// The target string path.
std::string file;
};
using DocumentLinks = std::vector<DocumentLink>;
/// Generate document link for main file.
DocumentLinks document_links(CompilationUnitRef unit);
/// Generate document link for all source file.
index::Shared<DocumentLinks> index_document_link(CompilationUnitRef unit);
} // namespace clice::feature

View File

@@ -0,0 +1,37 @@
#pragma once
#include "AST/SourceCode.h"
#include "AST/SymbolKind.h"
#include "Index/Shared.h"
namespace clice::feature {
struct DocumentSymbol {
/// The range of symbol name in source code.
LocalSourceRange selectionRange;
/// The range of whole symbol.
LocalSourceRange range;
/// The symbol kind of this document symbol.
SymbolKind kind;
/// The symbol name.
std::string name;
/// Extra information about this symbol.
std::string detail;
/// The symbols that this symbol contains
std::vector<DocumentSymbol> children;
};
using DocumentSymbols = std::vector<DocumentSymbol>;
/// Generate document symbols for only interested file.
DocumentSymbols document_symbols(CompilationUnitRef unit);
/// Generate document symbols for all file in unit.
index::Shared<DocumentSymbols> index_document_symbol(CompilationUnitRef unit);
} // namespace clice::feature

View File

@@ -0,0 +1,55 @@
#pragma once
#include "AST/SourceCode.h"
#include "Index/Shared.h"
#include "Support/Enum.h"
namespace clice::feature {
struct FoldingRangeKind : refl::Enum<FoldingRangeKind> {
enum Kind : uint8_t {
Invalid = 0,
Comment,
Imports,
Region,
Namespace,
Class,
Enum,
Struct,
Union,
LambdaCapture,
FunctionParams,
FunctionBody,
FunctionCall,
CompoundStmt,
AccessSpecifier,
ConditionDirective,
Initializer,
};
using Enum::Enum;
constexpr static auto InvalidEnum = Invalid;
};
/// We don't record the coalesced text for a range, because it's rarely useful.
struct FoldingRange {
/// The range to fold.
LocalSourceRange range;
/// Describes the kind of the folding range.
FoldingRangeKind kind;
/// The text to display when the folding range is collapsed.
std::string text;
};
using FoldingRanges = std::vector<FoldingRange>;
/// Generate folding range for interested file only.
FoldingRanges folding_ranges(CompilationUnitRef unit);
/// Generate folding range for all files.
index::Shared<FoldingRanges> index_folding_range(CompilationUnitRef unit);
} // namespace clice::feature

View File

@@ -0,0 +1,14 @@
#pragma once
#include "AST/SourceCode.h"
#include "Protocol/Feature/Formatting.h"
#include "llvm/ADT/StringRef.h"
namespace clice::feature {
std::vector<proto::TextEdit> document_format(llvm::StringRef file,
llvm::StringRef content,
std::optional<LocalSourceRange>);
}

74
include/Feature/Hover.h Normal file
View File

@@ -0,0 +1,74 @@
#pragma once
#include "AST/SourceCode.h"
#include "AST/SymbolKind.h"
#include "Index/Shared.h"
namespace clice::config {
struct HoverOptions {
/// Strip doxygen info and merge with lsp info
bool enable_doxygen_parsing = true;
/// If set `false`, the comment will be wrapped
/// in code block and keep ascii typesetting
bool parse_comment_as_markdown = true;
/// Show sugar type
bool show_aka = true;
};
} // namespace clice::config
namespace clice::feature {
struct HoverItem {
enum class HoverKind : uint8_t {
/// The typename of a variable or a type alias.
Type,
/// Size of type or variable.
Size,
/// Align of type or variable.
Align,
/// Offset of field in a class/struct.
Offset,
/// Bit width of a bit field.
BitWidth,
/// The index of a field in a class/struct.
FieldIndex,
/// The value of an enum item.
EnumValue,
};
using enum HoverKind;
HoverKind kind;
std::string value;
};
/// Hover information for a symbol.
struct Hover {
/// Title
SymbolKind kind;
std::string name;
/// Extra information.
std::vector<HoverItem> items;
/// Raw document in the source code.
std::string document;
/// The full qualified name of the declaration.
std::string qualifier;
/// The source code of the declaration.
std::string source;
};
/// Generate the hover information for the given declaration(for test).
Hover hover(CompilationUnitRef unit, const clang::NamedDecl* decl);
/// Generate the hover information for the symbol at the given offset.
Hover hover(CompilationUnitRef unit, std::uint32_t offset);
} // namespace clice::feature

