rbtr-lang-python 2026.7.0.dev0__py3-none-any.whl

rbtr_lang_python/tests/test_docstrings.py

@@ -0,0 +1,54 @@
+"""Python docstring extraction.
+
+rbtr folds a symbol's docstring into its chunk content. Python is an
+interior-doc language — the docstring is captured by the `@_docstring` query
+capture, orthogonal to the leading-comment sibling walk — so suppressing that
+walk leaves the doc (and `line_start`) unchanged. Data lives in
+`cases_docstrings.py`, sliced by tag.
+"""
+
+from __future__ import annotations
+
+from pytest_cases import parametrize_with_cases
+
+from rbtr.git import FileEntry
+from rbtr.languages.extract import extract_file
+
+
+@parametrize_with_cases(
+    "lang, source, name, snippet", cases=".cases_docstrings", has_tag="documented"
+)
+def test_documented_chunk_includes_doc_text(
+    lang: str, source: str, name: str, snippet: str
+) -> None:
+    """By default the chunk content carries the symbol's docs."""
+    chunks = extract_file(FileEntry("input", "sha1", source.encode()), lang)
+    chunk = next(c for c in chunks if c.name == name)
+    assert snippet in chunk.content, (
+        f"expected {snippet!r} in {lang}.{name} content; got:\n{chunk.content!r}"
+    )
+
+
+@parametrize_with_cases(
+    "lang, source, name, snippet", cases=".cases_docstrings", has_tag="undocumented"
+)
+def test_no_phantom_documentation(lang: str, source: str, name: str, snippet: str) -> None:
+    """Symbols without documentation do not gain any in content."""
+    chunks = extract_file(FileEntry("input", "sha1", source.encode()), lang)
+    chunk = next(c for c in chunks if c.name == name)
+    assert snippet not in chunk.content, (
+        f"unexpected {snippet!r} in {lang}.{name} content; got:\n{chunk.content!r}"
+    )
+
+
+@parametrize_with_cases(
+    "lang, source, name, snippet", cases=".cases_docstrings", has_tag="interior_doc"
+)
+def test_interior_docstring_folds_into_content(
+    lang: str, source: str, name: str, snippet: str
+) -> None:
+    """An interior (`@_docstring`) doc is part of its symbol's chunk content."""
+    chunk = next(
+        c for c in extract_file(FileEntry("input", "sha1", source.encode()), lang) if c.name == name
+    )
+    assert snippet in chunk.content

