kodit 0.1.11__py3-none-any.whl → 0.1.13__py3-none-any.whl

kodit/_version.py

@@ -17,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.1.11'
-__version_tuple__ = version_tuple = (0, 1, 11)
+__version__ = version = '0.1.13'
+__version_tuple__ = version_tuple = (0, 1, 13)

kodit/bm25/bm25.py

@@ -52,7 +52,7 @@ class BM25Service:
             self.log.warning("No documents to retrieve from, returning empty list")
             return []
 
-        top_k = min(top_k, len(doc_ids))
+        top_k = min(top_k, len(self.retriever.scores))
         self.log.debug(
             "Retrieving from index", query=query, top_k=top_k, num_docs=len(doc_ids)
         )

kodit/cli.py

@@ -1,6 +1,5 @@
 """Command line interface for kodit."""
 
-import os
 import signal
 from pathlib import Path
 from typing import Any
@@ -12,35 +11,21 @@ from pytable_formatter import Cell, Table
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from kodit.config import (
-    DEFAULT_BASE_DIR,
-    DEFAULT_DB_URL,
-    DEFAULT_DISABLE_TELEMETRY,
-    DEFAULT_EMBEDDING_MODEL_NAME,
-    DEFAULT_LOG_FORMAT,
-    DEFAULT_LOG_LEVEL,
     AppContext,
     with_app_context,
     with_session,
 )
-from kodit.indexing.repository import IndexRepository
-from kodit.indexing.service import IndexService
+from kodit.embedding.embedding import embedding_factory
+from kodit.indexing.indexing_repository import IndexRepository
+from kodit.indexing.indexing_service import IndexService
 from kodit.log import configure_logging, configure_telemetry, log_event
-from kodit.retreival.repository import RetrievalRepository
-from kodit.retreival.service import RetrievalRequest, RetrievalService
-from kodit.sources.repository import SourceRepository
-from kodit.sources.service import SourceService
+from kodit.search.search_repository import SearchRepository
+from kodit.search.search_service import SearchRequest, SearchService
+from kodit.source.source_repository import SourceRepository
+from kodit.source.source_service import SourceService
 
 
 @click.group(context_settings={"max_content_width": 100})
-@click.option("--log-level", help=f"Log level [default: {DEFAULT_LOG_LEVEL}]")
-@click.option("--log-format", help=f"Log format [default: {DEFAULT_LOG_FORMAT}]")
-@click.option(
-    "--disable-telemetry",
-    is_flag=True,
-    help=f"Disable telemetry [default: {DEFAULT_DISABLE_TELEMETRY}]",
-)
-@click.option("--db-url", help=f"Database URL [default: {DEFAULT_DB_URL}]")
-@click.option("--data-dir", help=f"Data directory [default: {DEFAULT_BASE_DIR}]")
 @click.option(
     "--env-file",
     help="Path to a .env file [default: .env]",
@@ -52,13 +37,8 @@ from kodit.sources.service import SourceService
     ),
 )
 @click.pass_context
-def cli(  # noqa: PLR0913
+def cli(
     ctx: click.Context,
-    log_level: str | None,
-    log_format: str | None,
-    disable_telemetry: bool | None,
-    db_url: str | None,
-    data_dir: str | None,
     env_file: Path | None,
 ) -> None:
     """kodit CLI - Code indexing for better AI code generation."""  # noqa: D403
@@ -67,17 +47,6 @@ def cli(  # noqa: PLR0913
     if env_file:
         config = AppContext(_env_file=env_file)  # type: ignore[reportCallIssue]
 
-    # Now override with CLI arguments, if set
-    if data_dir:
-        config.data_dir = Path(data_dir)
-    if db_url:
-        config.db_url = db_url
-    if log_level:
-        config.log_level = log_level
-    if log_format:
-        config.log_format = log_format
-    if disable_telemetry:
-        config.disable_telemetry = disable_telemetry
     configure_logging(config)
     configure_telemetry(config)
 
@@ -102,7 +71,7 @@ async def index(
         repository,
         source_service,
         app_context.get_data_dir(),
-        embedding_model_name=DEFAULT_EMBEDDING_MODEL_NAME,
+        embedding_service=embedding_factory(app_context.get_default_openai_client()),
     )
 
     if not sources:
@@ -159,14 +128,14 @@ async def code(
 
     This works best if your query is code.
     """
-    repository = RetrievalRepository(session)
-    service = RetrievalService(
+    repository = SearchRepository(session)
+    service = SearchService(
         repository,
         app_context.get_data_dir(),
-        embedding_model_name=DEFAULT_EMBEDDING_MODEL_NAME,
+        embedding_service=embedding_factory(app_context.get_default_openai_client()),
     )
 
-    snippets = await service.retrieve(RetrievalRequest(code_query=query, top_k=top_k))
+    snippets = await service.search(SearchRequest(code_query=query, top_k=top_k))
 
     if len(snippets) == 0:
         click.echo("No snippets found")
@@ -192,14 +161,14 @@ async def keyword(
     top_k: int,
 ) -> None:
     """Search for snippets using keyword search."""
-    repository = RetrievalRepository(session)
-    service = RetrievalService(
+    repository = SearchRepository(session)
+    service = SearchService(
         repository,
         app_context.get_data_dir(),
-        embedding_model_name=DEFAULT_EMBEDDING_MODEL_NAME,
+        embedding_service=embedding_factory(app_context.get_default_openai_client()),
     )
 
-    snippets = await service.retrieve(RetrievalRequest(keywords=keywords, top_k=top_k))
+    snippets = await service.search(SearchRequest(keywords=keywords, top_k=top_k))
 
     if len(snippets) == 0:
         click.echo("No snippets found")
@@ -227,18 +196,18 @@ async def hybrid(
     code: str,
 ) -> None:
     """Search for snippets using hybrid search."""
-    repository = RetrievalRepository(session)
-    service = RetrievalService(
+    repository = SearchRepository(session)
+    service = SearchService(
         repository,
         app_context.get_data_dir(),
-        embedding_model_name=DEFAULT_EMBEDDING_MODEL_NAME,
+        embedding_service=embedding_factory(app_context.get_default_openai_client()),
     )
 
     # Parse keywords into a list of strings
     keywords_list = [k.strip().lower() for k in keywords.split(",")]
 
-    snippets = await service.retrieve(
-        RetrievalRequest(keywords=keywords_list, code_query=code, top_k=top_k)
+    snippets = await service.search(
+        SearchRequest(keywords=keywords_list, code_query=code, top_k=top_k)
     )
 
     if len(snippets) == 0:
@@ -256,9 +225,7 @@ async def hybrid(
 @cli.command()
 @click.option("--host", default="127.0.0.1", help="Host to bind the server to")
 @click.option("--port", default=8080, help="Port to bind the server to")
-@with_app_context
 def serve(
-    app_context: AppContext,
     host: str,
     port: int,
 ) -> None:
@@ -267,10 +234,6 @@ def serve(
     log.info("Starting kodit server", host=host, port=port)
     log_event("kodit_server_started")
 