View File

@@ -0,0 +1,57 @@
#pragma once
#include "AST/SourceCode.h"
#include "AST/SymbolID.h"
#include "Index/Shared.h"
#include "Support/JSON.h"
namespace clice::config {
struct InlayHintsOptions {
/// If false, inlay hints are completely disabled.
bool enabled = true;
// Whether specific categories of hints are enabled.
bool parameters = true;
bool deduced_types = true;
bool designators = true;
bool block_end = false;
bool default_arguments = false;
// Limit the length of type names in inlay hints. (0 means no limit)
uint32_t type_name_limit = 32;
};
} // namespace clice::config
namespace clice::feature {
enum class InlayHintKind {
Parameter,
InvalidEnum,
DefaultArgument,
Type,
Designator,
BlockEnd,
};
struct InlayHint {
/// The position offset of the inlay hint in the source code.
std::uint32_t offset;
/// The kind/category of the inlay hint.
InlayHintKind kind;
/// The label parts of the inlay hint.
/// Each SymbolID consists of two parts: the symbol name and its USR hash.
/// For symbols without a USR (e.g., built-in types or function parameters),
/// the symbol hash will be empty.
/// Otherwise, the symbol hash is non-empty and can be used for "go-to-definition".
std::vector<index::SymbolID> parts;
};
auto inlay_hints(CompilationUnitRef unit,
LocalSourceRange target,
const config::InlayHintsOptions& options) -> std::vector<InlayHint>;
} // namespace clice::feature

85
include/Feature/Lookup.h Normal file
View File

@@ -0,0 +1,85 @@
#pragma once
#include "AST/SymbolKind.h"
#include "Server/Protocol.h"
#include "Support/Struct.h"
namespace clice::proto {
struct WorkDoneProgressOptions {
/// Report on work done progress.
bool workDoneProgress = false;
};
struct PartialResultParams {};
/// The options of the all lookup.
using LookupOptions = WorkDoneProgressOptions;
/// The parameters of the simple lookup(definition, declaration,
/// type definition, implementation and reference).
inherited_struct(ReferenceParams, TextDocumentPositionParams, PartialResultParams){};
/// The result of the simple lookup.
using ReferenceResult = std::vector<Location>;
/// The parameters of the all hierarchy resolve(call hierarchy and type hierarchy)
using HierarchyPrepareParams = TextDocumentPositionParams;
struct HierarchyItem {
/// The name of the item.
string name;
/// The kind of the item.
SymbolKind kind;
/// The resource identifier of this item.
DocumentUri uri;
/// The range enclosing this symbol not including leading/trailing whitespace
/// but everything else, e.g. comments and code.
Range range;
/// The range that should be selected and revealed when this symbol is being
/// picked, e.g. the name of a function. Must be contained by the
/// [`range`](#CallHierarchyItem.range).
Range selectionRange;
/// A customized data of the item. We use it to store
/// the USR hash of the item.
uint64_t data = 0;
};
using HierarchyPrepareResult = std::vector<HierarchyItem>;
/// The parameters of the both call hierarchy and type hierarchy.
inherited_struct(HierarchyParams, TextDocumentPositionParams, PartialResultParams) {
HierarchyItem item;
};
struct CallHierarchyIncomingCall {
/// The item that makes the call.
HierarchyItem from;
/// The ranges at which the calls appear. This is relative to the caller
/// denoted by [`this.from`](#CallHierarchyIncomingCall.from).
std::vector<Range> fromRanges;
};
using CallHierarchyIncomingCallsResult = std::vector<CallHierarchyIncomingCall>;
struct CallHierarchyOutgoingCall {
/// The item that is called.
HierarchyItem to;
/// The range at which this item is called. This is the range relative to
/// the caller, e.g the item passed to `callHierarchy/outgoingCalls` request.
std::vector<Range> fromRanges;
};
using CallHierarchyOutgoingCallsResult = std::vector<CallHierarchyOutgoingCall>;
/// The result of the both super and sub type hierarchy.
using TypeHierarchyResult = std::vector<HierarchyItem>;
} // namespace clice::proto

View File

@@ -0,0 +1,29 @@
#pragma once
#include "AST/SourceCode.h"
#include "AST/SymbolKind.h"
#include "Index/Shared.h"
namespace clice::config {
struct SemanticTokensOption {};
}; // namespace clice::config
namespace clice::feature {
struct SemanticToken {
LocalSourceRange range;
SymbolKind kind;
SymbolModifiers modifiers;
};
using SemanticTokens = std::vector<SemanticToken>;
/// Generate semantic tokens for the interested file only.
SemanticTokens semantic_tokens(CompilationUnitRef unit);
/// Generate semantic tokens for all files.
index::Shared<SemanticTokens> index_semantic_token(CompilationUnitRef unit);
} // namespace clice::feature

