Skip to content

C
coderai
Commit fcaa9452 authored by Your Name's avatar Your Name
Browse files
Options

Fix debug logging - use state module for global_debug 

- text.py had local global_debug variable that shadowed the state module
- Changed text.py to import get_global_debug from state module
- Changed set_global_debug() in text.py to call state module's function
- Changed all 'if global_debug:' to 'if get_global_debug():' in text.py
- log.py was already using get_global_debug() correctly
parent ebd925a8
Hide whitespace changes
Inline Side-by-side
Showing with 49 additions and 47 deletions
codai/api/text.py
View file @ fcaa9452
......@@ -16,13 +16,19 @@ from codai.queue.manager import QueueManager, queue_manager
from codai.pydantic.textrequest import ChatCompletionRequest, ToolFunction, Tool
from codai.models.parser import filter_malformed_content, filter_repetition, OpenAIFormatter, ModelParserAdapter, ToolCallParser
# Import global state from state module
from codai.api.state import (
set_global_args as _set_global_args,
get_global_debug,
set_global_debug as _set_global_debug,
set_global_system_prompt as _set_global_system_prompt,
set_global_tools_closer_prompt as _set_global_tools_closer_prompt,
get_grammar_guided_gen,
set_grammar_guided_gen as _set_grammar_guided_gen,
)
# Global reference to be set by coderai
global_args = None
global_debug = False
global_system_prompt = None
global_tools_closer_prompt = False
grammar_guided_gen = False
# =============================================================================
......@@ -36,27 +42,23 @@ def set_global_args(args):
def set_global_debug(debug: bool):
"""Set the global debug flag."""
global global_debug
global_debug = debug
"""Set the global debug flag (via state module)."""
_set_global_debug(debug)
def set_global_system_prompt(prompt):
"""Set the global system prompt."""
global global_system_prompt
global_system_prompt = prompt
"""Set the global system prompt (via state module)."""
_set_global_system_prompt(prompt)
def set_global_tools_closer_prompt(tools_closer: bool):
"""Set the global tools-closer-prompt flag."""
global global_tools_closer_prompt
global_tools_closer_prompt = tools_closer
"""Set the global tools-closer-prompt flag (via state module)."""
_set_global_tools_closer_prompt(tools_closer)
def set_grammar_guided_gen(enabled: bool):
"""Set the grammar-guided generation flag."""
global grammar_guided_gen
grammar_guided_gen = enabled
"""Set the grammar-guided generation flag (via state module)."""
_set_grammar_guided_gen(enabled)
# =============================================================================
......@@ -357,7 +359,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
reasoning_enabled = enable_thinking_api or (len(force_reasoning_args) > 0)
# DEBUG: Print force_reasoning status when debug mode is enabled
if global_debug:
if get_global_debug():
print(f"\n{'='*60}")
print(f"=== REASONING MODE DEBUG ===")
print(f"{'='*60}")
......@@ -379,7 +381,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
qwen3_system_addon = ""
if use_qwen3_penalties:
qwen3_system_addon = "\n\nCRITICAL: Do not repeat tool calls. If a tool fails with an [ERROR], do not retry the exact same parameters. Propose a different approach or ask for clarification."