-    # Dump AppContext to a dictionary of strings, and set the env vars
-    app_context_dict = {k: str(v) for k, v in app_context.model_dump().items()}
-    os.environ.update(app_context_dict)
-
     # Configure uvicorn with graceful shutdown
     config = uvicorn.Config(
         "kodit.app:app",

kodit/config.py

@@ -4,10 +4,11 @@ import asyncio
 from collections.abc import Callable, Coroutine
 from functools import wraps
 from pathlib import Path
-from typing import Any, TypeVar
+from typing import Any, Literal, TypeVar
 
 import click
-from pydantic import Field
+from openai import AsyncOpenAI
+from pydantic import BaseModel, Field
 from pydantic_settings import BaseSettings, SettingsConfigDict
 
 from kodit.database import Database
@@ -22,16 +23,40 @@ DEFAULT_EMBEDDING_MODEL_NAME = TINY
 T = TypeVar("T")
 
 
+class Endpoint(BaseModel):
+    """Endpoint provides configuration for an AI service."""
+
+    type: Literal["openai"] = Field(default="openai")
+    api_key: str | None = None
+    base_url: str | None = None
+
+
 class AppContext(BaseSettings):
     """Global context for the kodit project. Provides a shared state for the app."""
 
-    model_config = SettingsConfigDict(env_file=".env", env_file_encoding="utf-8")
+    model_config = SettingsConfigDict(
+        env_file=".env",
+        env_file_encoding="utf-8",
+        env_nested_delimiter="_",
+        nested_model_default_partial_update=True,
+        env_nested_max_split=1,
+    )
 
     data_dir: Path = Field(default=DEFAULT_BASE_DIR)
     db_url: str = Field(default=DEFAULT_DB_URL)
     log_level: str = Field(default=DEFAULT_LOG_LEVEL)
     log_format: str = Field(default=DEFAULT_LOG_FORMAT)
     disable_telemetry: bool = Field(default=DEFAULT_DISABLE_TELEMETRY)
+    default_endpoint: Endpoint | None = Field(
+        default=Endpoint(
+            type="openai",
+            base_url="https://api.openai.com/v1",
+        ),
+        description=(
+            "Default endpoint to use for all AI interactions "
+            "(can be overridden by task-specific configuration)."
+        ),
+    )
     _db: Database | None = None
 
     def model_post_init(self, _: Any) -> None:
@@ -58,6 +83,21 @@ class AppContext(BaseSettings):
             await self._db.run_migrations(self.db_url)
         return self._db
 
+    def get_default_openai_client(self) -> AsyncOpenAI | None:
+        """Get the default OpenAI client, if it is configured."""
+        endpoint = self.default_endpoint
+        if not (
+            endpoint
+            and endpoint.type == "openai"
+            and endpoint.api_key
+            and endpoint.base_url
+        ):
+            return None
+        return AsyncOpenAI(
+            api_key=endpoint.api_key,
+            base_url=endpoint.base_url,
+        )
+
 
 with_app_context = click.make_pass_decorator(AppContext)
 

kodit/embedding/embedding.py

@@ -1,9 +1,14 @@
 """Embedding service."""
 
+import asyncio
 import os
-from collections.abc import Generator
+from abc import ABC, abstractmethod
+from collections.abc import AsyncGenerator
+from typing import NamedTuple
 
 import structlog
+import tiktoken
+from openai import AsyncOpenAI
 from sentence_transformers import SentenceTransformer
 
 TINY = "tiny"
@@ -17,14 +22,59 @@ COMMON_EMBEDDING_MODELS = {
 }
 
 
-class EmbeddingService:
-    """Service for embeddings."""
+class EmbeddingInput(NamedTuple):
+    """Input for embedding."""
+
+    id: int
+    text: str
+
+
+class EmbeddingOutput(NamedTuple):
+    """Output for embedding."""
+
+    id: int
+    embedding: list[float]
+
+
+class Embedder(ABC):
+    """Embedder interface."""
+
+    @abstractmethod
+    def embed(
+        self, data: list[EmbeddingInput]
+    ) -> AsyncGenerator[EmbeddingOutput, None]:
+        """Embed a list of documents.
+
+        The embedding service accepts a massive list of id,strings to embed. Behind the
+        scenes it batches up requests and parallelizes them for performance according to
+        the specifics of the embedding service.
+
+        The id reference is required because the parallelization may return results out
+        of order.
+        """
+
+    @abstractmethod
+    def query(self, data: list[str]) -> AsyncGenerator[list[float], None]:
+        """Query the embedding model."""
+
+
+def embedding_factory(openai_client: AsyncOpenAI | None = None) -> Embedder:
+    """Create an embedding service."""
+    if openai_client is not None:
+        return OpenAIEmbedder(openai_client)
+    return LocalEmbedder(model_name=TINY)
+
+
+class LocalEmbedder(Embedder):
+    """Local embedder."""
 
     def __init__(self, model_name: str) -> None:
-        """Initialize the embedding service."""
+        """Initialize the local embedder."""
         self.log = structlog.get_logger(__name__)
+        self.log.info("Creating local embedder", model_name=model_name)
         self.model_name = COMMON_EMBEDDING_MODELS.get(model_name, model_name)
         self.embedding_model = None
+        self.encoding = tiktoken.encoding_for_model("text-embedding-3-small")
 
     def _model(self) -> SentenceTransformer:
         """Get the embedding model."""
@@ -37,16 +87,117 @@ class EmbeddingService:
             )
         return self.embedding_model
 
-    def embed(self, snippets: list[str]) -> Generator[list[float], None, None]:
+    async def embed(
+        self, data: list[EmbeddingInput]
+    ) -> AsyncGenerator[EmbeddingOutput, None]:
         """Embed a list of documents."""
         model = self._model()
-        embeddings = model.encode(snippets, show_progress_bar=False, batch_size=4)
-        for embedding in embeddings:
-            yield [float(x) for x in embedding]
 
