kodit 0.3.12__py3-none-any.whl → 0.3.14__py3-none-any.whl

kodit/_version.py

@@ -17,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.3.12'
-__version_tuple__ = version_tuple = (0, 3, 12)
+__version__ = version = '0.3.14'
+__version_tuple__ = version_tuple = (0, 3, 14)

kodit/app.py

@@ -26,10 +26,11 @@ async def app_lifespan(_: FastAPI) -> AsyncIterator[AppLifespanState]:
     """Manage application lifespan for auto-indexing and sync."""
     global _auto_indexing_service, _sync_scheduler_service  # noqa: PLW0603
 
+    # App context has already been configured by the CLI.
     app_context = AppContext()
-    db = await app_context.get_db()
 
     # Start auto-indexing service
+    db = await app_context.get_db()
     _auto_indexing_service = AutoIndexingService(
         app_context=app_context,
         session_factory=db.session_factory,

kodit/config.py

@@ -3,6 +3,7 @@
 from __future__ import annotations
 
 import asyncio
+from enum import Enum
 from functools import wraps
 from pathlib import Path
 from typing import TYPE_CHECKING, Annotated, Any, Literal, TypeVar
@@ -18,15 +19,22 @@ from pydantic_settings import (
     SettingsConfigDict,
 )
 
+from kodit.database import Database
+
 if TYPE_CHECKING:
     from collections.abc import Callable, Coroutine
 
 
-from kodit.database import Database
+class LogFormat(Enum):
+    """The format of the log output."""
+
+    PRETTY = "pretty"
+    JSON = "json"
+
 
 DEFAULT_BASE_DIR = Path.home() / ".kodit"
 DEFAULT_LOG_LEVEL = "INFO"
-DEFAULT_LOG_FORMAT = "pretty"
+DEFAULT_LOG_FORMAT = LogFormat.PRETTY
 DEFAULT_DISABLE_TELEMETRY = False
 T = TypeVar("T")
 
@@ -40,6 +48,7 @@ class Endpoint(BaseModel):
     base_url: str | None = None
     model: str | None = None
     api_key: str | None = None
+    num_parallel_tasks: int | None = None
 
 
 class Search(BaseModel):
@@ -165,7 +174,7 @@ class AppContext(BaseSettings):
         default_factory=lambda data: f"sqlite+aiosqlite:///{data['data_dir']}/kodit.db"
     )
     log_level: str = Field(default=DEFAULT_LOG_LEVEL)
-    log_format: str = Field(default=DEFAULT_LOG_FORMAT)
+    log_format: LogFormat = Field(default=DEFAULT_LOG_FORMAT)
     disable_telemetry: bool = Field(default=DEFAULT_DISABLE_TELEMETRY)
     default_endpoint: Endpoint | None = Field(
         default=None,

kodit/infrastructure/embedding/embedding_factory.py

@@ -14,6 +14,7 @@ from kodit.infrastructure.embedding.embedding_providers.local_embedding_provider
     LocalEmbeddingProvider,
 )
 from kodit.infrastructure.embedding.embedding_providers.openai_embedding_provider import (  # noqa: E501
+    OPENAI_NUM_PARALLEL_TASKS,
     OpenAIEmbeddingProvider,
 )
 from kodit.infrastructure.embedding.local_vector_search_repository import (
@@ -55,6 +56,7 @@ def embedding_domain_service_factory(
                 max_retries=2,
             ),
             model_name=endpoint.model or "text-embedding-3-small",
+            num_parallel_tasks=endpoint.num_parallel_tasks or OPENAI_NUM_PARALLEL_TASKS,
         )
     else:
         log_event("kodit.embedding", {"provider": "local"})

kodit/infrastructure/embedding/embedding_providers/openai_embedding_provider.py

@@ -18,14 +18,17 @@ MAX_TOKENS = 8192  # Conservative token limit for the embedding model
 BATCH_SIZE = (
     10  # Maximum number of items per API call (keeps existing test expectations)
 )
-OPENAI_NUM_PARALLEL_TASKS = 25  # Semaphore limit for concurrent OpenAI requests
+OPENAI_NUM_PARALLEL_TASKS = 10  # Semaphore limit for concurrent OpenAI requests
 
 
 class OpenAIEmbeddingProvider(EmbeddingProvider):
     """OpenAI embedding provider that uses OpenAI's embedding API."""
 
     def __init__(
-        self, openai_client: AsyncOpenAI, model_name: str = "text-embedding-3-small"
+        self,
+        openai_client: AsyncOpenAI,
+        model_name: str = "text-embedding-3-small",
+        num_parallel_tasks: int = OPENAI_NUM_PARALLEL_TASKS,
     ) -> None:
         """Initialize the OpenAI embedding provider.
 
@@ -36,6 +39,7 @@ class OpenAIEmbeddingProvider(EmbeddingProvider):
         """
         self.openai_client = openai_client
         self.model_name = model_name
+        self.num_parallel_tasks = num_parallel_tasks
         self.log = structlog.get_logger(__name__)
 
         # Lazily initialised token encoding
@@ -78,7 +82,7 @@ class OpenAIEmbeddingProvider(EmbeddingProvider):
         # Process batches concurrently (but bounded by a semaphore)
         # -----------------------------------------------------------------
 
-        sem = asyncio.Semaphore(OPENAI_NUM_PARALLEL_TASKS)
+        sem = asyncio.Semaphore(self.num_parallel_tasks)
 
         async def _process_batch(
             batch: list[EmbeddingRequest],

kodit/infrastructure/enrichment/enrichment_factory.py

@@ -9,6 +9,7 @@ from kodit.infrastructure.enrichment.local_enrichment_provider import (
     LocalEnrichmentProvider,
 )
 from kodit.infrastructure.enrichment.openai_enrichment_provider import (
+    OPENAI_NUM_PARALLEL_TASKS,
     OpenAIEnrichmentProvider,
 )
 from kodit.log import log_event
@@ -54,6 +55,7 @@ def enrichment_domain_service_factory(
                 max_retries=2,
             ),
             model_name=endpoint.model or "gpt-4o-mini",
+            num_parallel_tasks=endpoint.num_parallel_tasks or OPENAI_NUM_PARALLEL_TASKS,
         )
     else:
         log_event("kodit.enrichment", {"provider": "local"})

kodit/infrastructure/enrichment/openai_enrichment_provider.py

@@ -21,7 +21,12 @@ OPENAI_NUM_PARALLEL_TASKS = 40
 class OpenAIEnrichmentProvider(EnrichmentProvider):
     """OpenAI enrichment provider implementation."""
 
-    def __init__(self, openai_client: Any, model_name: str = "gpt-4o-mini") -> None:
+    def __init__(
+        self,
+        openai_client: Any,
+        model_name: str = "gpt-4o-mini",
+        num_parallel_tasks: int = OPENAI_NUM_PARALLEL_TASKS,
+    ) -> None:
         """Initialize the OpenAI enrichment provider.
 
         Args:
@@ -32,6 +37,7 @@ class OpenAIEnrichmentProvider(EnrichmentProvider):
         self.log = structlog.get_logger(__name__)
         self.openai_client = openai_client
         self.model_name = model_name
+        self.num_parallel_tasks = num_parallel_tasks
 
     async def enrich(
         self, requests: list[EnrichmentRequest]
@@ -50,7 +56,7 @@ class OpenAIEnrichmentProvider(EnrichmentProvider):
             return
 
         # Process batches in parallel with a semaphore to limit concurrent requests
-        sem = asyncio.Semaphore(OPENAI_NUM_PARALLEL_TASKS)
+        sem = asyncio.Semaphore(self.num_parallel_tasks)
 
         async def process_request(request: EnrichmentRequest) -> EnrichmentResponse:
             async with sem:

kodit/log.py

@@ -7,7 +7,6 @@ import shutil
 import subprocess
 import sys
 import uuid
-from enum import Enum
 from functools import lru_cache
 from pathlib import Path
 from typing import Any
@@ -17,7 +16,7 @@ import structlog
 from structlog.types import EventDict
 
 from kodit import _version
-from kodit.config import AppContext
+from kodit.config import AppContext, LogFormat
 
 _MAC_RE = re.compile(r"(?:[0-9A-Fa-f]{2}[:-]){5}[0-9A-Fa-f]{2}")
 
@@ -35,13 +34,6 @@ def drop_color_message_key(_, __, event_dict: EventDict) -> EventDict:  # noqa:
     return event_dict
 
 
-class LogFormat(Enum):
-    """The format of the log output."""
-
-    PRETTY = "pretty"
-    JSON = "json"
-
-
 def configure_logging(app_context: AppContext) -> None:
     """Configure logging for the application."""
     timestamper = structlog.processors.TimeStamper(fmt="iso")