rbtr_lang_python/tests/test_extraction.py

@@ -0,0 +1,110 @@
+"""Python extraction tests.
+
+Symbol, import, and mixed cases (`cases_extraction.py`) drive the shared
+checks; the functions below pin Python's module-variable, function-local,
+class-attribute, and tuple-unpacking edge behaviour.
+"""
+
+from __future__ import annotations
+
+from pytest_cases import parametrize_with_cases
+
+from rbtr.git import FileEntry
+from rbtr.index.models import ChunkKind, ImportMeta
+from rbtr.languages.extract import extract_file
+
+
+@parametrize_with_cases("lang, source, expected", cases=".cases_extraction", has_tag="symbol")
+def test_extracts_expected_symbols(lang: str, source: str, expected: list) -> None:
+    """Each expected (kind, name, scope) tuple appears in the output."""
+    chunks = extract_file(FileEntry("input", "sha1", source.encode()), lang)
+    symbols = [(c.kind, c.name, c.scope) for c in chunks]
+    for exp in expected:
+        assert exp in symbols, f"expected {exp} not found in {symbols}"
+
+
+@parametrize_with_cases(
+    "lang, source, expected_kinds, expected_methods", cases=".cases_extraction", has_tag="mixed"
+)
+def test_extracts_all_expected_kinds(
+    lang: str, source: str, expected_kinds: set[str], expected_methods: list[tuple[str, str]]
+) -> None:
+    """Realistic source produces all expected chunk kinds and method scoping."""
+    chunks = extract_file(FileEntry("input", "sha1", source.encode()), lang)
+    kinds = {c.kind for c in chunks}
+    for kind in expected_kinds:
+        assert kind in kinds, f"expected kind {kind!r} not in {kinds}"
+    methods = [(c.name, c.scope) for c in chunks if c.kind == ChunkKind.METHOD]
+    for name, scope in expected_methods:
+        assert (name, scope) in methods, f"expected method ({name}, {scope}) not in {methods}"
+
+
+@parametrize_with_cases("lang, source, expected", cases=".cases_extraction", has_tag="import")
+def test_extracts_import_metadata(lang: str, source: str, expected: dict) -> None:
+    """First import chunk has the expected metadata."""
+    chunks = extract_file(FileEntry("input", "sha1", source.encode()), lang)
+    imports = [c for c in chunks if c.kind == ChunkKind.IMPORT]
+    assert len(imports) >= 1, f"no import chunks extracted from {source!r}"
+    assert imports[0].metadata == ImportMeta(**expected)
+
+
+@parametrize_with_cases(
+    "lang, source, count, metadata_list", cases=".cases_extraction", has_tag="multi_import"
+)
+def test_extracts_multi_import(
+    lang: str, source: str, count: int, metadata_list: list[dict]
+) -> None:
+    """Multiple imports have correct count and per-import metadata."""
+    chunks = extract_file(FileEntry("input", "sha1", source.encode()), lang)
+    imports = [c for c in chunks if c.kind == ChunkKind.IMPORT]
+    assert len(imports) == count
+    for imp, expected in zip(imports, metadata_list, strict=True):
+        assert imp.metadata == ImportMeta(**expected)
+
+
+def test_py_module_variable_content_is_whole_statement() -> None:
+    """A module-level VARIABLE chunk spans the whole statement, named by LHS."""
+    src = """\
+MAX_SIZE = 100
+"""
+    chunks = extract_file(FileEntry("input", "sha1", src.encode()), "python")
+    variables = [c for c in chunks if c.kind == ChunkKind.VARIABLE]
+    assert len(variables) == 1
+    assert variables[0].name == "MAX_SIZE"
+    assert variables[0].content.strip() == "MAX_SIZE = 100"
+
+
+def test_py_function_local_not_captured_as_variable() -> None:
+    """Assignments inside a function stay part of the function chunk."""
+    src = """\
+def f():
+    tmp = 1
+    return tmp
+"""
+    chunks = extract_file(FileEntry("input", "sha1", src.encode()), "python")
+    assert [c for c in chunks if c.kind == ChunkKind.VARIABLE] == []
+
+
+def test_py_class_attribute_not_captured_as_variable() -> None:
+    """Class-body attributes stay part of the class chunk, not VARIABLE chunks."""
+    src = """\
+class Config:
+    DEFAULT = 30
+"""
+    chunks = extract_file(FileEntry("input", "sha1", src.encode()), "python")
+    assert [c for c in chunks if c.kind == ChunkKind.VARIABLE] == []
+
+
+def test_py_tuple_unpacking_captured_as_variables() -> None:
+    """Flat tuple-unpacking binds each target as its own VARIABLE chunk.
+
+    Both names come from one statement (tree-sitter fans the destructuring
+    into a match per identifier), and each chunk spans the whole statement.
+    """
+    src = """\
+a, b = compute()
+"""
+    chunks = extract_file(FileEntry("input", "sha1", src.encode()), "python")
+    variables = [c for c in chunks if c.kind == ChunkKind.VARIABLE]
+    assert {c.name for c in variables} == {"a", "b"}
+    assert all(c.content.strip() == "a, b = compute()" for c in variables)

rbtr_lang_python/tests/test_samples.py

@@ -0,0 +1,83 @@
+"""Python sample extraction: the `samples/python/` project through the real pipeline.
+
+The snapshots are the golden record of what Python extraction produces.
+Engine-wide invariants are covered once in core.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import TYPE_CHECKING
+
+import pytest
+from tree_sitter import Parser
+
+from rbtr.git import FileEntry
+from rbtr.index.models import Chunk, ChunkKind, Edge
+from rbtr.languages.edges import build_resolution_map, infer_import_edges
+from rbtr.languages.extract import extract_file
+from rbtr.languages.manager import get_manager
+from rbtr.testing import render_edges
+
+if TYPE_CHECKING:
+    from syrupy.assertion import SnapshotAssertion
+
+
+@pytest.fixture
+def project() -> list[tuple[str, str]]:
+    """The `(relative path, text)` files of the `samples/python/` project."""
+    root = Path(__file__).parent / "samples" / "python"
+    return [
+        (str(p.relative_to(root)), p.read_text()) for p in sorted(root.rglob("*")) if p.is_file()
+    ]
+
+
+@pytest.fixture
+def chunks(project: list[tuple[str, str]]) -> list[Chunk]:
+    """Chunks from every project file, each via the real `extract_file`."""
+    manager = get_manager()
+    out: list[Chunk] = []
+    for path, text in project:
+        lang = manager.detect_language(path) or "python"
+        out.extend(extract_file(FileEntry(path, "sha1", text.encode()), lang))
+    return out
+
+
+@pytest.fixture
+def edges(project: list[tuple[str, str]], chunks: list[Chunk]) -> list[Edge]:
+    """Import edges inferred across the project's files."""
+    manager = get_manager()
+    repo_files = {path for path, _ in project}
+    return infer_import_edges(chunks, repo_files, build_resolution_map(manager))
+
+
+def test_emits_expected_kinds(chunks: list[Chunk]) -> None:
+    """The sample exercises Python's function, class, method, variable, and import chunks."""
+    kinds = {c.kind for c in chunks}
+    assert {
+        ChunkKind.FUNCTION,
+        ChunkKind.CLASS,
+        ChunkKind.METHOD,
+        ChunkKind.VARIABLE,
+        ChunkKind.IMPORT,
+        ChunkKind.COMMENT,
+    } <= kinds
+
+
+def test_parses_cleanly(project: list[tuple[str, str]]) -> None:
+    """Every project file is valid source — no tree-sitter ERROR/MISSING nodes."""
+    manager = get_manager()
+    for path, text in project:
+        grammar = manager.grammar(manager.detect_language(path) or "python")
+        assert grammar is not None
+        assert not Parser(grammar).parse(text.encode()).root_node.has_error, path
+
+
+def test_extraction_matches_snapshot(chunks: list[Chunk], snapshot_json: SnapshotAssertion) -> None:
+    assert chunks == snapshot_json
+
+
+def test_edges_match_snapshot(
+    chunks: list[Chunk], edges: list[Edge], snapshot_json: SnapshotAssertion
+) -> None:
+    assert render_edges(edges, chunks) == snapshot_json