View File

@@ -0,0 +1,27 @@
#pragma once
#include <cstdint>
#include <vector>
#include "Protocol/Feature/SignatureHelp.h"
#include "llvm/ADT/StringRef.h"
namespace clice {
struct CompilationParams;
namespace config {
struct SignatureHelpOption {};
} // namespace config
namespace feature {
proto::SignatureHelp signature_help(CompilationParams& params,
const config::SignatureHelpOption& option);
} // namespace feature
} // namespace clice

View File

@@ -1,11 +1,8 @@
#pragma once
#include <cassert>
#include <cstdint>
#include <string>
#include <vector>
#include "syntax/token.h"
#include "AST/SourceCode.h"
#include "llvm/ADT/DenseMap.h"
@@ -57,7 +54,7 @@ struct IncludeGraph {
return it->second;
}
std::uint32_t path_id(clang::FileID fid) const {
std::uint32_t path_id(clang::FileID fid) {
auto include = include_location_id(fid);
if(include != -1) {
return locations[include].path_id;

View File

@@ -1,11 +1,6 @@
#pragma once
#include <chrono>
#include <cstdint>
#include <memory>
#include <vector>
#include "index/tu_index.h"
#include "TUIndex.h"
#include "llvm/Support/Allocator.h"
#include "llvm/Support/MemoryBuffer.h"
@@ -64,23 +59,15 @@ public:
/// Remove the index of specific path id.
void remove(this Self& self, std::uint32_t path_id);
/// Get the stored source content for position mapping.
llvm::StringRef content(this const Self& self);
/// Merge the index with given compilation context.
void merge(this Self& self,
std::uint32_t path_id,
std::chrono::milliseconds build_at,
std::vector<IncludeLocation> include_locations,
FileIndex& index,
llvm::StringRef content);
FileIndex& index);
/// Merge the index with given header context.
void merge(this Self& self,
std::uint32_t path_id,
std::uint32_t include_id,
FileIndex& index,
llvm::StringRef content);
void merge(this Self& self, std::uint32_t path_id, std::uint32_t include_id, FileIndex& index);
friend bool operator==(MergedIndex& lhs, MergedIndex& rhs);

View File

@@ -1,16 +1,6 @@
#pragma once
#include <algorithm>
#include <cstdint>
#include <vector>
#include "index/tu_index.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/SmallString.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/Support/Allocator.h"
#include "TUIndex.h"
namespace clice::index {
@@ -30,17 +20,6 @@ struct PathPool {
auto path_id(llvm::StringRef path) {
assert(!path.empty());
// Normalize backslashes to forward slashes so that paths from different
// sources (URI decoding, CDB, clang FileManager) compare equal on
// Windows where native separators are backslashes.
llvm::SmallString<256> normalized;
if(path.contains('\\')) {
normalized = path;
std::replace(normalized.begin(), normalized.end(), '\\', '/');
path = normalized;
}
auto [it, success] = cache.try_emplace(path, paths.size());
if(!success) {
return it->second;
@@ -55,18 +34,6 @@ struct PathPool {
llvm::StringRef path(std::uint32_t id) {
return paths[id];
}
/// Look up a path in the cache, normalizing backslashes first.
/// Returns cache.end() if the path is not interned.
auto find(llvm::StringRef path) {
llvm::SmallString<256> normalized;
if(path.contains('\\')) {
normalized = path;
std::replace(normalized.begin(), normalized.end(), '\\', '/');
path = normalized;
}
return cache.find(path);
}
};
struct FileInfo {

View File

@@ -1,18 +1,12 @@
#pragma once
#include <array>
#include <bit>
#include <chrono>
#include <cstdint>
#include <string>
#include <vector>
#include "index/include_graph.h"
#include "semantic/relation_kind.h"
#include "semantic/symbol_kind.h"
#include "support/bitmap.h"
#include "llvm/Support/raw_ostream.h"
#include "IncludeGraph.h"
#include "AST/RelationKind.h"
#include "AST/SourceCode.h"
#include "AST/SymbolKind.h"
#include "Support/Bitmap.h"
namespace clice::index {
@@ -79,17 +73,9 @@ struct TUIndex {
llvm::DenseMap<clang::FileID, FileIndex> file_indices;
/// File indices keyed by path_id, populated by from() for deserialized data.
/// When built from AST, this is empty and file_indices (keyed by FileID) is used.
llvm::DenseMap<std::uint32_t, FileIndex> path_file_indices;
FileIndex main_file_index;
static TUIndex build(CompilationUnitRef unit, bool interested_only = false);
void serialize(llvm::raw_ostream& os) const;
static TUIndex from(const void* data);
static TUIndex build(CompilationUnitRef unit);
};
} // namespace clice::index

