Emil_Shanaty/open-webui-ai4me
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

refac

Browse Source
This commit is contained in:
Timothy Jaeryang Baek
2026-01-26 18:04:58 +04:00
parent c5c4aef7b1
commit 4a55167759
2 changed files with 167 additions and 86 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/open_webui/utils/middleware.py
+146 -86
View File
@@ -73,6 +73,7 @@ from open_webui.models.models import Models
from open_webui.retrieval.utils import get_sources_from_items from open_webui.retrieval.utils import get_sources_from_items
from open_webui.utils.sanitize import strip_markdown_code_fences
from open_webui.utils.chat import generate_chat_completion from open_webui.utils.chat import generate_chat_completion
from open_webui.utils.task import ( from open_webui.utils.task import (
get_task_model_id, get_task_model_id,
@@ -1408,12 +1409,12 @@ async def convert_url_images_to_base64(form_data):
def process_messages_with_output(messages: list[dict]) -> list[dict]: def process_messages_with_output(messages: list[dict]) -> list[dict]:
""" """
Process messages with OR-aligned output items for LLM consumption. Process messages with OR-aligned output items for LLM consumption.
For assistant messages with 'output' field, produces properly formatted For assistant messages with 'output' field, produces properly formatted
OpenAI-style messages (tool_calls + tool results). Strips 'output' before LLM. OpenAI-style messages (tool_calls + tool results). Strips 'output' before LLM.
""" """
processed = [] processed = []
for message in messages: for message in messages:
if message.get("role") == "assistant" and message.get("output"): if message.get("role") == "assistant" and message.get("output"):
# Use output items for clean OpenAI-format messages # Use output items for clean OpenAI-format messages
@@ -1421,11 +1422,11 @@ def process_messages_with_output(messages: list[dict]) -> list[dict]:
if output_messages: if output_messages:
processed.extend(output_messages) processed.extend(output_messages)
continue continue
# Strip 'output' field before adding (LLM shouldn't see it) # Strip 'output' field before adding (LLM shouldn't see it)
clean_message = {k: v for k, v in message.items() if k != "output"} clean_message = {k: v for k, v in message.items() if k != "output"}
processed.append(clean_message) processed.append(clean_message)
return processed return processed
@@ -1841,9 +1842,8 @@ async def process_chat_payload(request, form_data, user, metadata, model):
# Inject builtin tools for native function calling based on enabled features and model capability # Inject builtin tools for native function calling based on enabled features and model capability
# Check if builtin_tools capability is enabled for this model (defaults to True if not specified) # Check if builtin_tools capability is enabled for this model (defaults to True if not specified)
builtin_tools_enabled = ( builtin_tools_enabled = (
(model.get("info", {}).get("meta", {}).get("capabilities") or {}) model.get("info", {}).get("meta", {}).get("capabilities") or {}
.get("builtin_tools", True) ).get("builtin_tools", True)
)
if ( if (
metadata.get("params", {}).get("function_calling") == "native" metadata.get("params", {}).get("function_calling") == "native"
and builtin_tools_enabled and builtin_tools_enabled
@@ -1887,9 +1887,8 @@ async def process_chat_payload(request, form_data, user, metadata, model):
# Check if file context extraction is enabled for this model (default True) # Check if file context extraction is enabled for this model (default True)
file_context_enabled = ( file_context_enabled = (
(model.get("info", {}).get("meta", {}).get("capabilities") or {}) model.get("info", {}).get("meta", {}).get("capabilities") or {}
.get("file_context", True) ).get("file_context", True)
)
if file_context_enabled: if file_context_enabled:
try: try:
@@ -2238,7 +2237,7 @@ async def process_chat_response(
) )
title = Chats.get_chat_title_by_id(metadata["chat_id"]) title = Chats.get_chat_title_by_id(metadata["chat_id"])
# Use output from backend if provided (OR-compliant backends) # Use output from backend if provided (OR-compliant backends)
response_output = response_data.get("output") response_output = response_data.get("output")
@@ -2248,7 +2247,11 @@ async def process_chat_response(
"data": { "data": {
"done": True, "done": True,
"content": content, "content": content,
**({"output": response_output} if response_output else {}), **(
{"output": response_output}
if response_output
else {}
),
"title": title, "title": title,
}, },
} }
@@ -2261,7 +2264,11 @@ async def process_chat_response(
{ {
"role": "assistant", "role": "assistant",
"content": content, "content": content,
**({"output": response_output} if response_output else {}), **(
{"output": response_output}
if response_output
else {}
),
}, },
) )
@@ -2521,33 +2528,33 @@ async def process_chat_response(
For LLM consumption, use convert_output_to_messages() instead. For LLM consumption, use convert_output_to_messages() instead.