rbtr_lang_python-2026.7.0.dev0.dist-info/METADATA

@@ -0,0 +1,8 @@
+Metadata-Version: 2.4
+Name: rbtr-lang-python
+Version: 2026.7.0.dev0
+Summary: rbtr — Python language plugin
+License-Expression: MIT
+Requires-Dist: rbtr==2026.7.0.dev0
+Requires-Dist: tree-sitter-python
+Requires-Python: >=3.13

rbtr_lang_python-2026.7.0.dev0.dist-info/RECORD

@@ -0,0 +1,18 @@
+rbtr_lang_python/__init__.py,sha256=mdH22CFTQNelM0_lyVdigVKZOL4ErqZoCFVDc01DvHE,38
+rbtr_lang_python/plugin.py,sha256=FTnc6Vt7nog5Z2WhK3G0AbWx3kRJ-CPMvIo4gsDVaVw,4157
+rbtr_lang_python/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+rbtr_lang_python/python.scm,sha256=aIeFFw1i0nvcAWkdqJLB-B0N7FNiOdF0wKuKjxVmJiM,1602
+rbtr_lang_python/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+rbtr_lang_python/tests/__snapshots__/test_samples/test_edges_match_snapshot.json,sha256=7KeVtn7c62Hfz8Kv3p-Tp8Q88IMympsICYe935jDXwc,77
+rbtr_lang_python/tests/__snapshots__/test_samples/test_extraction_matches_snapshot.json,sha256=I4jAqsjOvt5wLIda25Tk9BpoIt0Jv9mctTAx3q2DIN0,12687
+rbtr_lang_python/tests/cases_docstrings.py,sha256=zqIjJHom1mSP0X42RpfsFf4jjrPSrM3X7jd7pMQeaxQ,4851
+rbtr_lang_python/tests/cases_extraction.py,sha256=dlGp-_M-BB1TQYG5w6BSXP2F0qfaPqa0mhC1I_rMG2Y,14436
+rbtr_lang_python/tests/samples/python/config.py,sha256=vq19QIAEQ4RSj5JNq5sprVVMACq7nwgEZ6ikCFpHnpM,59
+rbtr_lang_python/tests/samples/python/python.py,sha256=Q2Ggu-RMMM_iIHpg_YttbBK835sQ1AcfQ86BYFxemw0,2442
+rbtr_lang_python/tests/test_docstrings.py,sha256=YvjbJVWsASmSmEBGTJUgYaCeQqgHoxzyisEsplQg95U,2081
+rbtr_lang_python/tests/test_extraction.py,sha256=4OGNGIE8rOMaB00jdr5UkwNtgFxmsBg8Fyj8LfJX2bs,4580
+rbtr_lang_python/tests/test_samples.py,sha256=lrT_iSY85rl1wNOMqpfwtUB4Ys2d5OKSRJrjTYenxnE,2826
+rbtr_lang_python-2026.7.0.dev0.dist-info/WHEEL,sha256=CoDSoyhtC_eO_tlxRYzsTraPv1fPJRXFx91k6ISeAvA,81
+rbtr_lang_python-2026.7.0.dev0.dist-info/entry_points.txt,sha256=RNfHUZ0LplE85FNOnadS_BV0BM1jhRtsz_qPal60sTk,58
+rbtr_lang_python-2026.7.0.dev0.dist-info/METADATA,sha256=LIJFl35OAPWU5JuQQ5zkiPnMJ75C3mnT_rZJ2Z-aXOU,226
+rbtr_lang_python-2026.7.0.dev0.dist-info/RECORD,,

rbtr_lang_python-2026.7.0.dev0.dist-info/WHEEL

@@ -0,0 +1,4 @@
+Wheel-Version: 1.0
+Generator: uv 0.11.28
+Root-Is-Purelib: true
+Tag: py3-none-any

rbtr_lang_python-2026.7.0.dev0.dist-info/entry_points.txt

@@ -0,0 +1,3 @@
+[rbtr.languages]
+python = rbtr_lang_python.plugin:python
+