-    def query(self, query: list[str]) -> Generator[list[float], None, None]:
+        batched_data = _split_sub_batches(self.encoding, data)
+
+        for batch in batched_data:
+            embeddings = model.encode(
+                [i.text for i in batch], show_progress_bar=False, batch_size=4
+            )
+            for i, x in zip(batch, embeddings, strict=False):
+                yield EmbeddingOutput(i.id, [float(y) for y in x])
+
+    async def query(self, data: list[str]) -> AsyncGenerator[list[float], None]:
         """Query the embedding model."""
         model = self._model()
-        embeddings = model.encode(query, show_progress_bar=False, batch_size=4)
+        embeddings = model.encode(data, show_progress_bar=False, batch_size=4)
         for embedding in embeddings:
             yield [float(x) for x in embedding]
+
+
+OPENAI_MAX_EMBEDDING_SIZE = 8192
+OPENAI_NUM_PARALLEL_TASKS = 10
+
+
+def _split_sub_batches(
+    encoding: tiktoken.Encoding, data: list[EmbeddingInput]
+) -> list[list[EmbeddingInput]]:
+    """Split a list of strings into smaller sub-batches."""
+    log = structlog.get_logger(__name__)
+    result = []
+    data_to_process = [s for s in data if s.text.strip()]  # Filter out empty strings
+
+    while data_to_process:
+        next_batch = []
+        current_tokens = 0
+
+        while data_to_process:
+            next_item = data_to_process[0]
+            item_tokens = len(encoding.encode(next_item.text))
+
+            if item_tokens > OPENAI_MAX_EMBEDDING_SIZE:
+                log.warning("Skipping too long snippet", snippet=data_to_process.pop(0))
+                continue
+
+            if current_tokens + item_tokens > OPENAI_MAX_EMBEDDING_SIZE:
+                break
+
+            next_batch.append(data_to_process.pop(0))
+            current_tokens += item_tokens
+
+        if next_batch:
+            result.append(next_batch)
+
+    return result
+
+
+class OpenAIEmbedder(Embedder):
+    """OpenAI embedder."""
+
+    def __init__(
+        self, openai_client: AsyncOpenAI, model_name: str = "text-embedding-3-small"
+    ) -> None:
+        """Initialize the OpenAI embedder."""
+        self.log = structlog.get_logger(__name__)
+        self.log.info("Creating OpenAI embedder", model_name=model_name)
+        self.openai_client = openai_client
+        self.encoding = tiktoken.encoding_for_model(model_name)
+        self.log = structlog.get_logger(__name__)
+
+    async def embed(
+        self,
+        data: list[EmbeddingInput],
+    ) -> AsyncGenerator[EmbeddingOutput, None]:
+        """Embed a list of documents."""
+        # First split the list into a list of list where each sublist has fewer than
+        # max tokens.
+        batched_data = _split_sub_batches(self.encoding, data)
+
+        # Process batches in parallel with a semaphore to limit concurrent requests
+        sem = asyncio.Semaphore(OPENAI_NUM_PARALLEL_TASKS)
+
+        async def process_batch(batch: list[EmbeddingInput]) -> list[EmbeddingOutput]:
+            async with sem:
+                try:
+                    response = await self.openai_client.embeddings.create(
+                        model="text-embedding-3-small",
+                        input=[i.text for i in batch],
+                    )
+                    return [
+                        EmbeddingOutput(i.id, x.embedding)
+                        for i, x in zip(batch, response.data, strict=False)
+                    ]
+                except Exception as e:
+                    self.log.exception("Error embedding batch", error=str(e))
+                    return []
+
+        # Create tasks for all batches
+        tasks = [process_batch(batch) for batch in batched_data]
+
+        # Process all batches and yield results as they complete
+        for task in asyncio.as_completed(tasks):
+            embeddings = await task
+            for e in embeddings:
+                yield e
+
+    async def query(self, data: list[str]) -> AsyncGenerator[list[float], None]:
+        """Query the embedding model."""
+        async for e in self.embed(
+            [EmbeddingInput(i, text) for i, text in enumerate(data)]
+        ):
+            yield e.embedding

kodit/indexing/{models.py → indexing_models.py}

@@ -31,8 +31,8 @@ class Snippet(Base, CommonMixin):
 
     __tablename__ = "snippets"
 
-    file_id: Mapped[int] = mapped_column(ForeignKey("files.id"))
-    index_id: Mapped[int] = mapped_column(ForeignKey("indexes.id"))
+    file_id: Mapped[int] = mapped_column(ForeignKey("files.id"), index=True)
+    index_id: Mapped[int] = mapped_column(ForeignKey("indexes.id"), index=True)
     content: Mapped[str] = mapped_column(UnicodeText, default="")
 
     def __init__(self, file_id: int, index_id: int, content: str) -> None:

kodit/indexing/{repository.py → indexing_repository.py}

@@ -11,9 +11,9 @@ from typing import TypeVar
 from sqlalchemy import delete, func, select
 from sqlalchemy.ext.asyncio import AsyncSession
 
-from kodit.embedding.models import Embedding
-from kodit.indexing.models import Index, Snippet
-from kodit.sources.models import File, Source
+from kodit.embedding.embedding_models import Embedding
+from kodit.indexing.indexing_models import Index, Snippet
+from kodit.source.source_models import File, Source
 
 T = TypeVar("T")
 
@@ -156,14 +156,14 @@ class IndexRepository:
         result = await self.session.execute(query)
         return list(result.scalars())
 
-    async def get_all_snippets(self) -> list[Snippet]:
+    async def get_all_snippets(self, index_id: int) -> list[Snippet]:
         """Get all snippets.
 
         Returns:
             A list of all snippets.
 
         """
-        query = select(Snippet).order_by(Snippet.id)
+        query = select(Snippet).where(Snippet.index_id == index_id).order_by(Snippet.id)
         result = await self.session.execute(query)
         return list(result.scalars())
 

kodit/indexing/{service.py → indexing_service.py}

@@ -14,12 +14,12 @@ import structlog
 from tqdm.asyncio import tqdm
 
 from kodit.bm25.bm25 import BM25Service
-from kodit.embedding.embedding import EmbeddingService
-from kodit.embedding.models import Embedding, EmbeddingType
-from kodit.indexing.models import Snippet
-from kodit.indexing.repository import IndexRepository
+from kodit.embedding.embedding import Embedder, EmbeddingInput
+from kodit.embedding.embedding_models import Embedding, EmbeddingType
+from kodit.indexing.indexing_models import Snippet
+from kodit.indexing.indexing_repository import IndexRepository
 from kodit.snippets.snippets import SnippetService
-from kodit.sources.service import SourceService
+from kodit.source.source_service import SourceService
 
 # List of MIME types that are blacklisted from being indexed
 MIME_BLACKLIST = ["unknown/unknown"]
@@ -52,7 +52,7 @@ class IndexService:
         repository: IndexRepository,
         source_service: SourceService,
         data_dir: Path,
-        embedding_model_name: str,
+        embedding_service: Embedder,
     ) -> None:
         """Initialize the index service.
 