View File

@@ -1,9 +1,7 @@
#pragma once
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/StringRef.h"
#include "clang/AST/Decl.h"
#include "clang/Basic/SourceManager.h"
namespace clice::index {

103
include/Index/schema.fbs Normal file
View File

@@ -0,0 +1,103 @@
namespace clice.index.binary;
struct Range {
begin: uint;
end: uint;
}
struct Occurrence {
range: Range;
target: ulong;
}
struct Relation {
kind: uint;
padding: uint;
range: Range;
target_symbol: ulong;
}
table CacheEntry {
sha256: string;
canonical_id: uint;
}
struct IncludeContext {
include_id: uint;
canonical_id: uint;
}
table HeaderContextEntry {
path_id: uint;
version: uint;
includes: [IncludeContext];
}
struct IncludeLocation {
path_id: uint;
line: uint;
include_id: uint;
}
table CompilationContextEntry {
path_id: uint;
version: uint;
canonical_id: uint;
build_at: ulong;
include_locations: [IncludeLocation];
}
table OccurrenceEntry {
occurrence: Occurrence;
context: [ubyte];
}
table RelationEntry {
relation: Relation;
context: [ubyte];
}
table SymbolRelationsEntry {
symbol: ulong;
relations: [RelationEntry];
}
table Symbol {
kind: ubyte;
refs: [ubyte];
}
table SymbolEntry {
symbol_id: ulong;
symbol: Symbol;
}
table MergedIndex {
max_canonical_id: uint;
canonical_cache: [CacheEntry];
header_contexts: [HeaderContextEntry];
compilation_contexts: [CompilationContextEntry];
occurrences: [OccurrenceEntry];
relations: [SymbolRelationsEntry];
}
table PathEntry {
path: string;
id: uint;
}
struct PathMapEntry {
source: uint;
index: uint;
}
table ProjectIndex {
paths: [PathEntry];
indices: [PathMapEntry];
symbols: [SymbolEntry];
}

139
include/Protocol/Basic.h Normal file
View File

@@ -0,0 +1,139 @@
#pragma once
#include <array>
#include <cstdint>
#include <optional>
#include <string>
#include <vector>
#include "llvm/Support/JSON.h"
namespace clice::proto {
using integer = std::int32_t;
/// range in [0, 2^31- 1]
using uinteger = std::uint32_t;
using decimal = double;
using string = std::string;
using any = llvm::json::Value;
template <typename T>
using array = std::vector<T>;
template <typename T>
using optional = std::optional<T>;
using PositionEncodingKind = string;
struct WorkDoneProgressOptions {
bool workDoneProgress;
};
using URI = string;
using DocumentUri = string;
enum class ErrorCodes : integer {
/// Defined by JSON-RPC.
ParseError = -32700,
InvalidRequest = -32600,
MethodNotFound = -32601,
InvalidParams = -32602,
InternalError = -32603,
/// JSON-RPC error code indicating a server error.
serverErrorStart = -32099,
serverErrorEnd = -32000,
ServerNotInitialized = -32002,
UnknownErrorCode = -32001,
/// Defined by the protocol.
RequestFailed = -32803,
ServerCancelled = -32802,
ContentModified = -32801,
RequestCancelled = -32800
};
struct Position {
/// Line position in a document (zero-based).
uinteger line;
/// Character offset on a line in a document (zero-based).
/// The meaning of this offset is determined by the negotiated
/// `PositionEncodingKind`.
uinteger character;
constexpr friend bool operator==(const Position&, const Position&) = default;
};
struct Range {
/// The range's start position.
Position start;
/// The range's end position.
Position end;
constexpr friend bool operator==(const Range&, const Range&) = default;
};
struct Location {
DocumentUri uri;
Range range;
};
struct TextEdit {
/// The range of the text document to be manipulated. To insert
/// text into a document create a range where start === end.
Range range;
// The string to be inserted. For delete operations use an
// empty string.
string newText;
};
struct TextDocumentItem {
/// The text document's URI.
DocumentUri uri;
/// The text document's language identifier.
string languageId;
/// The version number of this document (it will strictly increase after each
/// change, including undo/redo).
uinteger version;
/// The content of the opened text document.
string text;
};
struct TextDocumentIdentifier {
/// The text document's URI.
DocumentUri uri;
};
struct VersionedTextDocumentIdentifier {
/// The text document's URI.
DocumentUri uri;
/// The version of document.
integer version;
};
struct TextDocumentPositionParams {
/// The text document.
TextDocumentIdentifier textDocument;
/// The position inside the text document.
Position position;
};
struct MarkupContent {
/// The type of the Markup.
string kind;
/// The content itself.
string value;
};
} // namespace clice::proto

Some files were not shown because too many files have changed in this diff Show More