if global_debug:
if get_global_debug():
print(f"QWEEN3: Adding penalties and system addon for qwen3 with force_reasoning")
# Handle 'chat' - enable thinking API parameter
......@@ -388,7 +390,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
if "chat" in force_reasoning_args or enable_thinking_api:
if hasattr(request, 'thinking'):
request.thinking = {"type": "enabled"}
if global_debug:
if get_global_debug():
print(f"CHAT: Reasoning API param enabled")
# Handle 'inject' - system prompt injection
......@@ -422,7 +424,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
if not system_found:
messages = [ChatMessage(role="system", content=system_content)] + list(messages)
if global_debug:
if get_global_debug():
print(f"INJECT: System prompt injected with agentic instructions")
print(f"\n--- INJECTED SYSTEM PROMPT ---")
print(system_content)
......@@ -456,7 +458,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
from codai.pydantic.textrequest import ChatMessage
messages = [ChatMessage(role="user", content=seeded_prompt)]
if global_debug:
if get_global_debug():
print(f"PROMPT: Prompt seeding applied (ends with thought tag)")
print(f"\n--- SEEDED PROMPT (last 80 chars) ---")
print(f"...{seeded_prompt[-80:]}")
......@@ -488,7 +490,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
if "</think>" not in stop_sequences:
stop_sequences.append("</think>\n")
if global_debug:
if get_global_debug():
print(f"STOP: Added reasoning stop tokens: {additional_stops}")
# Format messages with tools if provided - BUT SKIP for raw mode
......@@ -604,11 +606,11 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
)
use_raw_mode = True
if global_debug:
if get_global_debug():
print(f"RAW: Using template_manager.format_for_raw_completion")
print(f"RAW: Prompt ends with: ...{raw_prompt_for_generation[-80:]}")
else:
if global_debug:
if get_global_debug():
print(f"RAW: template_manager.format_for_raw_completion not available")
# Get resolved model name for response (with coderai/ prefix and proper formatting)
......@@ -618,7 +620,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
# Handle raw mode - two pass: first capture reasoning, then get final answer
if use_raw_mode and raw_prompt_for_generation:
if global_debug:
if get_global_debug():
print(f"RAW: Starting two-pass generation")
print(f"RAW: First pass prompt: ...{raw_prompt_for_generation[-100:]}")
......@@ -639,7 +641,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
thought_tag, close_tag, _ = get_reasoning_stop_tokens(model_family)
reasoning_text = ""
if global_debug:
if get_global_debug():
print(f"DEBUG: raw_stream_generate started, stream=True")
# Use the backend's async generate if available
......@@ -655,19 +657,19 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
reasoning_text += chunk
# Debug: log first pass chunks
if global_debug:
if get_global_debug():
print(f"DEBUG FIRST PASS: chunk length={len(chunk)}, total reasoning so far={len(reasoning_text)}")
yield f"data: {json.dumps({'choices': [{'delta': {'content': chunk}, 'finish_reason': None}]})}\n\n"
# Check if we hit the close tag
if close_tag and close_tag in reasoning_text:
if global_debug:
if get_global_debug():
print(f"DEBUG: Close tag detected in first pass, reasoning length={len(reasoning_text)}")
break
else:
# Fallback: non-streaming
if global_debug:
if get_global_debug():
print(f"DEBUG: Using non-streaming fallback for first pass")
first_pass_result = current_manager.generate(
prompt=raw_prompt_for_generation,
......@@ -686,7 +688,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
# Second pass: get the rest
full_prompt = raw_prompt_for_generation + reasoning_text + (close_tag or "")
if global_debug:
if get_global_debug():
print(f"DEBUG: raw_stream_generate second pass, full_prompt length: {len(full_prompt)}")
second_pass_result = current_manager.generate(
......@@ -719,11 +721,11 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
reasoning_text = reasoning_text[:earliest_tool_idx].strip()
# Prepend the tool part to second_pass_result so it can be extracted as a tool call
second_pass_result = tool_part + second_pass_result
if global_debug:
if get_global_debug():
print(f"DEBUG: Moved tool call from reasoning to second_pass_result: {tool_part[:100]}...")
