cli2api 0.1.0__tar.gz → 0.2.0__tar.gz

{cli2api-0.1.0 → cli2api-0.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cli2api
-Version: 0.1.0
+Version: 0.2.0
 Summary: OpenAI-compatible API over Claude Code CLI
 Requires-Python: >=3.11
 Requires-Dist: fastapi>=0.115.0
@@ -35,45 +35,22 @@ Claude Code CLI is powerful but not all tools support it directly. CLI2API bridg
 
 ## Installation
 
-### Quick Start
+### Quick Start (pip)
 
 ```bash
-# Clone the repository
-git clone https://github.com/anoxis/CLI2API.git
-cd CLI2API
-
-# Install dependencies
-pip install -e .
-
-# Start server
-./cli2api.sh
-```
-
-### Add to PATH (Recommended)
-
-For easy access from anywhere:
-
-```bash
-# Add to your shell profile (~/.bashrc, ~/.zshrc, etc.)
-export PATH="$PATH:/path/to/CLI2API"
-
-# Then run from anywhere
-cli2api.sh
+pip install cli2api
+cli2api
 ```
 
-### Running After Installation
+Server starts at http://localhost:8000
 
-Once installed, you can start the server in several ways:
+### From Source
 
 ```bash
-# Using the script (recommended)
-./cli2api.sh
-
-# Using Python module
-python -m cli2api
-
-# Using uvicorn directly
-uvicorn cli2api.main:app --host 0.0.0.0 --port 8000
+git clone https://github.com/anoxis/CLI2API.git
+cd CLI2API
+pip install -e .
+cli2api
 ```
 
 ### Docker
@@ -111,7 +88,7 @@ pytest tests/ -v
 
 ### Kilo Code / Roo Code
 
-1. Start CLI2API server: `./cli2api.sh`
+1. Start CLI2API server: `cli2api`
 2. Open extension settings
 3. Add custom provider:
    - **Provider Name:** `CLI2API` (or any name)

{cli2api-0.1.0 → cli2api-0.2.0}/README.md

@@ -20,45 +20,22 @@ Claude Code CLI is powerful but not all tools support it directly. CLI2API bridg
 
 ## Installation
 
-### Quick Start
+### Quick Start (pip)
 
 ```bash
-# Clone the repository
-git clone https://github.com/anoxis/CLI2API.git
-cd CLI2API
-
-# Install dependencies
-pip install -e .
-
-# Start server
-./cli2api.sh
-```
-
-### Add to PATH (Recommended)
-
-For easy access from anywhere:
-
-```bash
-# Add to your shell profile (~/.bashrc, ~/.zshrc, etc.)
-export PATH="$PATH:/path/to/CLI2API"
-
-# Then run from anywhere
-cli2api.sh
+pip install cli2api
+cli2api
 ```
 
-### Running After Installation
+Server starts at http://localhost:8000
 
-Once installed, you can start the server in several ways:
+### From Source
 
 ```bash
-# Using the script (recommended)
-./cli2api.sh
-
-# Using Python module
-python -m cli2api
-
-# Using uvicorn directly
-uvicorn cli2api.main:app --host 0.0.0.0 --port 8000
+git clone https://github.com/anoxis/CLI2API.git
+cd CLI2API
+pip install -e .
+cli2api
 ```
 
 ### Docker
@@ -96,7 +73,7 @@ pytest tests/ -v
 
 ### Kilo Code / Roo Code
 
-1. Start CLI2API server: `./cli2api.sh`
+1. Start CLI2API server: `cli2api`
 2. Open extension settings
 3. Add custom provider:
    - **Provider Name:** `CLI2API` (or any name)

cli2api-0.2.0/cli2api/api/v1/chat.py

@@ -0,0 +1,54 @@
+"""Chat completions endpoint - OpenAI compatible."""
+
+from fastapi import APIRouter, Depends, HTTPException
+from fastapi.responses import StreamingResponse
+
+from cli2api.api.dependencies import get_provider
+from cli2api.api.utils import parse_model_name
+from cli2api.constants import HTTP_GATEWAY_TIMEOUT, HTTP_INTERNAL_ERROR
+from cli2api.providers.claude import ClaudeCodeProvider
+from cli2api.schemas.openai import ChatCompletionRequest
+from cli2api.services.completion import CompletionService
+
+router = APIRouter()
+
+
+@router.post("/chat/completions")
+async def chat_completions(
+    request: ChatCompletionRequest,
+    provider: ClaudeCodeProvider = Depends(get_provider),
+):
+    """Create a chat completion (OpenAI-compatible endpoint)."""
+    actual_model = parse_model_name(request.model)
+    service = CompletionService(provider)
+    completion_id = service.generate_completion_id()
+
+    if request.stream:
+        return StreamingResponse(
+            service.stream_completion(
+                messages=request.messages,
+                model=actual_model,
+                completion_id=completion_id,
+                tools=request.tools,
+                reasoning_effort=request.reasoning_effort,
+            ),
+            media_type="text/event-stream",
+            headers={
+                "Cache-Control": "no-cache",
+                "Connection": "keep-alive",
+                "X-Accel-Buffering": "no",
+            },
+        )
+
+    try:
+        response = await service.create_completion(
+            messages=request.messages,
+            model=actual_model,
+            completion_id=completion_id,
+            tools=request.tools,
+        )
+        return response.model_dump(exclude_none=True)
+    except TimeoutError as e:
+        raise HTTPException(status_code=HTTP_GATEWAY_TIMEOUT, detail=str(e))
+    except RuntimeError as e:
+        raise HTTPException(status_code=HTTP_INTERNAL_ERROR, detail=str(e))