""" """
content = "" content = ""
# First pass: collect function_call_output items by call_id for lookup # First pass: collect function_call_output items by call_id for lookup
tool_outputs = {} tool_outputs = {}
for item in output: for item in output:
if item.get("type") == "function_call_output": if item.get("type") == "function_call_output":
tool_outputs[item.get("call_id")] = item tool_outputs[item.get("call_id")] = item
# Second pass: render items in order # Second pass: render items in order
for item in output: for item in output:
item_type = item.get("type", "") item_type = item.get("type", "")
if item_type == "message": if item_type == "message":
for content_part in item.get("content", []): for content_part in item.get("content", []):
if content_part.get("type") == "output_text": if content_part.get("type") == "output_text":
text = content_part.get("text", "").strip() text = content_part.get("text", "").strip()
if text: if text:
content = f"{content}{text}\n" content = f"{content}{text}\n"
elif item_type == "function_call": elif item_type == "function_call":
# Render tool call inline with its result (if available) # Render tool call inline with its result (if available)
if content and not content.endswith("\n"): if content and not content.endswith("\n"):
content += "\n" content += "\n"
call_id = item.get("call_id", "") call_id = item.get("call_id", "")
name = item.get("name", "") name = item.get("name", "")
arguments = item.get("arguments", "") arguments = item.get("arguments", "")
result_item = tool_outputs.get(call_id) result_item = tool_outputs.get(call_id)
if result_item: if result_item:
result_text = "" result_text = ""
@@ -2556,59 +2563,65 @@ async def process_chat_response(
result_text += out.get("text", "") result_text += out.get("text", "")
files = result_item.get("files") files = result_item.get("files")
embeds = result_item.get("embeds", "") embeds = result_item.get("embeds", "")
content += f'<details type="tool_calls" done="true" id="{call_id}" name="{name}" arguments="{html.escape(json.dumps(arguments))}" result="{html.escape(json.dumps(result_text, ensure_ascii=False))}" files="{html.escape(json.dumps(files)) if files else ""}" embeds="{html.escape(json.dumps(embeds))}">\n<summary>Tool Executed</summary>\n</details>\n' content += f'<details type="tool_calls" done="true" id="{call_id}" name="{name}" arguments="{html.escape(json.dumps(arguments))}" result="{html.escape(json.dumps(result_text, ensure_ascii=False))}" files="{html.escape(json.dumps(files)) if files else ""}" embeds="{html.escape(json.dumps(embeds))}">\n<summary>Tool Executed</summary>\n</details>\n'
else: else:
content += f'<details type="tool_calls" done="false" id="{call_id}" name="{name}" arguments="{html.escape(json.dumps(arguments))}">\n<summary>Executing...</summary>\n</details>\n' content += f'<details type="tool_calls" done="false" id="{call_id}" name="{name}" arguments="{html.escape(json.dumps(arguments))}">\n<summary>Executing...</summary>\n</details>\n'
elif item_type == "function_call_output": elif item_type == "function_call_output":
# Already handled inline with function_call above # Already handled inline with function_call above
pass pass
elif item_type == "reasoning": elif item_type == "reasoning":
reasoning_content = "" reasoning_content = ""
for content_part in item.get("content", []): for content_part in item.get("content", []):
if content_part.get("type") == "output_text": if content_part.get("type") == "output_text":
reasoning_content = content_part.get("text", "").strip() reasoning_content = content_part.get("text", "").strip()
duration = item.get("duration") duration = item.get("duration")
status = item.get("status", "in_progress") status = item.get("status", "in_progress")
if content and not content.endswith("\n"): if content and not content.endswith("\n"):
content += "\n" content += "\n"
display = html.escape("\n".join( display = html.escape(
(f"> {line}" if not line.startswith(">") else line) "\n".join(
for line in reasoning_content.splitlines() (f"> {line}" if not line.startswith(">") else line)
)) for line in reasoning_content.splitlines()
)
)
if status == "completed" or duration is not None: if status == "completed" or duration is not None:
content = f'{content}<details type="reasoning" done="true" duration="{duration or 0}">\n<summary>Thought for {duration or 0} seconds</summary>\n{display}\n</details>\n' content = f'{content}<details type="reasoning" done="true" duration="{duration or 0}">\n<summary>Thought for {duration or 0} seconds</summary>\n{display}\n</details>\n'
else: else:
content = f'{content}<details type="reasoning" done="false">\n<summary>Thinking…</summary>\n{display}\n</details>\n' content = f'{content}<details type="reasoning" done="false">\n<summary>Thinking…</summary>\n{display}\n</details>\n'
elif item_type == "open_webui:code_interpreter": elif item_type == "open_webui:code_interpreter":
code = item.get("code", "") code = item.get("code", "")
output_val = item.get("output") output_val = item.get("output")
lang = item.get("lang", "") lang = item.get("lang", "")
content_stripped, original_whitespace = split_content_and_whitespace(content) content_stripped, original_whitespace = (
split_content_and_whitespace(content)
)
if is_opening_code_block(content_stripped): if is_opening_code_block(content_stripped):
content = content_stripped.rstrip("`").rstrip() + original_whitespace content = (
content_stripped.rstrip("`").rstrip()
+ original_whitespace
)
else: else:
content = content_stripped + original_whitespace content = content_stripped + original_whitespace
if content and not content.endswith("\n"): if content and not content.endswith("\n"):
content += "\n" content += "\n"
if output_val: if output_val:
output_escaped = html.escape(json.dumps(output_val)) output_escaped = html.escape(json.dumps(output_val))
content = f'{content}<details type="code_interpreter" done="true" output="{output_escaped}">\n<summary>Analyzed</summary>\n```{lang}\n{code}\n```\n</details>\n' content = f'{content}<details type="code_interpreter" done="true" output="{output_escaped}">\n<summary>Analyzed</summary>\n```{lang}\n{code}\n```\n</details>\n'
else: else:
content = f'{content}<details type="code_interpreter" done="false">\n<summary>Analyzing...</summary>\n```{lang}\n{code}\n```\n</details>\n' content = f'{content}<details type="code_interpreter" done="false">\n<summary>Analyzing...</summary>\n```{lang}\n{code}\n```\n</details>\n'
return content.strip()
return content.strip()
def convert_content_blocks_to_messages(content_blocks, raw=False): def convert_content_blocks_to_messages(content_blocks, raw=False):
messages = [] messages = []
@@ -2668,13 +2681,17 @@ async def process_chat_response(
if block_type == "text": if block_type == "text":
text_content = block.get("content", "").strip() text_content = block.get("content", "").strip()
if text_content: if text_content:
output_items.append({ output_items.append(
"type": "message", {
"id": block_id or next_id("msg"), "type": "message",
"status": "completed", "id": block_id or next_id("msg"),
"role": "assistant", "status": "completed",
"content": [{"type": "output_text", "text": text_content}], "role": "assistant",
}) "content": [
{"type": "output_text", "text": text_content}
],
}
)
elif block_type == "tool_calls": elif block_type == "tool_calls":
tool_calls = block.get("content", []) tool_calls = block.get("content", [])
@@ -2684,50 +2701,86 @@ async def process_chat_response(
for tool_call in tool_calls: for tool_call in tool_calls:
call_id = tool_call.get("id", "") call_id = tool_call.get("id", "")
func = tool_call.get("function", {}) func = tool_call.get("function", {})
output_items.append({ output_items.append(
"type": "function_call", {
"id": call_id or next_id("fc"), # Use call_id as item id if available "type": "function_call",
"call_id": call_id, "id": call_id
"name": func.get("name", ""), or next_id(
"arguments": func.get("arguments", "{}"), "fc"
"status": "completed" if results else "in_progress", ), # Use call_id as item id if available
}) "call_id": call_id,
"name": func.get("name", ""),
"arguments": func.get("arguments", "{}"),
"status": "completed" if results else "in_progress",
}
)
# Emit function_call_output items # Emit function_call_output items
for result in results: for result in results:
output_items.append({ output_items.append(
"type": "function_call_output", {
"id": result.get("id") or next_id("fco"), "type": "function_call_output",
"call_id": result.get("tool_call_id", ""), "id": result.get("id") or next_id("fco"),