{kodit-0.3.12.dist-info → kodit-0.3.14.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kodit
-Version: 0.3.12
+Version: 0.3.14
 Summary: Code indexing for better AI code generation
 Project-URL: Homepage, https://docs.helixml.tech/kodit/
 Project-URL: Documentation, https://docs.helixml.tech/kodit/

{kodit-0.3.12.dist-info → kodit-0.3.14.dist-info}/RECORD

@@ -1,11 +1,11 @@
 kodit/.gitignore,sha256=ztkjgRwL9Uud1OEi36hGQeDGk3OLK1NfDEO8YqGYy8o,11
 kodit/__init__.py,sha256=aEKHYninUq1yh6jaNfvJBYg-6fenpN132nJt1UU6Jxs,59
-kodit/_version.py,sha256=LY4sFGHLjqvfo3vSpOS7ED-Kifi9goigErF1wVOVwKc,513
-kodit/app.py,sha256=um69vDbaUdy_auEubKLzFo7oXUK5MIjjMl9O6n8T5wc,3859
+kodit/_version.py,sha256=dAYECGus8vEags4_X66QYJmiBFCY-yJJXDyexsFv72w,513
+kodit/app.py,sha256=aK6TP-7L3aHCiatNm_SYSw0zU4G3EuOcvdBZ3xPHO_U,3917
 kodit/cli.py,sha256=ZOS_VzCHGjJRZzZpaVR00QXSPIwRXPYu-pTrbEtlyR0,19328
-kodit/config.py,sha256=qk242V-ys2cLRq3vy6SrTi8_HFaDNL-f5i63HIpyh2s,8756
+kodit/config.py,sha256=kuGdl-q5r9b7spt89-1mfVNuoQLqkXQ9-nuJFfRflUU,8940
 kodit/database.py,sha256=kI9yBm4uunsgV4-QeVoCBL0wLzU4kYmYv5qZilGnbPE,1740
-kodit/log.py,sha256=r0o7IpNvV-dNW-cTNWu1ouJF71vD9wHYzvqDPzeDYfw,8768
+kodit/log.py,sha256=XyuseZk90gUBj1B7np2UO2EW9eE_ApayIpPRvI19KCE,8651
 kodit/mcp.py,sha256=aEcPc8dQiZaR0AswCZZNxcm_rhhUZNsEBimYti0ibSI,7221
 kodit/middleware.py,sha256=xBmC6keFeNsS0y8XUcIKzJzuefkE9bq2UhW1fR5cqxg,2770
 kodit/reporting.py,sha256=icce1ZyiADsA_Qz-mSjgn2H4SSqKuGfLKnw-yrl9nsg,2722
@@ -50,19 +50,19 @@ kodit/infrastructure/cloning/metadata.py,sha256=GD2UnCC1oR82RD0SVUqk9CJOqzXPxhOA
 kodit/infrastructure/cloning/git/__init__.py,sha256=20ePcp0qE6BuLsjsv4KYB1DzKhMIMsPXwEqIEZtjTJs,34
 kodit/infrastructure/cloning/git/working_copy.py,sha256=qYcrR5qP1rhWZiYGMT1p-1Alavi_YvQLXx4MgIV7eXs,2611
 kodit/infrastructure/embedding/__init__.py,sha256=F-8nLlWAerYJ0MOIA4tbXHLan8bW5rRR84vzxx6tRKI,39
-kodit/infrastructure/embedding/embedding_factory.py,sha256=SOVX3wGxW8Qo2lJmMZ0kRssXKn6A4W7DmSpgo0zWTfM,3625
+kodit/infrastructure/embedding/embedding_factory.py,sha256=BsePuOAU4UmFP1hbn3WwZkY6dTG-oxlte0l7P_LsuuA,3745
 kodit/infrastructure/embedding/local_vector_search_repository.py,sha256=ExweyNEL5cP-g3eDhGqZSih7zhdOrop2WdFPPJL-tB4,3505
 kodit/infrastructure/embedding/vectorchord_vector_search_repository.py,sha256=PIoU0HsDlaoXDXnGjOR0LAkAcW4JiE3ymJy_SBhEopc,8030
 kodit/infrastructure/embedding/embedding_providers/__init__.py,sha256=qeZ-oAIAxMl5QqebGtO1lq-tHjl_ucAwOXePklcwwGk,34
 kodit/infrastructure/embedding/embedding_providers/batching.py,sha256=a8CL9PX2VLmbeg616fc_lQzfC4BWTVn32m4SEhXpHxc,3279
 kodit/infrastructure/embedding/embedding_providers/hash_embedding_provider.py,sha256=V6OdCuWyQQOvo3OJGRi-gBKDApIcrELydFg7T696P5s,2257
 kodit/infrastructure/embedding/embedding_providers/local_embedding_provider.py,sha256=9aLV1Zg4KMhYWlGRwgAUtswW4aIabNqbsipWhAn64RI,4133
-kodit/infrastructure/embedding/embedding_providers/openai_embedding_provider.py,sha256=MA68eoAyIKIfazWgm4XrCyWF8BIzjp8dNi_aoLHPaB0,3844
+kodit/infrastructure/embedding/embedding_providers/openai_embedding_provider.py,sha256=QfvdXXbWa-1DQriCzV-TmFm3XGqnVP_7PSQ6u6fKnuw,3973
 kodit/infrastructure/enrichment/__init__.py,sha256=8acZKNzql8Fs0lceFu9U3KoUrOptRBtVIxr_Iw6lz3Y,40
-kodit/infrastructure/enrichment/enrichment_factory.py,sha256=2h-c96FWgSi0UG6IeL9Yjae80KYdzJqes5HIJQrkpc8,1919
+kodit/infrastructure/enrichment/enrichment_factory.py,sha256=_JWkna3g8q8hzaPq1NRfZU9Y_sv99xcloM4BCJAWmHw,2039
 kodit/infrastructure/enrichment/local_enrichment_provider.py,sha256=7Vlwu1jPJ5KNUn1a51M1P-laUd5YVFJA8EeH6KO-95k,3960
 kodit/infrastructure/enrichment/null_enrichment_provider.py,sha256=DhZkJBnkvXg_XSAs-oKiFnKqYFPnmTl3ikdxrqeEfbc,713
-kodit/infrastructure/enrichment/openai_enrichment_provider.py,sha256=fenq4HiJ2UkrzsE2D0A0qpmro38z9mKaIzKKU5v7hnY,3189
+kodit/infrastructure/enrichment/openai_enrichment_provider.py,sha256=quu9GLGhxGk3ilrg5kPZdGiCTfuQ20JuwM3Grv6qaQE,3332
 kodit/infrastructure/git/__init__.py,sha256=0iMosFzudj4_xNIMe2SRbV6l5bWqkjnUsZoFsoZFuM8,33
 kodit/infrastructure/git/git_utils.py,sha256=KERwmhWDR4ooMQKS-nSPxjvdCzoWF9NS6nhdeXyzdtY,571
 kodit/infrastructure/ignore/__init__.py,sha256=VzFv8XOzHmsu0MEGnWVSF6KsgqLBmvHlRqAkT1Xb1MY,36
@@ -95,8 +95,8 @@ kodit/migrations/versions/c3f5137d30f5_index_all_the_things.py,sha256=r7ukmJ_axX
 kodit/utils/__init__.py,sha256=DPEB1i8evnLF4Ns3huuAYg-0pKBFKUFuiDzOKG9r-sw,33
 kodit/utils/dump_openapi.py,sha256=29VdjHpNSaGAg7RjQw0meq1OLhljCx1ElgBlTC8xoF4,1247
 kodit/utils/path_utils.py,sha256=thK6YGGNvQThdBaCYCCeCvS1L8x-lwl3AoGht2jnjGw,1645
-kodit-0.3.12.dist-info/METADATA,sha256=xfv-ASXWA_i_TkP-TtPQDkqL2Yl_dKohqwtkXZqktR8,7672
-kodit-0.3.12.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-kodit-0.3.12.dist-info/entry_points.txt,sha256=hoTn-1aKyTItjnY91fnO-rV5uaWQLQ-Vi7V5et2IbHY,40
-kodit-0.3.12.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-kodit-0.3.12.dist-info/RECORD,,
+kodit-0.3.14.dist-info/METADATA,sha256=H9I5v4D4kVAJnF1-ofQZpxvMif86Vn-CNNWGvV9VI7M,7672
+kodit-0.3.14.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+kodit-0.3.14.dist-info/entry_points.txt,sha256=hoTn-1aKyTItjnY91fnO-rV5uaWQLQ-Vi7V5et2IbHY,40
+kodit-0.3.14.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+kodit-0.3.14.dist-info/RECORD,,