feat: enhance comfyui workflow (#10085)

This commit is contained in:
非法操作 2024-10-31 10:00:22 +08:00 committed by GitHub
parent 6692e8c508
commit bd6175157c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 68 additions and 24 deletions

View File

@ -1,5 +1,3 @@
import base64
import io
import json
import random
import uuid
@ -8,7 +6,7 @@ import httpx
from websocket import WebSocket
from yarl import URL
from core.file.file_manager import _get_encoded_string
from core.file.file_manager import download
from core.file.models import File
@ -29,8 +27,7 @@ class ComfyUiClient:
return response.content
def upload_image(self, image_file: File) -> dict:
image_content = base64.b64decode(_get_encoded_string(image_file))
file = io.BytesIO(image_content)
file = download(image_file)
files = {"image": (image_file.filename, file, image_file.mime_type), "overwrite": "true"}
res = httpx.post(str(self.base_url / "upload/image"), files=files)
return res.json()
@ -47,12 +44,7 @@ class ComfyUiClient:
ws.connect(ws_address)
return ws, client_id
def set_prompt(
self, origin_prompt: dict, positive_prompt: str, negative_prompt: str = "", image_name: str = ""
) -> dict:
"""
find the first KSampler, then can find the prompt node through it.
"""
def set_prompt_by_ksampler(self, origin_prompt: dict, positive_prompt: str, negative_prompt: str = "") -> dict:
prompt = origin_prompt.copy()
id_to_class_type = {id: details["class_type"] for id, details in prompt.items()}
k_sampler = [key for key, value in id_to_class_type.items() if value == "KSampler"][0]
@ -64,9 +56,20 @@ class ComfyUiClient:
negative_input_id = prompt.get(k_sampler)["inputs"]["negative"][0]
prompt.get(negative_input_id)["inputs"]["text"] = negative_prompt
if image_name != "":
image_loader = [key for key, value in id_to_class_type.items() if value == "LoadImage"][0]
prompt.get(image_loader)["inputs"]["image"] = image_name
return prompt
def set_prompt_images_by_ids(self, origin_prompt: dict, image_names: list[str], image_ids: list[str]) -> dict:
prompt = origin_prompt.copy()
for index, image_node_id in enumerate(image_ids):
prompt[image_node_id]["inputs"]["image"] = image_names[index]
return prompt
def set_prompt_images_by_default(self, origin_prompt: dict, image_names: list[str]) -> dict:
prompt = origin_prompt.copy()
id_to_class_type = {id: details["class_type"] for id, details in prompt.items()}
load_image_nodes = [key for key, value in id_to_class_type.items() if value == "LoadImage"]
for load_image, image_name in zip(load_image_nodes, image_names):
prompt.get(load_image)["inputs"]["image"] = image_name
return prompt
def track_progress(self, prompt: dict, ws: WebSocket, prompt_id: str):

View File

@ -1,7 +1,9 @@
import json
from typing import Any
from core.file import FileType
from core.tools.entities.tool_entities import ToolInvokeMessage
from core.tools.errors import ToolParameterValidationError
from core.tools.provider.builtin.comfyui.tools.comfyui_client import ComfyUiClient
from core.tools.tool.builtin_tool import BuiltinTool
@ -10,19 +12,46 @@ class ComfyUIWorkflowTool(BuiltinTool):
def _invoke(self, user_id: str, tool_parameters: dict[str, Any]) -> ToolInvokeMessage | list[ToolInvokeMessage]:
comfyui = ComfyUiClient(self.runtime.credentials["base_url"])
positive_prompt = tool_parameters.get("positive_prompt")
negative_prompt = tool_parameters.get("negative_prompt")
positive_prompt = tool_parameters.get("positive_prompt", "")
negative_prompt = tool_parameters.get("negative_prompt", "")
images = tool_parameters.get("images") or []
workflow = tool_parameters.get("workflow_json")
image_name = ""
if image := tool_parameters.get("image"):
image_names = []
for image in images:
if image.type != FileType.IMAGE:
continue
image_name = comfyui.upload_image(image).get("name")
image_names.append(image_name)
set_prompt_with_ksampler = True
if "{{positive_prompt}}" in workflow:
set_prompt_with_ksampler = False
workflow = workflow.replace("{{positive_prompt}}", positive_prompt)
workflow = workflow.replace("{{negative_prompt}}", negative_prompt)
try:
origin_prompt = json.loads(workflow)
prompt = json.loads(workflow)
except:
return self.create_text_message("the Workflow JSON is not correct")
prompt = comfyui.set_prompt(origin_prompt, positive_prompt, negative_prompt, image_name)
if set_prompt_with_ksampler:
try:
prompt = comfyui.set_prompt_by_ksampler(prompt, positive_prompt, negative_prompt)
except:
raise ToolParameterValidationError(
"Failed set prompt with KSampler, try replace prompt to {{positive_prompt}} in your workflow json"
)
if image_names:
if image_ids := tool_parameters.get("image_ids"):
image_ids = image_ids.split(",")
try:
prompt = comfyui.set_prompt_images_by_ids(prompt, image_names, image_ids)
except:
raise ToolParameterValidationError("the Image Node ID List not match your upload image files.")
else:
prompt = comfyui.set_prompt_images_by_default(prompt, image_names)
images = comfyui.generate_image_by_prompt(prompt)
result = []
for img in images:

View File

@ -24,12 +24,12 @@ parameters:
zh_Hans: 负面提示词
llm_description: Negative prompt, you should describe the image you don't want to generate as a list of words as possible as detailed, the prompt must be written in English.
form: llm
- name: image
type: file
- name: images
type: files
label:
en_US: Input Image
en_US: Input Images
zh_Hans: 输入的图片
llm_description: The input image, used to transfer to the comfyui workflow to generate another image.
llm_description: The input images, used to transfer to the comfyui workflow to generate another image.
form: llm
- name: workflow_json
type: string
@ -40,3 +40,15 @@ parameters:
en_US: exported from ComfyUI workflow
zh_Hans: 从ComfyUI的工作流中导出
form: form
- name: image_ids
type: string
label:
en_US: Image Node ID List
zh_Hans: 图片节点ID列表
placeholder:
en_US: Use commas to separate multiple node ID
zh_Hans: 多个节点ID时使用半角逗号分隔
human_description:
en_US: When the workflow has multiple image nodes, enter the ID list of these nodes, and the images will be passed to ComfyUI in the order of the list.
zh_Hans: 当工作流有多个图片节点时输入这些节点的ID列表图片将按列表顺序传给ComfyUI
form: form