"output": [{"type": "input_text", "text": result.get("content", "")}], "call_id": result.get("tool_call_id", ""),
"status": "completed", "output": [
**({"files": result.get("files")} if result.get("files") else {}), {
**({"embeds": result.get("embeds")} if result.get("embeds") else {}), "type": "input_text",
}) "text": result.get("content", ""),
}
],
"status": "completed",
**(
{"files": result.get("files")}
if result.get("files")
else {}
),
**(
{"embeds": result.get("embeds")}
if result.get("embeds")
else {}
),
}
)
elif block_type == "reasoning": elif block_type == "reasoning":
reasoning_content = block.get("content", "").strip() reasoning_content = block.get("content", "").strip()
duration = block.get("duration") duration = block.get("duration")
output_items.append({ output_items.append(
"type": "reasoning", {
"id": block_id or next_id("r"), "type": "reasoning",
"status": "completed" if duration is not None else "in_progress", "id": block_id or next_id("r"),
"content": [{"type": "output_text", "text": reasoning_content}] if reasoning_content else None, "status": (
"summary": None, "completed"
}) if duration is not None
else "in_progress"
),
"content": (
[{"type": "output_text", "text": reasoning_content}]
if reasoning_content
else None
),
"summary": None,
}
)
elif block_type == "code_interpreter": elif block_type == "code_interpreter":
code = block.get("content", "") code = block.get("content", "")
output_val = block.get("output") output_val = block.get("output")
attrs = block.get("attributes", {}) attrs = block.get("attributes", {})
output_items.append({ output_items.append(
"type": "open_webui:code_interpreter", {
"id": block_id or next_id("ci"), "type": "open_webui:code_interpreter",
"status": "completed" if output_val is not None else "in_progress", "id": block_id or next_id("ci"),
"lang": attrs.get("lang", ""), "status": (
"code": code, "completed"
"output": output_val, if output_val is not None
}) else "in_progress"
),
"lang": attrs.get("lang", ""),
"code": code,
"output": output_val,
}
)
return output_items return output_items
@@ -2934,14 +2987,16 @@ async def process_chat_response(
output = existing_output output = existing_output
else: else:
# Always create an initial message item (even if content is empty) # Always create an initial message item (even if content is empty)
output = [{ output = [
"type": "message", {
"id": output_id("msg"), "type": "message",
"status": "in_progress", "id": output_id("msg"),
"role": "assistant", "status": "in_progress",
"content": [{"type": "output_text", "text": content}], "role": "assistant",
}] "content": [{"type": "output_text", "text": content}],
}
]
# Keep content_blocks for backward compatibility during transition # Keep content_blocks for backward compatibility during transition
content_blocks = [ content_blocks = [
{ {
@@ -3354,7 +3409,9 @@ async def process_chat_response(
if ENABLE_REALTIME_CHAT_SAVE: if ENABLE_REALTIME_CHAT_SAVE:
# Save message in the database # Save message in the database
output = convert_content_blocks_to_output(content_blocks) output = convert_content_blocks_to_output(
content_blocks
)
Chats.upsert_message_to_chat_by_id_and_message_id( Chats.upsert_message_to_chat_by_id_and_message_id(
metadata["chat_id"], metadata["chat_id"],
metadata["message_id"], metadata["message_id"],
@@ -3692,6 +3749,9 @@ async def process_chat_response(
try: try:
if content_blocks[-1]["attributes"].get("type") == "code": if content_blocks[-1]["attributes"].get("type") == "code":
code = content_blocks[-1]["content"] code = content_blocks[-1]["content"]
# Strip markdown fences if model included them
code = strip_markdown_code_fences(code)
if CODE_INTERPRETER_BLOCKED_MODULES: if CODE_INTERPRETER_BLOCKED_MODULES:
blocking_code = textwrap.dedent( blocking_code = textwrap.dedent(
f""" f"""
backend/open_webui/utils/sanitize.py
+21
View File
@@ -0,0 +1,21 @@
import re
def strip_markdown_code_fences(code: str) -> str:
"""
Strip markdown code fences if present.
This is a defensive, non-breaking change — if the code doesn't
contain fences, it passes through unchanged.
Handles patterns like:
- ```python
- ```py
- ```
"""
code = code.strip()
# Remove opening fence (```python, ```py, ``` etc.)
code = re.sub(r"^```\w*\n?", "", code)
# Remove closing fence
code = re.sub(r"\n?```\s*$", "", code)
return code.strip()