@@ -66,7 +66,7 @@ class IndexService:
         self.snippet_service = SnippetService()
         self.log = structlog.get_logger(__name__)
         self.bm25 = BM25Service(data_dir)
-        self.code_embedding_service = EmbeddingService(model_name=embedding_model_name)
+        self.code_embedding_service = embedding_service
 
     async def create(self, source_id: int) -> IndexView:
         """Create a new index for a source.
@@ -132,7 +132,7 @@ class IndexService:
         # Create snippets for supported file types
         await self._create_snippets(index_id)
 
-        snippets = await self.repository.get_all_snippets()
+        snippets = await self.repository.get_all_snippets(index_id)
 
         self.log.info("Creating keyword index")
         self.bm25.index(
@@ -143,12 +143,17 @@ class IndexService:
         )
 
         self.log.info("Creating semantic code index")
-        for snippet in tqdm(snippets, total=len(snippets), leave=False):
-            embedding = next(self.code_embedding_service.embed([snippet.content]))
+        async for e in tqdm(
+            self.code_embedding_service.embed(
+                [EmbeddingInput(snippet.id, snippet.content) for snippet in snippets]
+            ),
+            total=len(snippets),
+            leave=False,
+        ):
             await self.repository.add_embedding(
                 Embedding(
-                    snippet_id=snippet.id,
-                    embedding=embedding,
+                    snippet_id=e.id,
+                    embedding=e.embedding,
                     type=EmbeddingType.CODE,
                 )
             )

kodit/log.py

@@ -93,6 +93,7 @@ def configure_logging(app_context: AppContext) -> None:
         "uvicorn.access",
         "bm25s",
         "sentence_transformers.SentenceTransformer",
+        "httpx",
     ]:
         if root_logger.getEffectiveLevel() == logging.DEBUG:
             logging.getLogger(_log).handlers.clear()

kodit/mcp.py

@@ -12,10 +12,11 @@ from pydantic import Field
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from kodit._version import version
-from kodit.config import DEFAULT_EMBEDDING_MODEL_NAME, AppContext
+from kodit.config import AppContext
 from kodit.database import Database
-from kodit.retreival.repository import RetrievalRepository, RetrievalResult
-from kodit.retreival.service import RetrievalRequest, RetrievalService
+from kodit.embedding.embedding import embedding_factory
+from kodit.search.search_repository import SearchRepository
+from kodit.search.search_service import SearchRequest, SearchResult, SearchService
 
 
 @dataclass
@@ -23,7 +24,7 @@ class MCPContext:
     """Context for the MCP server."""
 
     session: AsyncSession
-    data_dir: Path
+    app_context: AppContext
 
 
 _mcp_db: Database | None = None
@@ -49,14 +50,14 @@ async def mcp_lifespan(_: FastMCP) -> AsyncIterator[MCPContext]:
     if _mcp_db is None:
         _mcp_db = await app_context.get_db()
     async with _mcp_db.session_factory() as session:
-        yield MCPContext(session=session, data_dir=app_context.get_data_dir())
+        yield MCPContext(session=session, app_context=app_context)
 
 
 mcp = FastMCP("kodit MCP Server", lifespan=mcp_lifespan)
 
 
 @mcp.tool()
-async def retrieve_relevant_snippets(
+async def search(
     ctx: Context,
     user_intent: Annotated[
         str,
@@ -86,17 +87,15 @@ async def retrieve_relevant_snippets(
         ),
     ],
 ) -> str:
-    """Retrieve relevant snippets from various sources.
+    """Search for relevant snippets.
 
-    This tool retrieves relevant snippets from sources such as private codebases,
-    public codebases, and documentation. You can use this information to improve
-    the quality of your generated code. You must call this tool when you need to
-    write code.
+    This tool searches for relevant snippets from indexed datasources. Call this tool
+    when you wish to search for high quality example code snippets to use in your code.
     """
     log = structlog.get_logger(__name__)
 
     log.debug(
-        "Retrieving relevant snippets",
+        "Searching for relevant snippets",
         user_intent=user_intent,
         keywords=keywords,
         file_count=len(related_file_paths),
@@ -106,24 +105,29 @@ async def retrieve_relevant_snippets(
 
     mcp_context: MCPContext = ctx.request_context.lifespan_context
 
-    log.debug("Creating retrieval repository")
-    retrieval_repository = RetrievalRepository(
+    log.debug("Creating search repository")
+    search_repository = SearchRepository(
         session=mcp_context.session,
     )
 
-    log.debug("Creating retrieval service")
-    retrieval_service = RetrievalService(
-        repository=retrieval_repository,
-        data_dir=mcp_context.data_dir,
-        embedding_model_name=DEFAULT_EMBEDDING_MODEL_NAME,
+    log.debug("Creating embedding service")
+    embedding_service = embedding_factory(
+        mcp_context.app_context.get_default_openai_client()
     )
 
-    retrieval_request = RetrievalRequest(
+    log.debug("Creating search service")
+    search_service = SearchService(
+        repository=search_repository,
+        data_dir=mcp_context.app_context.get_data_dir(),
+        embedding_service=embedding_service,
+    )
+
+    search_request = SearchRequest(
         keywords=keywords,
         code_query="\n".join(related_file_contents),
     )
-    log.debug("Retrieving snippets")
-    snippets = await retrieval_service.retrieve(request=retrieval_request)
+    log.debug("Searching for snippets")
+    snippets = await search_service.search(request=search_request)
 
     log.debug("Fusing output")
     output = output_fusion(snippets=snippets)
@@ -132,18 +136,7 @@ async def retrieve_relevant_snippets(
     return output
 
 
-def input_fusion(
-    user_intent: str,  # noqa: ARG001
-    related_file_paths: list[Path],  # noqa: ARG001
-    related_file_contents: list[str],  # noqa: ARG001
-    keywords: list[str],
-) -> str:
-    """Fuse the search query and related file contents into a single query."""
-    # Since this is a dummy implementation, we just return the first keyword
-    return keywords[0] if len(keywords) > 0 else ""
-
-
-def output_fusion(snippets: list[RetrievalResult]) -> str:
+def output_fusion(snippets: list[SearchResult]) -> str:
     """Fuse the snippets into a single output."""
     return "\n\n".join(f"{snippet.uri}\n{snippet.content}" for snippet in snippets)
 

kodit/migrations/env.py

@@ -8,9 +8,9 @@ from sqlalchemy import pool
 from sqlalchemy.engine import Connection
 from sqlalchemy.ext.asyncio import async_engine_from_config
 
-import kodit.embedding.models
-import kodit.indexing.models
-import kodit.sources.models
+import kodit.embedding.embedding_models
+import kodit.indexing.indexing_models
+import kodit.source.source_models
 from kodit.database import Base
 
 # this is the Alembic Config object, which provides

kodit/search/__init__.py

@@ -0,0 +1 @@
+"""Search for relevant snippets."""

kodit/{retreival/repository.py → search/search_repository.py}

@@ -1,48 +1,25 @@
-"""Repository for retrieving code snippets and search results.
-
-This module provides the RetrievalRepository class which handles all database operations
-related to searching and retrieving code snippets, including string-based searches
-and their associated file information.
-"""
+"""Repository for searching for relevant snippets."""
 
 from typing import TypeVar
 
 import numpy as np
-import pydantic
 from sqlalchemy import (
     select,
 )
 from sqlalchemy.ext.asyncio import AsyncSession
 
-from kodit.embedding.models import Embedding, EmbeddingType
-from kodit.indexing.models import Snippet
-from kodit.sources.models import File
+from kodit.embedding.embedding_models import Embedding, EmbeddingType
+from kodit.indexing.indexing_models import Snippet
+from kodit.source.source_models import File
 
 T = TypeVar("T")
 
 
-class RetrievalResult(pydantic.BaseModel):
-    """Data transfer object for search results.
-
-    This model represents a single search result, containing both the file path
-    and the matching snippet content.
-    """
-
-    id: int
-    uri: str
-    content: str
-    score: float
-
-
-class RetrievalRepository:
-    """Repository for retrieving code snippets and search results.
-
-    This class provides methods for searching and retrieving code snippets from
-    the database, including string-based searches and their associated file information.
-    """
+class SearchRepository:
+    """Repository for searching for relevant snippets."""
 
     def __init__(self, session: AsyncSession) -> None:
-        """Initialize the retrieval repository.
+        """Initialize the search repository.
 