# In debug mode, dump the full generated text (second pass result)
if global_debug:
if get_global_debug():
print(f"\n{'='*80}")
print(f"=== RAW STREAM: FULL GENERATED TEXT (DEBUG) ===")
print(f"{'='*80}")
......@@ -746,7 +748,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
# CRITICAL: Only extract from second pass, never from reasoning
# Reasoning may contain partial/incomplete tool calls that confuse the parser
if global_debug:
if get_global_debug():
print(f"DEBUG: Tool extraction - using second_pass_result only")
print(f"DEBUG: Second pass result length: {len(second_pass_result) if second_pass_result else 0}")
print(f"DEBUG: Reasoning text length: {len(reasoning_text) if reasoning_text else 0}")
......@@ -794,7 +796,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
validated_calls.append(tc)
if len(validated_calls) != len(extracted_tool_calls):
if global_debug:
if get_global_debug():
print(f"DEBUG: Filtered out {len(extracted_tool_calls) - len(validated_calls)} invalid tool calls")
extracted_tool_calls = validated_calls if validated_calls else None
......@@ -804,7 +806,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
print(f"{'='*80}")
print(json.dumps(extracted_tool_calls, indent=2))
print(f"{'='*80}\n")
elif global_debug:
elif get_global_debug():
print(f"DEBUG: No tool calls found in raw stream")
if extracted_tool_calls:
......@@ -829,7 +831,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
**extra_params,
)
if global_debug:
if get_global_debug():
print(f"RAW: First pass result: ...{first_pass_result[-200:]}")
# Dump first pass result if --dump is enabled
......@@ -866,13 +868,13 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
if earliest_tool_tag:
# Split at tool tag
if global_debug:
if get_global_debug():
print(f"RAW: No close tag found, using tool tag '{earliest_tool_tag}' as fallback")
parts = first_pass_result.split(earliest_tool_tag, 1)
reasoning_text = parts[0]
final_text = earliest_tool_tag + (parts[1] if len(parts) > 1 else "")
if global_debug:
if get_global_debug():
print(f"RAW: Extracted reasoning: {reasoning_text[:100]}...")
print(f"RAW: Final text before cleanup: {final_text[:100]}...")
......@@ -912,10 +914,10 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
reasoning_text = reasoning_text[:earliest_tool_idx].strip()
# Prepend the tool part to final_text so it can be extracted as a tool call
final_text = tool_part + final_text
if global_debug:
if get_global_debug():
print(f"RAW: Moved tool call from reasoning to final_text: {tool_part[:100]}...")
if global_debug:
if get_global_debug():
print(f"RAW: Final text after cleanup: {final_text[:100]}...")
# If we have reasoning, continue with second pass to get more complete answer
......@@ -944,7 +946,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
# Additional cleanup of the full generated text
generated_text = cleanup_control_tokens(generated_text)
if global_debug:
if get_global_debug():
print(f"RAW: Generated text after cleanup: {generated_text[:100]}...")
# Pass through the formatter/parser (same as regular mode)
......@@ -1007,7 +1009,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
if extracted_tool_calls:
# Strip tool calls from the text
clean_text = adapter.strip_tool_calls_from_content(final_text)
if global_debug:
if get_global_debug():
print(f"RAW: Extracted {len(extracted_tool_calls)} tool calls from final_text (after reasoning)")
# Estimate token counts
......@@ -1027,7 +1029,7 @@ async def chat_completions(request: ChatCompletionRequest, http_request: Request
print(f"RAW: ERROR in formatter.format_full: {e}")
formatted_response = None
if global_debug:
if get_global_debug():
if formatted_response and isinstance(formatted_response, dict):
try:
choices = formatted_response.get('choices', [])
......@@ -1263,7 +1265,7 @@ async def stream_chat_response(
chunk_count = 0
# Debug: Print what is being passed to the model
if global_debug:
if get_global_debug():
print(f"\n{'='*80}")
print(f"=== MODEL INPUT (DEBUG) ===")
print(f"{'='*80}")
......@@ -1326,7 +1328,7 @@ async def stream_chat_response(
print(f"DEBUG: Warning - no content generated!")
# In debug mode, dump the full generated text
if global_debug:
if get_global_debug():
print(f"\n{'='*80}")
print(f"=== FULL GENERATED TEXT (DEBUG) ===")
print(f"{'='*80}")
......@@ -1386,7 +1388,7 @@ async def stream_chat_response(
tool_calls = None
if tool_calls:
# In debug mode, dump tool calls
if global_debug:
if get_global_debug():
print(f"\n{'='*80}")
print(f"=== EXTRACTED TOOL CALLS (DEBUG) ===")
print(f"{'='*80}")
......@@ -1500,7 +1502,7 @@ async def generate_chat_response(
created = int(time.time())
# Debug: Print what is being passed to the model
if global_debug:
if get_global_debug():
print(f"\n{'='*80}")
print(f"=== MODEL INPUT (DEBUG) ===")
print(f"{'='*80}")
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment