codemap-rust 0.1.0a1__tar.gz

codemap_rust-0.1.0a1/.gitignore

@@ -0,0 +1,43 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# Build artifacts
+build/
+dist/
+*.egg-info/
+*.egg
+.eggs/
+
+# Test / coverage
+.pytest_cache/
+.coverage
+.coverage.*
+htmlcov/
+coverage.xml
+.tox/
+.mypy_cache/
+.ruff_cache/
+.benchmarks/
+
+# Virtualenv
+.venv/
+venv/
+env/
+
+# uv / pdm lockfiles (commit uv.lock once we settle)
+# uv.lock
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# OS
+.DS_Store
+Thumbs.db
+
+# CodeMap own index when dogfooding
+.codemap/

codemap_rust-0.1.0a1/PKG-INFO

@@ -0,0 +1,72 @@
+Metadata-Version: 2.4
+Name: codemap-rust
+Version: 0.1.0a1
+Summary: Rust indexer plugin for CodeMap
+Project-URL: Homepage, https://github.com/qxbyte/codemap
+Author: CodeMap Contributors
+License: MIT
+Keywords: codemap,indexer,rust,tree-sitter
+Classifier: Development Status :: 3 - Alpha
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Rust
+Classifier: Topic :: Software Development
+Requires-Python: >=3.11
+Requires-Dist: codemap-core<0.2,>=0.1.0a1
+Requires-Dist: tree-sitter-rust>=0.23
+Requires-Dist: tree-sitter>=0.25
+Provides-Extra: dev
+Requires-Dist: pytest>=8.0; extra == 'dev'
+Description-Content-Type: text/markdown
+
+# codemap-rust
+
+> A Rust indexer for [CodeMap](https://github.com/qxbyte/codemap), shipped
+> as an independent PyPI package.
+
+## What it captures
+
+Backed by `tree-sitter-rust`:
+
+| AST node | Symbol kind |
+|---|---|
+| `function_item` (free) | `function` |
+| `function_item` (inside `impl`) | `method` (attached to the impl'd type) |
+| `function_signature_item` (inside `trait`) | `method` (attached to the trait) |
+| `struct_item` | `class` (with `extra.rust_kind=struct`) |
+| `enum_item` | `class` (with `extra.rust_kind=enum`) |
+| `trait_item` | `class` (with `extra.rust_kind=trait`) |
+| `const_item` | `variable` (with `extra.rust_kind=const`) |
+| `static_item` | `variable` (with `extra.rust_kind=static`) |
+
+`impl Trait for Type` blocks attach methods to `Type` (the receiver), so
+both inherent and trait impls end up addressable under the same
+`Type#method()` shape.
+
+## SymbolID encoding
+
+```
+scip-rust . . . src/user.rs/User#login().
+```
+
+## Install
+
+```bash
+pip install "git+https://github.com/qxbyte/codemap.git#subdirectory=plugins/codemap-rust"
+```
+
+After install, `codemap doctor` lists `rust` alongside the other
+indexers on identical terms (ADR-004 + ADR-L001).
+
+## Limits
+
+* Methods on generic types (`impl<T> Foo<T>`) attach to `Foo` without
+  the generic parameter — adequate for code search, lossy for type
+  checking.
+* Macros are not expanded; their declarations are not captured as
+  symbols.
+* `mod` declarations are not turned into namespace prefixes; the file
+  path is the only namespace.
+
+## License
+
+MIT.

codemap_rust-0.1.0a1/README.md

@@ -0,0 +1,52 @@
+# codemap-rust
+
+> A Rust indexer for [CodeMap](https://github.com/qxbyte/codemap), shipped
+> as an independent PyPI package.
+
+## What it captures
+
+Backed by `tree-sitter-rust`:
+
+| AST node | Symbol kind |
+|---|---|
+| `function_item` (free) | `function` |
+| `function_item` (inside `impl`) | `method` (attached to the impl'd type) |
+| `function_signature_item` (inside `trait`) | `method` (attached to the trait) |
+| `struct_item` | `class` (with `extra.rust_kind=struct`) |
+| `enum_item` | `class` (with `extra.rust_kind=enum`) |
+| `trait_item` | `class` (with `extra.rust_kind=trait`) |
+| `const_item` | `variable` (with `extra.rust_kind=const`) |
+| `static_item` | `variable` (with `extra.rust_kind=static`) |
+
+`impl Trait for Type` blocks attach methods to `Type` (the receiver), so
+both inherent and trait impls end up addressable under the same
+`Type#method()` shape.
+
+## SymbolID encoding
+
+```
+scip-rust . . . src/user.rs/User#login().
+```
+
+## Install
+
+```bash
+pip install "git+https://github.com/qxbyte/codemap.git#subdirectory=plugins/codemap-rust"
+```
+
+After install, `codemap doctor` lists `rust` alongside the other
+indexers on identical terms (ADR-004 + ADR-L001).
+
+## Limits
+
+* Methods on generic types (`impl<T> Foo<T>`) attach to `Foo` without
+  the generic parameter — adequate for code search, lossy for type
+  checking.
+* Macros are not expanded; their declarations are not captured as
+  symbols.
+* `mod` declarations are not turned into namespace prefixes; the file
+  path is the only namespace.
+
+## License
+
+MIT.