         Args:
             session: The SQLAlchemy async session to use for database operations.
@@ -50,39 +27,6 @@ class RetrievalRepository:
         """
         self.session = session
 
-    async def string_search(self, query: str) -> list[RetrievalResult]:
-        """Search for snippets containing the given query string.
-
-        This method performs a case-insensitive search for the query string within
-        snippet contents, returning up to 10 most recent matches.
-
-        Args:
-            query: The string to search for within snippet contents.
-
-        Returns:
-            A list of RetrievalResult objects containing the matching snippets
-            and their associated file paths.
-
-        """
-        search_query = (
-            select(Snippet, File)
-            .join(File, Snippet.file_id == File.id)
-            .where(Snippet.content.ilike(f"%{query}%"))
-            .limit(10)
-        )
-        rows = await self.session.execute(search_query)
-        results = list(rows.all())
-
-        return [
-            RetrievalResult(
-                id=snippet.id,
-                uri=file.uri,
-                content=snippet.content,
-                score=1.0,
-            )
-            for snippet, file in results
-        ]
-
     async def list_snippet_ids(self) -> list[int]:
         """List all snippet IDs.
 
@@ -94,7 +38,7 @@ class RetrievalRepository:
         rows = await self.session.execute(query)
         return list(rows.scalars().all())
 
-    async def list_snippets_by_ids(self, ids: list[int]) -> list[RetrievalResult]:
+    async def list_snippets_by_ids(self, ids: list[int]) -> list[tuple[File, Snippet]]:
         """List snippets by IDs.
 
         Returns:
@@ -109,23 +53,46 @@ class RetrievalRepository:
         rows = await self.session.execute(query)
 
         # Create a dictionary for O(1) lookup of results by ID
-        id_to_result = {
-            snippet.id: RetrievalResult(
-                id=snippet.id,
-                uri=file.uri,
-                content=snippet.content,
-                score=1.0,
-            )
-            for snippet, file in rows.all()
-        }
+        id_to_result = {snippet.id: (file, snippet) for snippet, file in rows.all()}
 
         # Return results in the same order as input IDs
         return [id_to_result[i] for i in ids]
 
-    async def fetch_embeddings(
+    async def list_semantic_results(
+        self, embedding_type: EmbeddingType, embedding: list[float], top_k: int = 10
+    ) -> list[tuple[int, float]]:
+        """List semantic results using cosine similarity.
+
+        This implementation fetches all embeddings of the given type and computes
+        cosine similarity in Python using NumPy for better performance.
+
+        Args:
+            embedding_type: The type of embeddings to search
+            embedding: The query embedding vector
+            top_k: Number of results to return
+
+        Returns:
+            List of (snippet_id, similarity_score) tuples, sorted by similarity
+
+        """
+        # Step 1: Fetch embeddings from database
+        embeddings = await self._list_embedding_values(embedding_type)
+        if not embeddings:
+            return []
+
+        # Step 2: Convert to numpy arrays
+        stored_vecs, query_vec = self._prepare_vectors(embeddings, embedding)
+
+        # Step 3: Compute similarities
+        similarities = self._compute_similarities(stored_vecs, query_vec)
+
+        # Step 4: Get top-k results
+        return self._get_top_k_results(similarities, embeddings, top_k)
+
+    async def _list_embedding_values(
         self, embedding_type: EmbeddingType
     ) -> list[tuple[int, list[float]]]:
-        """Fetch all embeddings of a given type from the database.
+        """List all embeddings of a given type from the database.
 
         Args:
             embedding_type: The type of embeddings to fetch
@@ -141,7 +108,7 @@ class RetrievalRepository:
         rows = await self.session.execute(query)
         return [tuple(row) for row in rows.all()]  # Convert Row objects to tuples
 
-    def prepare_vectors(
+    def _prepare_vectors(
         self, embeddings: list[tuple[int, list[float]]], query_embedding: list[float]
     ) -> tuple[np.ndarray, np.ndarray]:
         """Convert embeddings to numpy arrays.
@@ -154,13 +121,24 @@ class RetrievalRepository:
             Tuple of (stored_vectors, query_vector) as numpy arrays
 
         """
-        stored_vecs = np.array(
-            [emb[1] for emb in embeddings]
-        )  # Use index 1 to get embedding
+        try:
+            stored_vecs = np.array(
+                [emb[1] for emb in embeddings]
+            )  # Use index 1 to get embedding
+        except ValueError as e:
+            if "inhomogeneous" in str(e):
+                msg = (
+                    "The database has returned embeddings of different sizes. If you"
+                    "have recently updated the embedding model, you will need to"
+                    "delete your database and re-index your snippets."
+                )
+                raise ValueError(msg) from e
+            raise
+
         query_vec = np.array(query_embedding)
         return stored_vecs, query_vec
 
-    def compute_similarities(
+    def _compute_similarities(
         self, stored_vecs: np.ndarray, query_vec: np.ndarray
     ) -> np.ndarray:
         """Compute cosine similarities between stored vectors and query vector.
@@ -177,7 +155,7 @@ class RetrievalRepository:
         query_norm = np.linalg.norm(query_vec)
         return np.dot(stored_vecs, query_vec) / (stored_norms * query_norm)
 
