diff --git a/src/semble/__init__.py b/src/semble/__init__.py index 6265930..826a11e 100644 --- a/src/semble/__init__.py +++ b/src/semble/__init__.py @@ -1,14 +1,14 @@ from semble.index import SembleIndex from semble.types import Chunk, EmbeddingMatrix, Encoder, IndexStats, SearchMode, SearchResult -from semble.version import __version__, __version_triple__ +from semble.version import __version__ __all__ = [ - "SembleIndex", "Chunk", "EmbeddingMatrix", "Encoder", "IndexStats", "SearchMode", "SearchResult", + "SembleIndex", "__version__", ] diff --git a/src/semble/index/create.py b/src/semble/index/create.py index 76ac6a3..80eac92 100644 --- a/src/semble/index/create.py +++ b/src/semble/index/create.py @@ -1,5 +1,3 @@ -from __future__ import annotations - import contextlib from pathlib import Path diff --git a/src/semble/index/dense.py b/src/semble/index/dense.py index 2f864ee..13c9620 100644 --- a/src/semble/index/dense.py +++ b/src/semble/index/dense.py @@ -1,5 +1,3 @@ -from __future__ import annotations - import numpy as np import numpy.typing as npt from huggingface_hub.utils.tqdm import disable_progress_bars diff --git a/src/semble/index/sparse.py b/src/semble/index/sparse.py index 9fd839d..921c1da 100644 --- a/src/semble/index/sparse.py +++ b/src/semble/index/sparse.py @@ -1,5 +1,3 @@ -from __future__ import annotations - import contextlib from pathlib import Path diff --git a/src/semble/mcp.py b/src/semble/mcp.py index e8b52a2..17921b6 100644 --- a/src/semble/mcp.py +++ b/src/semble/mcp.py @@ -121,10 +121,7 @@ def __init__(self, model: Encoder) -> None: async def get(self, source: str, ref: str | None = None) -> SembleIndex: """Return an index for the requested source, building and caching it on first access.""" is_git = _is_git_url(source) - if is_git: - cache_key = f"{source}@{ref}" if ref else source - else: - cache_key = str(Path(source).resolve()) + cache_key = (f"{source}@{ref}" if ref else source) if is_git else str(Path(source).resolve()) if cache_key not in self._tasks: if is_git: diff --git a/src/semble/tokens.py b/src/semble/tokens.py index bb787c8..417bcb4 100644 --- a/src/semble/tokens.py +++ b/src/semble/tokens.py @@ -20,15 +20,15 @@ def _split_identifier(token: str) -> list[str]: lower = token.lower() parts: list[str] = [] - # snake_case splitting if "_" in token: + # snake_case splitting parts = [p for p in lower.split("_") if p] else: # camelCase / PascalCase splitting parts = [m.lower() for m in _CAMEL_RE.findall(token)] if len(parts) >= 2: - return [lower] + parts + return [lower, *parts] return [lower]