codemap_rust-0.1.0a1/pyproject.toml

@@ -0,0 +1,36 @@
+[build-system]
+requires = ["hatchling>=1.21"]
+build-backend = "hatchling.build"
+
+[project]
+name = "codemap-rust"
+version = "0.1.0a1"
+description = "Rust indexer plugin for CodeMap"
+readme = "README.md"
+requires-python = ">=3.11"
+license = { text = "MIT" }
+authors = [{ name = "CodeMap Contributors" }]
+keywords = ["codemap", "rust", "indexer", "tree-sitter"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Rust",
+    "Topic :: Software Development",
+]
+dependencies = [
+    "codemap-core>=0.1.0a1,<0.2",
+    "tree-sitter>=0.25",
+    "tree-sitter-rust>=0.23",
+]
+
+[project.optional-dependencies]
+dev = ["pytest>=8.0"]
+
+[project.entry-points."codemap.indexers"]
+rust = "codemap_rust:RustIndexer"
+
+[project.urls]
+Homepage = "https://github.com/qxbyte/codemap"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/codemap_rust"]

codemap_rust-0.1.0a1/src/codemap_rust/__init__.py

@@ -0,0 +1,8 @@
+"""Rust indexer plugin for CodeMap."""
+
+from __future__ import annotations
+
+from codemap_rust.indexer import RustIndexer
+
+__all__ = ["RustIndexer"]
+__version__ = "0.1.0"