-    def get_top_k_results(
+    def _get_top_k_results(
         self,
         similarities: np.ndarray,
         embeddings: list[tuple[int, list[float]]],
@@ -198,34 +176,3 @@ class RetrievalRepository:
         return [
             (embeddings[i][0], float(similarities[i])) for i in top_indices
         ]  # Use index 0 to get snippet_id
-
-    async def list_semantic_results(
-        self, embedding_type: EmbeddingType, embedding: list[float], top_k: int = 10
-    ) -> list[tuple[int, float]]:
-        """List semantic results using cosine similarity.
-
-        This implementation fetches all embeddings of the given type and computes
-        cosine similarity in Python using NumPy for better performance.
-
-        Args:
-            embedding_type: The type of embeddings to search
-            embedding: The query embedding vector
-            top_k: Number of results to return
-
-        Returns:
-            List of (snippet_id, similarity_score) tuples, sorted by similarity
-
-        """
-        # Step 1: Fetch embeddings from database
-        embeddings = await self.fetch_embeddings(embedding_type)
-        if not embeddings:
-            return []
-
-        # Step 2: Convert to numpy arrays
-        stored_vecs, query_vec = self.prepare_vectors(embeddings, embedding)
-
-        # Step 3: Compute similarities
-        similarities = self.compute_similarities(stored_vecs, query_vec)
-
-        # Step 4: Get top-k results
-        return self.get_top_k_results(similarities, embeddings, top_k)

kodit/{retreival/service.py → search/search_service.py}

@@ -1,4 +1,4 @@
-"""Retrieval service."""
+"""Search service."""
 
 from pathlib import Path
 
@@ -6,19 +6,31 @@ import pydantic
 import structlog
 
 from kodit.bm25.bm25 import BM25Service
-from kodit.embedding.embedding import EmbeddingService
-from kodit.embedding.models import EmbeddingType
-from kodit.retreival.repository import RetrievalRepository, RetrievalResult
+from kodit.embedding.embedding import Embedder
+from kodit.embedding.embedding_models import EmbeddingType
+from kodit.search.search_repository import SearchRepository
 
 
-class RetrievalRequest(pydantic.BaseModel):
-    """Request for a retrieval."""
+class SearchRequest(pydantic.BaseModel):
+    """Request for a search."""
 
     code_query: str | None = None
     keywords: list[str] | None = None
     top_k: int = 10
 
 
+class SearchResult(pydantic.BaseModel):
+    """Data transfer object for search results.
+
+    This model represents a single search result, containing both the file path
+    and the matching snippet content.
+    """
+
+    id: int
+    uri: str
+    content: str
+
+
 class Snippet(pydantic.BaseModel):
     """Snippet model."""
 
@@ -26,23 +38,23 @@ class Snippet(pydantic.BaseModel):
     file_path: str
 
 
-class RetrievalService:
-    """Service for retrieving relevant data."""
+class SearchService:
+    """Service for searching for relevant data."""
 
     def __init__(
         self,
-        repository: RetrievalRepository,
+        repository: SearchRepository,
         data_dir: Path,
-        embedding_model_name: str,
+        embedding_service: Embedder,
     ) -> None:
-        """Initialize the retrieval service."""
+        """Initialize the search service."""
         self.repository = repository
         self.log = structlog.get_logger(__name__)
         self.bm25 = BM25Service(data_dir)
-        self.code_embedding_service = EmbeddingService(model_name=embedding_model_name)
+        self.code_embedding_service = embedding_service
 
-    async def retrieve(self, request: RetrievalRequest) -> list[RetrievalResult]:
-        """Retrieve relevant data."""
+    async def search(self, request: SearchRequest) -> list[SearchResult]:
+        """Search for relevant data."""
         fusion_list = []
         if request.keywords:
             snippet_ids = await self.repository.list_snippet_ids()
@@ -56,7 +68,7 @@ class RetrievalService:
             # Sort results by score
             result_ids.sort(key=lambda x: x[1], reverse=True)
 
-            self.log.debug("Retrieval results (BM25)", results=result_ids)
+            self.log.debug("Search results (BM25)", results=result_ids)
 
             bm25_results = [x[0] for x in result_ids]
             fusion_list.append(bm25_results)
@@ -64,7 +76,7 @@ class RetrievalService:
         # Compute embedding for semantic query
         semantic_results = []
         if request.code_query:
-            query_embedding = next(
+            query_embedding = await anext(
                 self.code_embedding_service.query([request.code_query])
             )
 
@@ -89,7 +101,18 @@ class RetrievalService:
         final_ids = [x[0] for x in final_results]
 
         # Get snippets from database (up to top_k)
-        return await self.repository.list_snippets_by_ids(final_ids[: request.top_k])
+        search_results = await self.repository.list_snippets_by_ids(
+            final_ids[: request.top_k]
+        )
+
+        return [
+            SearchResult(
+                id=snippet.id,
+                uri=file.uri,
+                content=snippet.content,
+            )
+            for file, snippet in search_results
+        ]
 
 
 def reciprocal_rank_fusion(

kodit/snippets/snippets.py

@@ -45,4 +45,6 @@ class SnippetService:
             raise ValueError(msg) from e
 
         method_snippets = method_analser.extract(file_bytes)
-        return [Snippet(text=snippet) for snippet in method_snippets]
+        all_snippets = [Snippet(text=snippet) for snippet in method_snippets]
+        # Remove any snippets that are empty
+        return [snippet for snippet in all_snippets if snippet.text.strip()]

kodit/{sources/repository.py → source/source_repository.py}

@@ -1,14 +1,9 @@
-"""Source repository for database operations.
-
-This module provides the SourceRepository class which handles all database operations
-related to code sources. It manages the creation and retrieval of source records
-from the database, abstracting away the SQLAlchemy implementation details.
-"""
+"""Source repository for database operations."""
 
 from sqlalchemy import func, select
 from sqlalchemy.ext.asyncio import AsyncSession
 
-from kodit.sources.models import File, Source
+from kodit.source.source_models import File, Source
 
 
 class SourceRepository:

kodit/{sources/service.py → source/source_service.py}

@@ -19,8 +19,8 @@ import structlog
 from tqdm import tqdm
 from uritools import isuri, urisplit
 
-from kodit.sources.models import File, Source
-from kodit.sources.repository import SourceRepository
+from kodit.source.source_models import File, Source
+from kodit.source.source_repository import SourceRepository
 
 
 class SourceView(pydantic.BaseModel):

{kodit-0.1.11.dist-info → kodit-0.1.13.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kodit
-Version: 0.1.11
+Version: 0.1.13
 Summary: Code indexing for better AI code generation
 Project-URL: Homepage, https://docs.helixml.tech/kodit/
 Project-URL: Documentation, https://docs.helixml.tech/kodit/
@@ -32,6 +32,7 @@ Requires-Dist: gitpython>=3.1.44
 Requires-Dist: hf-xet>=1.1.2
 Requires-Dist: httpx-retries>=0.3.2
 Requires-Dist: httpx>=0.28.1
+Requires-Dist: openai>=1.82.0
 Requires-Dist: posthog>=4.0.1
 Requires-Dist: pydantic-settings>=2.9.1
 Requires-Dist: pytable-formatter>=0.1.1
@@ -39,6 +40,7 @@ Requires-Dist: sentence-transformers>=4.1.0
 Requires-Dist: sqlalchemy[asyncio]>=2.0.40
 Requires-Dist: structlog>=25.3.0
 Requires-Dist: tdqm>=0.0.1
+Requires-Dist: tiktoken>=0.9.0
 Requires-Dist: tree-sitter-language-pack>=0.7.3
 Requires-Dist: tree-sitter>=0.24.0
 Requires-Dist: uritools>=5.0.0