{cli2api-0.1.0 → cli2api-0.2.0}/cli2api/api/v1/responses.py

@@ -10,13 +10,17 @@ from typing import Any, AsyncIterator, Optional
 
 from fastapi import APIRouter, Depends, HTTPException
 from fastapi.responses import StreamingResponse
-from pydantic import BaseModel, ConfigDict
+from pydantic import BaseModel, ConfigDict, Field, field_validator
 
 from cli2api.api.dependencies import get_provider
 from cli2api.api.utils import parse_model_name
+from cli2api.constants import ID_HEX_LENGTH, MESSAGE_ID_PREFIX, RESPONSE_ID_PREFIX
 from cli2api.providers.claude import ClaudeCodeProvider
 from cli2api.schemas.openai import ChatMessage
 from cli2api.streaming.sse import sse_encode, sse_error
+from cli2api.utils.logging import get_logger
+
+logger = get_logger(__name__)
 
 router = APIRouter()
 
@@ -25,8 +29,6 @@ router = APIRouter()
 
 
 class ResponsesInputMessage(BaseModel):
-    """Input message for Responses API."""
-
     model_config = ConfigDict(extra="ignore")
 
     role: str
@@ -42,20 +44,28 @@ class ResponsesRequest(BaseModel):
     input: list[ResponsesInputMessage] | str
     stream: bool = False
     instructions: Optional[str] = None
-    temperature: Optional[float] = None
-    max_output_tokens: Optional[int] = None
+    temperature: Optional[float] = Field(default=None, ge=0.0, le=2.0)
+    max_output_tokens: Optional[int] = Field(default=None, gt=0)
     # Additional fields
     tools: Optional[list[Any]] = None
     tool_choice: Optional[Any] = None
     metadata: Optional[dict] = None
 
+    @field_validator("input")
+    @classmethod
+    def input_not_empty(cls, v: list | str) -> list | str:
+        """Validate that input is not empty."""
+        if isinstance(v, str) and not v.strip():
+            raise ValueError("input cannot be empty string")
+        if isinstance(v, list) and len(v) == 0:
+            raise ValueError("input cannot be empty list")
+        return v
+
 
 # === Response Models ===
 
 
 class ResponsesOutput(BaseModel):
-    """Output content in response."""
-
     type: str = "message"
     id: str
     role: str = "assistant"
@@ -63,8 +73,6 @@ class ResponsesOutput(BaseModel):
 
 
 class ResponsesResponse(BaseModel):
-    """Response body for /v1/responses."""
-
     id: str
     object: str = "response"
     created_at: int
@@ -74,40 +82,56 @@ class ResponsesResponse(BaseModel):
 
 
 def convert_to_chat_messages(request: ResponsesRequest) -> list[ChatMessage]:
-    """Convert Responses API input to ChatMessage list."""
-    messages = []
+    """Convert Responses API input to ChatMessage list.
+
+    Args:
+        request: The ResponsesRequest with input messages.
+
+    Returns:
+        List of ChatMessage objects for the provider.
+    """
+    messages: list[ChatMessage] = []
 
     # Add instructions as system message
     if request.instructions:
         messages.append(ChatMessage(role="system", content=request.instructions))
 
-    # Handle input
+    # Handle input - can be string or list of messages
     if isinstance(request.input, str):
         messages.append(ChatMessage(role="user", content=request.input))
     else:
         for msg in request.input:
-            content = msg.content
-            if isinstance(content, list):
-                # Extract text from content blocks
-                texts = []
-                for item in content:
-                    if isinstance(item, dict) and item.get("type") == "text":
-                        texts.append(item.get("text", ""))
-                    elif isinstance(item, str):
-                        texts.append(item)
-                content = "\n".join(texts)
-            elif content is None:
-                content = ""
-
-            role = msg.role
-            if role not in ("system", "user", "assistant"):
-                role = "user"  # Default to user for unknown roles
-
-            messages.append(ChatMessage(role=role, content=str(content)))
+            content: str = _extract_message_content(msg.content)
+            role: str = msg.role if msg.role in ("system", "user", "assistant") else "user"
+            messages.append(ChatMessage(role=role, content=content))
 