codemap_rust-0.1.0a1/src/codemap_rust/indexer.py

@@ -0,0 +1,324 @@
+"""Rust indexer built on tree-sitter-rust.
+
+Notable Rust-specific behaviour: ``impl_item`` blocks (both inherent and
+trait impls) attach their inner functions to the impl'd type's namespace,
+so ``impl User { fn login(&self) {} }`` and ``impl Greeter for User {
+fn hello(&self) {} }`` both produce ``User#xxx().`` SymbolIDs. Methods
+defined directly on a trait (``trait T { fn m(&self); }``) attach to the
+trait the same way.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path, PurePosixPath
+from typing import ClassVar
+
+import tree_sitter
+import tree_sitter_rust
+
+from codemap.core.models import Diagnostic, Edge, IndexResult, Range, Symbol
+from codemap.core.symbol import Descriptor, DescriptorKind, SymbolID
+from codemap.indexers.base import IndexContext
+
+SCHEME = "scip-rust"
+LANG = "rust"
+
+_RUST_LANG = tree_sitter.Language(tree_sitter_rust.language())
+
+
+class RustIndexer:
+    name: ClassVar[str] = "rust"
+    version: ClassVar[str] = "0.1.0"
+    file_patterns: ClassVar[list[str]] = ["*.rs"]
+    languages: ClassVar[list[str]] = [LANG]
+
+    def supports(self, path: Path) -> bool:
+        return path.suffix == ".rs"
+
+    def index_file(
+        self,
+        path: Path,
+        source: bytes,
+        ctx: IndexContext,
+    ) -> IndexResult:
+        try:
+            source.decode("utf-8")
+        except UnicodeDecodeError as exc:
+            return IndexResult(
+                diagnostics=[
+                    Diagnostic(
+                        severity="error",
+                        file=ctx.relative_path,
+                        code="RS002",
+                        message=f"not valid UTF-8: {exc}",
+                        producer=self.name,
+                    )
+                ]
+            )
+        parser = tree_sitter.Parser(_RUST_LANG)
+        tree = parser.parse(source)
+        visitor = _Visitor(ctx.relative_path)
+        visitor.visit(tree.root_node)
+        diagnostics = list(visitor.diagnostics)
+        if tree.root_node.has_error:
+            diagnostics.append(
+                Diagnostic(
+                    severity="warning",
+                    file=ctx.relative_path,
+                    range=Range(start_line=1, end_line=1),
+                    code="RS001",
+                    message="tree-sitter reported parse errors; symbols may be incomplete",
+                    producer=self.name,
+                )
+            )
+        return IndexResult(
+            symbols=visitor.symbols,
+            edges=visitor.edges,
+            diagnostics=diagnostics,
+        )
+
+
+class _Visitor:
+    def __init__(self, relative_path: PurePosixPath) -> None:
+        self.relative_path = relative_path
+        self.symbols: list[Symbol] = []
+        self.edges: list[Edge] = []
+        self.diagnostics: list[Diagnostic] = []
+        self._type_stack: list[str] = []
+
+    def visit(self, node: tree_sitter.Node) -> None:
+        kind = node.type
+        if kind == "function_item":
+            self._visit_function_item(node)
+            return
+        if kind == "function_signature_item" and self._type_stack:
+            self._visit_function_signature_item(node)
+            return
+        if kind == "struct_item":
+            self._visit_type_decl(node, rust_kind="struct")
+            return
+        if kind == "enum_item":
+            self._visit_type_decl(node, rust_kind="enum")
+            return
+        if kind == "trait_item":
+            self._visit_trait_item(node)
+            return
+        if kind == "impl_item":
+            self._visit_impl_item(node)
+            return
+        if kind == "const_item":
+            self._visit_const_or_static(node, rust_kind="const")
+            return
+        if kind == "static_item":
+            self._visit_const_or_static(node, rust_kind="static")
+            return
+        for child in node.children:
+            self.visit(child)
+
+    # ---------------------------------------------------- functions
+
+    def _visit_function_item(self, node: tree_sitter.Node) -> None:
+        name_node = node.child_by_field_name("name")
+        if name_node is None:
+            return
+        name = _node_text(name_node)
+        if not name:
+            return
+        kind: str = "method" if self._type_stack else "function"
+        sid = self._make_id(name, kind=DescriptorKind.METHOD)
+        self.symbols.append(
+            Symbol(
+                id=sid,
+                kind=kind,  # type: ignore[arg-type]
+                language=LANG,
+                file=self.relative_path,
+                range=_node_range(node),
+                signature=_fn_signature(node, name),
+            )
+        )
+
+    def _visit_function_signature_item(self, node: tree_sitter.Node) -> None:
+        name_node = node.child_by_field_name("name")
+        if name_node is None:
+            return
+        name = _node_text(name_node)
+        if not name:
+            return
+        sid = self._make_id(name, kind=DescriptorKind.METHOD)
+        self.symbols.append(
+            Symbol(
+                id=sid,
+                kind="method",
+                language=LANG,
+                file=self.relative_path,
+                range=_node_range(node),
+                signature=_fn_signature(node, name),
+                extra={"rust_kind": "trait_method"},
+            )
+        )
+
+    # ------------------------------------------------------ types
+
+    def _visit_type_decl(self, node: tree_sitter.Node, *, rust_kind: str) -> None:
+        name_node = node.child_by_field_name("name")
+        if name_node is None:
+            return
+        name = _node_text(name_node)
+        if not name:
+            return
+        sid = self._make_id(name, kind=DescriptorKind.TYPE)
+        self.symbols.append(
+            Symbol(
+                id=sid,
+                kind="class",
+                language=LANG,
+                file=self.relative_path,
+                range=_node_range(node),
+                extra={"rust_kind": rust_kind},
+            )
+        )
+
+    def _visit_trait_item(self, node: tree_sitter.Node) -> None:
+        name_node = node.child_by_field_name("name")
+        if name_node is None:
+            return
+        name = _node_text(name_node)
+        if not name:
+            return
+        sid = self._make_id(name, kind=DescriptorKind.TYPE)
+        self.symbols.append(
+            Symbol(
+                id=sid,
+                kind="class",
+                language=LANG,
+                file=self.relative_path,
+                range=_node_range(node),
+                extra={"rust_kind": "trait"},
+            )
+        )
+        body = node.child_by_field_name("body")
+        if body is None:
+            return
+        self._type_stack.append(name)
+        try:
+            for child in body.children:
+                self.visit(child)
+        finally:
+            self._type_stack.pop()
+
+    def _visit_impl_item(self, node: tree_sitter.Node) -> None:
+        owner = _impl_owner_type(node)
+        body = node.child_by_field_name("body")
+        if owner is None or body is None:
+            # Walk anyway to capture function symbols at file scope.
+            for child in node.children:
+                self.visit(child)
+            return
+        self._type_stack.append(owner)
+        try:
+            for child in body.children:
+                self.visit(child)
+        finally:
+            self._type_stack.pop()
+
+    # ----------------------------------------------- const / static
+
+    def _visit_const_or_static(self, node: tree_sitter.Node, *, rust_kind: str) -> None:
+        name_node = node.child_by_field_name("name")
+        if name_node is None:
+            return
+        name = _node_text(name_node)
+        if not name:
+            return
+        sid = self._make_id(name, kind=DescriptorKind.TERM)
+        self.symbols.append(
+            Symbol(
+                id=sid,
+                kind="variable",
+                language=LANG,
+                file=self.relative_path,
+                range=_node_range(node),
+                extra={"rust_kind": rust_kind},
+            )
+        )
+
+    # ----------------------------------------------------- helpers
+
+    def _make_id(self, name: str, *, kind: DescriptorKind) -> SymbolID:
+        descriptors = list(_path_namespaces(self.relative_path))
+        descriptors.extend(Descriptor(name=t, kind=DescriptorKind.TYPE) for t in self._type_stack)
+        descriptors.append(Descriptor(name=name, kind=kind))
+        return SymbolID(scheme=SCHEME, descriptors=tuple(descriptors))
+
+
+# ---------------------------------------------------------------------------
+# Pure helpers
+# ---------------------------------------------------------------------------
+
+
+def _path_namespaces(path: PurePosixPath) -> list[Descriptor]:
+    return [Descriptor(name=part, kind=DescriptorKind.NAMESPACE) for part in path.parts]
+
+
+def _node_range(node: tree_sitter.Node) -> Range:
+    sr, sc = node.start_point
+    er, ec = node.end_point
+    return Range(
+        start_line=sr + 1,
+        start_col=sc,
+        end_line=max(er + 1, sr + 1),
+        end_col=ec,
+    )
+
+
+def _node_text(node: tree_sitter.Node) -> str:
+    return node.text.decode("utf-8") if node.text is not None else ""
+
+
+def _fn_signature(node: tree_sitter.Node, name: str) -> str:
+    params = node.child_by_field_name("parameters")
+    params_text = _node_text(params) if params is not None else "()"
+    return_type = node.child_by_field_name("return_type")
+    rt_text = (" -> " + _node_text(return_type)) if return_type is not None else ""
+    return f"fn {name}{params_text}{rt_text}"
+
+
+def _impl_owner_type(impl_node: tree_sitter.Node) -> str | None:
+    """Return the owning type of an impl block.
+
+    For ``impl Foo { ... }`` returns ``"Foo"``. For
+    ``impl Trait for Foo { ... }`` returns ``"Foo"`` (the receiver, not
+    the trait).
+    """
+    # Field-named child ``type`` is the receiver of the impl (both inherent
+    # and trait impls).
+    type_node = impl_node.child_by_field_name("type")
+    if type_node is not None:
+        return _unwrap_type(type_node)
+    # Fallback: walk children, take the last `type_identifier` before the
+    # `declaration_list`.
+    last: tree_sitter.Node | None = None
+    for child in impl_node.children:
+        if child.type == "declaration_list":
+            break
+        if child.type in {"type_identifier", "generic_type", "scoped_type_identifier"}:
+            last = child
+    if last is None:
+        return None
+    return _unwrap_type(last)
+
+
+def _unwrap_type(node: tree_sitter.Node) -> str | None:
+    if node.type == "type_identifier":
+        return _node_text(node)
+    if node.type == "generic_type":
+        for c in node.children:
+            if c.type == "type_identifier":
+                return _node_text(c)
+        return None
+    if node.type == "scoped_type_identifier":
+        for c in reversed(node.children):
+            if c.type == "type_identifier":
+                return _node_text(c)
+        return None
+    return None

codemap_rust-0.1.0a1/tests/test_indexer.py

@@ -0,0 +1,165 @@
+"""Unit tests for the Rust indexer plugin."""
+
+from __future__ import annotations
+
+import textwrap
+from pathlib import Path, PurePosixPath
+
+from codemap_rust import RustIndexer
+from codemap_rust.indexer import SCHEME
+
+from codemap.core.models import IndexResult
+from codemap.indexers.base import IndexContext
+
+
+def _index(source: str, *, path: str = "src/lib.rs") -> IndexResult:
+    code = textwrap.dedent(source).lstrip("\n")
+    return RustIndexer().index_file(
+        Path(path),
+        code.encode("utf-8"),
+        IndexContext(
+            project_root=Path("/tmp/proj"),
+            relative_path=PurePosixPath(path),
+            language="rust",
+        ),
+    )
+
+
+# ---------------------------------------------------------------------------
+# Metadata
+# ---------------------------------------------------------------------------
+
+
+def test_indexer_metadata() -> None:
+    ix = RustIndexer()
+    assert ix.name == "rust"
+    assert ix.languages == ["rust"]
+    assert ix.supports(Path("a.rs"))
+    assert not ix.supports(Path("a.go"))
+
+
+def test_scheme_is_consistent() -> None:
+    r = _index(
+        """
+        fn f() {}
+        struct S;
+        const C: u32 = 1;
+        """
+    )
+    for s in r.symbols:
+        assert str(s.id).startswith(f"{SCHEME} ")
+
+
+# ---------------------------------------------------------------------------
+# Function / impl
+# ---------------------------------------------------------------------------
+
+
+def test_free_function() -> None:
+    r = _index("fn helper(x: u32) -> u32 { x + 1 }")
+    funcs = [s for s in r.symbols if s.kind == "function"]
+    assert len(funcs) == 1
+    assert "helper()." in str(funcs[0].id)
+
+
+def test_inherent_impl_method_attaches_to_type() -> None:
+    r = _index(
+        """
+        struct User;
+        impl User {
+            fn new() -> Self { User }
+            fn login(&mut self) -> bool { false }
+        }
+        """
+    )
+    methods = [s for s in r.symbols if s.kind == "method"]
+    method_ids = {str(s.id) for s in methods}
+    assert any("User#new()." in i for i in method_ids)
+    assert any("User#login()." in i for i in method_ids)
+
+
+def test_trait_impl_method_attaches_to_receiver_type() -> None:
+    r = _index(
+        """
+        trait Greeter { fn hello(&self) -> String; }
+        struct User;
+        impl Greeter for User {
+            fn hello(&self) -> String { String::new() }
+        }
+        """
+    )
+    methods = [s for s in r.symbols if s.kind == "method"]
+    # User#hello (impl) + Greeter#hello (trait signature) → 2 methods
+    ids = {str(s.id) for s in methods}
+    assert any("User#hello()." in i for i in ids)
+    assert any("Greeter#hello()." in i for i in ids)
+
+
+# ---------------------------------------------------------------------------
+# Type-level items
+# ---------------------------------------------------------------------------
+
+
+def test_struct_decl() -> None:
+    r = _index("struct User { id: u32 }")
+    classes = [s for s in r.symbols if s.kind == "class"]
+    assert classes[0].extra.get("rust_kind") == "struct"
+
+
+def test_enum_decl() -> None:
+    r = _index("enum Color { Red, Green, Blue }")
+    classes = [s for s in r.symbols if s.kind == "class"]
+    assert classes[0].extra.get("rust_kind") == "enum"
+
+
+def test_trait_decl() -> None:
+    r = _index("trait Greeter { fn hello(&self) -> String; }")
+    classes = [s for s in r.symbols if s.kind == "class"]
+    assert classes[0].extra.get("rust_kind") == "trait"
+
+
+# ---------------------------------------------------------------------------
+# Const / static
+# ---------------------------------------------------------------------------
+
+
+def test_const_decl() -> None:
+    r = _index("const MAX: u32 = 10;")
+    vars_ = [s for s in r.symbols if s.kind == "variable"]
+    assert vars_[0].extra.get("rust_kind") == "const"
+
+
+def test_static_decl() -> None:
+    r = _index("static COUNTER: u32 = 0;")
+    vars_ = [s for s in r.symbols if s.kind == "variable"]
+    assert vars_[0].extra.get("rust_kind") == "static"
+
+
+# ---------------------------------------------------------------------------
+# SymbolID + edge cases
+# ---------------------------------------------------------------------------
+
+
+def test_symbol_id_uses_path_namespaces() -> None:
+    r = _index("fn helper() {}", path="src/user.rs")
+    assert str(r.symbols[0].id) == "scip-rust . . . src/user.rs/helper()."
+
+
+def test_empty_file_yields_no_symbols() -> None:
+    r = _index("")
+    assert r.symbols == []
+
+
+def test_invalid_utf8_yields_diagnostic() -> None:
+    ix = RustIndexer()
+    r = ix.index_file(
+        Path("bad.rs"),
+        b"\xff\xfe fn",
+        IndexContext(
+            project_root=Path("/tmp/proj"),
+            relative_path=PurePosixPath("bad.rs"),
+            language="rust",
+        ),
+    )
+    assert r.symbols == []
+    assert r.diagnostics[0].code == "RS002"