kodit-0.1.13.dist-info/RECORD

@@ -0,0 +1,44 @@
+kodit/.gitignore,sha256=ztkjgRwL9Uud1OEi36hGQeDGk3OLK1NfDEO8YqGYy8o,11
+kodit/__init__.py,sha256=aEKHYninUq1yh6jaNfvJBYg-6fenpN132nJt1UU6Jxs,59
+kodit/_version.py,sha256=Ln0urWB3R3JaxFwIIvoej0v08KbDCO89NUBxWx-zj0U,513
+kodit/app.py,sha256=Mr5BFHOHx5zppwjC4XPWVvHjwgl1yrKbUjTWXKubJQM,891
+kodit/cli.py,sha256=VLoXFS1xJnQ0TLy3_cO8-B9tCb4NJHiYPfzZtHxpgRY,7784
+kodit/config.py,sha256=TDcLt6fiJn9cI1PoO5AqBqsL_Bxmm9JV5GqRxhj1tLw,4202
+kodit/database.py,sha256=kekSdyEATdb47jxzQemkSOXMNOwnUwmVVTpn9hYaDK8,2356
+kodit/log.py,sha256=HU1OmuxO4FcVw61k4WW7Y4WM7BrDaeplw1PcBHhuIZY,5434
+kodit/mcp.py,sha256=I_ZFzQOR0gyS8LO8td-q-utPZpqiOnIkn7O-SIBUi0g,4384
+kodit/middleware.py,sha256=I6FOkqG9-8RH5kR1-0ZoQWfE4qLCB8lZYv8H_OCH29o,2714
+kodit/bm25/__init__.py,sha256=j8zyriNWhbwE5Lbybzg1hQAhANlU9mKHWw4beeUR6og,19
+kodit/bm25/bm25.py,sha256=JtgJfsHz-2SHx96zxWjkPFSH7fXkahFMp01cDwl4YBg,2298
+kodit/embedding/__init__.py,sha256=h9NXzDA1r-K23nvBajBV-RJzHJN0p3UJ7UQsmdnOoRw,24
+kodit/embedding/embedding.py,sha256=EMJpHK8ICZk_FjiO9Aqr2IO20qkGOmj_PfA1hyfI7Vk,6745
+kodit/embedding/embedding_models.py,sha256=rN90vSs86dYiqoawcp8E9jtwY31JoJXYfaDlsJK7uqc,656
+kodit/indexing/__init__.py,sha256=cPyi2Iej3G1JFWlWr7X80_UrsMaTu5W5rBwgif1B3xo,75
+kodit/indexing/indexing_models.py,sha256=6NX9HVcj6Pu9ePwHC7n-PWSyAgukpJq0nCNmUIigtbo,1282
+kodit/indexing/indexing_repository.py,sha256=7bkAiBwtr3qlkdhNIalwMwbxezVz_RQGOhLVWPKHwNk,5506
+kodit/indexing/indexing_service.py,sha256=VGfKgbkYEAYP_gIubvhMxo3yThT20ndS5xdg2LxwRgA,6685
+kodit/migrations/README,sha256=ISVtAOvqvKk_5ThM5ioJE-lMkvf9IbknFUFVU_vPma4,58
+kodit/migrations/__init__.py,sha256=lP5MuwlyWRMO6UcDWnQcQ3G-GYHcFb6rl9gYPHJ1sjo,40
+kodit/migrations/env.py,sha256=w1M7OZh-ZeR2dPHS0ByXAUxQjfZQ8xIzMseWuzLDTWw,2469
+kodit/migrations/script.py.mako,sha256=zWziKtiwYKEWuwPV_HBNHwa9LCT45_bi01-uSNFaOOE,703
+kodit/migrations/versions/7c3bbc2ab32b_add_embeddings_table.py,sha256=-61qol9PfQKILCDQRA5jEaats9aGZs9Wdtp-j-38SF4,1644
+kodit/migrations/versions/85155663351e_initial.py,sha256=Cg7zlF871o9ShV5rQMQ1v7hRV7fI59veDY9cjtTrs-8,3306
+kodit/migrations/versions/__init__.py,sha256=9-lHzptItTzq_fomdIRBegQNm4Znx6pVjwD4MiqRIdo,36
+kodit/search/__init__.py,sha256=4QbdjbrlhNKMovmuKHxJnUeZT7KNjTTFU0GdnuwUHdQ,36
+kodit/search/search_repository.py,sha256=r1fkV6-cy9BKsy5J4WTHaY_FcjMaT1PV5qqqq0gvjZw,5833
+kodit/search/search_service.py,sha256=KePkqCAc3CUcrpNsbDc5DqbF6W2m0TG6TDa9-VSJZS0,4227
+kodit/snippets/__init__.py,sha256=-2coNoCRjTixU9KcP6alpmt7zqf37tCRWH3D7FPJ8dg,48
+kodit/snippets/method_snippets.py,sha256=EVHhSNWahAC5nSXv9fWVFJY2yq25goHdCSCuENC07F8,4145
+kodit/snippets/snippets.py,sha256=mwN0bM1Msu8ZeEsUHyQ7tx3Hj3vZsm8G7Wu4eWSkLY8,1539
+kodit/snippets/languages/__init__.py,sha256=Bj5KKZSls2MQ8ZY1S_nHg447MgGZW-2WZM-oq6vjwwA,1187
+kodit/snippets/languages/csharp.scm,sha256=gbBN4RiV1FBuTJF6orSnDFi8H9JwTw-d4piLJYsWUsc,222
+kodit/snippets/languages/python.scm,sha256=ee85R9PBzwye3IMTE7-iVoKWd_ViU3EJISTyrFGrVeo,429
+kodit/source/__init__.py,sha256=1NTZyPdjThVQpZO1Mp1ColVsS7sqYanOVLqnoqV9Ipo,83
+kodit/source/source_models.py,sha256=xb42CaNDO1CUB8SIW-xXMrB6Ji8cFw-yeJ550xBEg9Q,2398
+kodit/source/source_repository.py,sha256=0EksMpoLzdkfe8S4eeCm4Sf7TuxsOzOzaF4BBsMYo-4,3163
+kodit/source/source_service.py,sha256=qBV9FCFQbJppeFrVo4uMgvC_mzWRIKldymp5yqLx9pw,9255
+kodit-0.1.13.dist-info/METADATA,sha256=Od1OTG0tkd0Cf82juR2DGKBQ8l1RwHQ5VLgtiIW5qeA,2349
+kodit-0.1.13.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+kodit-0.1.13.dist-info/entry_points.txt,sha256=hoTn-1aKyTItjnY91fnO-rV5uaWQLQ-Vi7V5et2IbHY,40
+kodit-0.1.13.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+kodit-0.1.13.dist-info/RECORD,,

kodit/retreival/__init__.py

@@ -1 +0,0 @@
-"""Retrieval package for code search and retrieval functionality."""