     return messages
 
 
+def _extract_message_content(content: Any) -> str:
+    """Extract text content from various message content formats.
+
+    Args:
+        content: Message content - can be str, list of content blocks, or None.
+
+    Returns:
+        Extracted text content as string.
+    """
+    if content is None:
+        return ""
+    if isinstance(content, str):
+        return content
+    if isinstance(content, list):
+        texts: list[str] = []
+        for item in content:
+            if isinstance(item, dict) and item.get("type") == "text":
+                texts.append(item.get("text", ""))
+            elif isinstance(item, str):
+                texts.append(item)
+        return "\n".join(texts)
+    return str(content)
+
+
 @router.post("/responses")
 async def create_response(
     request: ResponsesRequest,
@@ -121,7 +145,7 @@ async def create_response(
 
     # Convert to chat messages
     messages = convert_to_chat_messages(request)
-    response_id = f"resp-{uuid.uuid4().hex[:24]}"
+    response_id = f"{RESPONSE_ID_PREFIX}{uuid.uuid4().hex[:ID_HEX_LENGTH]}"
 
     if request.stream:
         return StreamingResponse(
@@ -150,7 +174,7 @@ async def create_response(
         except RuntimeError as e:
             raise HTTPException(status_code=500, detail=str(e))
 
-        output_id = f"msg-{uuid.uuid4().hex[:24]}"
+        output_id = f"{MESSAGE_ID_PREFIX}{uuid.uuid4().hex[:ID_HEX_LENGTH]}"
 
         return ResponsesResponse(
             id=response_id,
@@ -174,7 +198,7 @@ async def stream_response(
 ) -> AsyncIterator[str]:
     """Generate SSE events for streaming response."""
     created = int(time.time())
-    output_id = f"msg-{uuid.uuid4().hex[:24]}"
+    output_id = f"{MESSAGE_ID_PREFIX}{uuid.uuid4().hex[:ID_HEX_LENGTH]}"
     content_buffer = ""
 
     try:
@@ -248,8 +272,14 @@ async def stream_response(
             }
         })
 
+        logger.info(f"[{response_id}] Stream completed successfully")
         yield "data: [DONE]\n\n"
 
+    except RuntimeError as e:
+        logger.error(f"[{response_id}] Provider error: {e}")
+        yield sse_error(str(e))
+        yield "data: [DONE]\n\n"
     except Exception as e:
+        logger.error(f"[{response_id}] Stream error: {e}")
         yield sse_error(str(e))
         yield "data: [DONE]\n\n"

cli2api-0.2.0/cli2api/constants.py

@@ -0,0 +1,61 @@
+"""Application constants.
+
+Centralized location for magic numbers and strings used throughout the codebase.
+"""
+
+# ====================
+# ID Generation
+# ====================
+
+# Length of hex suffix for generated IDs (24 hex chars = 12 bytes of randomness)
+ID_HEX_LENGTH = 24
+
+# ID prefixes for different entity types
+CHAT_COMPLETION_ID_PREFIX = "chatcmpl-"
+RESPONSE_ID_PREFIX = "resp-"
+MESSAGE_ID_PREFIX = "msg-"
+TOOL_CALL_ID_PREFIX = "call_"
+
+# ====================
+# Streaming
+# ====================
+
+# Maximum size of content chunks when splitting for streaming
+# Chosen to balance between responsiveness and overhead
+STREAM_CHUNK_MAX_SIZE = 150
+
+# Preferred split points for chunking (in priority order)
+CHUNK_SPLIT_SEPARATORS = (" ", "\n", ".", ",", ";")
+
+# Minimum position ratio for split point (don't split too early)
+CHUNK_SPLIT_MIN_RATIO = 0.5
+
+# ====================
+# HTTP Status Codes
+# ====================
+
+HTTP_OK = 200
+HTTP_BAD_REQUEST = 400
+HTTP_INTERNAL_ERROR = 500
+HTTP_GATEWAY_TIMEOUT = 504
+
+# ====================
+# OpenAI API
+# ====================
+
+# Default object types for OpenAI compatibility
+OBJECT_CHAT_COMPLETION = "chat.completion"
+OBJECT_CHAT_COMPLETION_CHUNK = "chat.completion.chunk"
+OBJECT_MODEL = "model"
+OBJECT_LIST = "list"
+
+# Finish reasons
+FINISH_REASON_STOP = "stop"
+FINISH_REASON_TOOL_CALLS = "tool_calls"
+
+# ====================
+# Tool Call Markers
+# ====================
+
+TOOL_CALL_START_MARKER = "<tool_call>"
+TOOL_CALL_END_MARKER = "</tool_call>"