kodit-0.1.11.dist-info/RECORD

@@ -1,44 +0,0 @@
-kodit/.gitignore,sha256=ztkjgRwL9Uud1OEi36hGQeDGk3OLK1NfDEO8YqGYy8o,11
-kodit/__init__.py,sha256=aEKHYninUq1yh6jaNfvJBYg-6fenpN132nJt1UU6Jxs,59
-kodit/_version.py,sha256=xfwL5IZGNNwnNDAQtGFjpvlNxqYn3U9IM9B98Du9pJw,513
-kodit/app.py,sha256=Mr5BFHOHx5zppwjC4XPWVvHjwgl1yrKbUjTWXKubJQM,891
-kodit/cli.py,sha256=qEQy_Sd64cEV5KzYsKlGLyMxFQ4fFi-as4QO8CRrKYo,8978
-kodit/config.py,sha256=hQshTMW_8jpk94zP-1JaxowgmW_LrT534ipHFaRUGMw,3006
-kodit/database.py,sha256=kekSdyEATdb47jxzQemkSOXMNOwnUwmVVTpn9hYaDK8,2356
-kodit/log.py,sha256=PhyzQktEyyHaNr78W0wmL-RSRuq311DQ-d0l-EKTGmQ,5417
-kodit/mcp.py,sha256=qp16vRb0TY46-xQy179iWgYebr6Ju_Z91ZSzZnWPHuk,4771
-kodit/middleware.py,sha256=I6FOkqG9-8RH5kR1-0ZoQWfE4qLCB8lZYv8H_OCH29o,2714
-kodit/bm25/__init__.py,sha256=j8zyriNWhbwE5Lbybzg1hQAhANlU9mKHWw4beeUR6og,19
-kodit/bm25/bm25.py,sha256=NtlcLrgqJja11qDGKz_U6tuYWaS9sfbyS-TcA__rBKs,2284
-kodit/embedding/__init__.py,sha256=h9NXzDA1r-K23nvBajBV-RJzHJN0p3UJ7UQsmdnOoRw,24
-kodit/embedding/embedding.py,sha256=X2Fa-eXhQwp__QFj9yxIhvlCAiYVQSaZ2y18ZtG5_1Y,1810
-kodit/embedding/models.py,sha256=rN90vSs86dYiqoawcp8E9jtwY31JoJXYfaDlsJK7uqc,656
-kodit/indexing/__init__.py,sha256=cPyi2Iej3G1JFWlWr7X80_UrsMaTu5W5rBwgif1B3xo,75
-kodit/indexing/models.py,sha256=sZIhGwvL4Dw0QTWFxrjfWctSLkAoDT6fv5DlGz8-Fr8,1258
-kodit/indexing/repository.py,sha256=eIaIbqNs9Z3XTVymZ5Zl5uPWveqiEXNo0JTa-y-Tl24,5430
-kodit/indexing/service.py,sha256=hhQ_6vI7J7LnNgOLbsO4B07TOJvEePqqFviiqr3TL_M,6579
-kodit/migrations/README,sha256=ISVtAOvqvKk_5ThM5ioJE-lMkvf9IbknFUFVU_vPma4,58
-kodit/migrations/__init__.py,sha256=lP5MuwlyWRMO6UcDWnQcQ3G-GYHcFb6rl9gYPHJ1sjo,40
-kodit/migrations/env.py,sha256=bzB6vod_tO-X2F_G671FwYSAn0pyhNw8M1kG4MgidO8,2444
-kodit/migrations/script.py.mako,sha256=zWziKtiwYKEWuwPV_HBNHwa9LCT45_bi01-uSNFaOOE,703
-kodit/migrations/versions/7c3bbc2ab32b_add_embeddings_table.py,sha256=-61qol9PfQKILCDQRA5jEaats9aGZs9Wdtp-j-38SF4,1644
-kodit/migrations/versions/85155663351e_initial.py,sha256=Cg7zlF871o9ShV5rQMQ1v7hRV7fI59veDY9cjtTrs-8,3306
-kodit/migrations/versions/__init__.py,sha256=9-lHzptItTzq_fomdIRBegQNm4Znx6pVjwD4MiqRIdo,36
-kodit/retreival/__init__.py,sha256=33PhJU-3gtsqYq6A1UkaLNKbev_Zee9Lq6dYC59-CsA,69
-kodit/retreival/repository.py,sha256=XHkkeUsnXSrrcthJOL9FXgivn5kkaPnC9Qci6ebwjZc,7294
-kodit/retreival/service.py,sha256=gGp74jnqhyCDF5vKOrN2dJKDnhlfR4HZaxADSrjTb4s,3778
-kodit/snippets/__init__.py,sha256=-2coNoCRjTixU9KcP6alpmt7zqf37tCRWH3D7FPJ8dg,48
-kodit/snippets/method_snippets.py,sha256=EVHhSNWahAC5nSXv9fWVFJY2yq25goHdCSCuENC07F8,4145
-kodit/snippets/snippets.py,sha256=QumvhltWoxXw41SyKb-RbSvAr3m6V3lUy9n0AI8jcto,1409
-kodit/snippets/languages/__init__.py,sha256=Bj5KKZSls2MQ8ZY1S_nHg447MgGZW-2WZM-oq6vjwwA,1187
-kodit/snippets/languages/csharp.scm,sha256=gbBN4RiV1FBuTJF6orSnDFi8H9JwTw-d4piLJYsWUsc,222
-kodit/snippets/languages/python.scm,sha256=ee85R9PBzwye3IMTE7-iVoKWd_ViU3EJISTyrFGrVeo,429
-kodit/sources/__init__.py,sha256=1NTZyPdjThVQpZO1Mp1ColVsS7sqYanOVLqnoqV9Ipo,83
-kodit/sources/models.py,sha256=xb42CaNDO1CUB8SIW-xXMrB6Ji8cFw-yeJ550xBEg9Q,2398
-kodit/sources/repository.py,sha256=mGJrHWH6Uo8YABdoojHFbzaf_jW-2ywJpAHIa1gnc3U,3401
-kodit/sources/service.py,sha256=aV_qiqkU2kMBNPvye5_v4NnZiK-lJ64rQdmFtBtsQaY,9243
-kodit-0.1.11.dist-info/METADATA,sha256=yUO645VYUiVrJMRtwNB71O-6qvC94nS7_ILQ8eQEvoY,2288
-kodit-0.1.11.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-kodit-0.1.11.dist-info/entry_points.txt,sha256=hoTn-1aKyTItjnY91fnO-rV5uaWQLQ-Vi7V5et2IbHY,40
-kodit-0.1.11.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-kodit-0.1.11.dist-info/RECORD,,