docker_test / app.py
朱东升
update10
33da27c
raw
history blame
37.1 kB
import gradio as gr
import json
import importlib
import os
import sys
import time
from pathlib import Path
import concurrent.futures
import multiprocessing
import threading
import queue
from datetime import datetime, timedelta
from src.containerized_eval import eval_string_script
# 添加当前目录和src目录到模块搜索路径
current_dir = os.path.dirname(os.path.abspath(__file__))
src_dir = os.path.join(current_dir, "src")
if current_dir not in sys.path:
sys.path.append(current_dir)
if src_dir not in sys.path:
sys.path.append(src_dir)
# 添加全局任务队列和任务状态跟踪
task_queue = queue.Queue()
task_results = {}
active_tasks = {}
completed_tasks = []
task_id_counter = 0
task_lock = threading.Lock()
last_update_time = datetime.now() # 替换update_event为时间戳跟踪
# 持久化的状态用于API访问和轮询更新
status_poll_counter = 0 # 用于强制更新UI,即使状态没有变化
last_background_update = datetime.now() # 最后一次后台更新时间
def trigger_ui_update():
"""触发UI更新事件"""
global last_update_time, status_poll_counter, last_background_update
with task_lock:
last_update_time = datetime.now() # 更新时间戳
status_poll_counter += 1 # 递增计数器,确保每次更新都被捕获
last_background_update = last_update_time # 更新后台状态
print(f"UI更新被触发: {datetime.now().strftime('%H:%M:%S')} [计数: {status_poll_counter}]")
def get_next_task_id():
global task_id_counter
with task_lock:
task_id_counter += 1
return f"task_{task_id_counter}"
def submit_task(input_data):
"""提交任务到队列
Args:
input_data: 列表(批量处理多个测试用例)
Returns:
str: 任务ID
"""
try:
if not isinstance(input_data, list):
return {"status": "error", "message": "Input must be a list"}
task_id = get_next_task_id()
with task_lock:
estimated_time = estimate_completion_time(input_data)
task_info = {
"id": task_id,
"data": input_data,
"status": "queued",
"submitted_at": datetime.now(),
"estimated_completion_time": estimated_time,
"items_count": len(input_data)
}
active_tasks[task_id] = task_info
task_queue.put(task_info)
# 触发UI更新
trigger_ui_update()
# 如果这是第一个任务,启动处理线程
if len(active_tasks) == 1:
threading.Thread(target=process_task_queue, daemon=True).start()
return {"status": "success", "task_id": task_id}
except Exception as e:
return {"status": "error", "message": str(e)}
def estimate_completion_time(input_data):
"""估计完成任务所需的时间
Args:
input_data: 任务数据
Returns:
timedelta: 估计的完成时间
"""
# 在Hugging Face Spaces环境中,资源通常受限,调整处理时间预估
# 假设每个任务项平均需要5秒处理(HF环境中可能更慢)
avg_time_per_item = 5
total_items = len(input_data)
# Hugging Face Spaces通常有限制的CPU资源
# 保守估计并行处理能力
try:
cpu_count = multiprocessing.cpu_count()
except:
# 如果获取失败,假设只有2个CPU
cpu_count = 2
# 在HF环境中,即使有多核也可能性能受限,降低并行因子
parallel_factor = min(2, total_items) # 限制最多2个并行任务
if parallel_factor > 0:
estimated_seconds = (total_items * avg_time_per_item) / parallel_factor
# 为了避免过于乐观的估计,增加30%的缓冲时间
estimated_seconds = estimated_seconds * 1.3
return timedelta(seconds=round(estimated_seconds))
else:
return timedelta(seconds=0)
def process_task_queue():
"""处理任务队列的后台线程"""
while True:
try:
if task_queue.empty():
time.sleep(0.5)
continue
task_info = task_queue.get()
task_id = task_info["id"]
# 更新任务状态
with task_lock:
if task_id in active_tasks: # 确保任务仍存在
active_tasks[task_id]["status"] = "processing"
print(f"任务 {task_id} 开始处理,当前时间: {datetime.now().strftime('%H:%M:%S')}")
trigger_ui_update() # 状态变更为处理中时更新UI
else:
print(f"警告: 任务 {task_id} 不在活跃任务列表中")
task_queue.task_done()
continue
# 处理任务
print(f"开始评估任务 {task_id},数据项数: {len(task_info['data'])}")
result = evaluate(task_info["data"])
print(f"任务 {task_id} 评估完成,结果数: {len(result) if isinstance(result, list) else 'Not a list'}")
# 更新任务结果
with task_lock:
if task_id in active_tasks: # 再次确保任务仍存在
completed_time = datetime.now()
active_tasks[task_id]["status"] = "completed"
active_tasks[task_id]["completed_at"] = completed_time
active_tasks[task_id]["result"] = result
# 计算处理持续时间
start_time = active_tasks[task_id]["submitted_at"]
duration = completed_time - start_time
print(f"任务 {task_id} 已完成,耗时: {duration},当前时间: {completed_time.strftime('%H:%M:%S')}")
# 将任务移至已完成列表
completed_tasks.append(active_tasks[task_id])
del active_tasks[task_id]
# 保留最近的20个已完成任务
if len(completed_tasks) > 20:
completed_tasks.pop(0)
# 状态更新后强制触发UI更新
trigger_ui_update()
print(f"任务 {task_id} 完成后触发UI更新: {last_update_time}")
else:
print(f"警告: 任务 {task_id} 在处理完成后不在活跃任务列表中")
task_queue.task_done()
except Exception as e:
print(f"处理任务队列出错: {str(e)}")
import traceback
traceback.print_exc()
time.sleep(1)
def evaluate(input_data):
"""评估代码的主函数
Args:
input_data: 列表(批量处理多个测试用例)
Returns:
list: 包含评估结果的列表
"""
# 打印Gradio版本,用于调试
import gradio
print(f"Gradio version: {gradio.__version__}")
try:
if not isinstance(input_data, list):
return {"status": "Exception", "error": "Input must be a list"}
results = []
# 在HF Spaces环境中可能受限,降低并行数量
try:
max_workers = min(multiprocessing.cpu_count(), 2) # 最多2个并行任务
except:
max_workers = 2 # 如果无法获取,默认为2
# 增加超时处理
with concurrent.futures.ThreadPoolExecutor(max_workers=max_workers) as executor:
future_to_item = {}
# 分批处理,每批最多5个任务,避免资源耗尽
batch_size = 5
for i in range(0, len(input_data), batch_size):
batch = input_data[i:i+batch_size]
# 为每个任务提交并记录
for item in batch:
future = executor.submit(evaluate_single_case, item)
future_to_item[future] = item
# 等待当前批次完成
for future in concurrent.futures.as_completed(future_to_item):
item = future_to_item[future]
try:
# 设置较短的超时时间,避免任务卡死
result = future.result(timeout=60) # 60秒超时
item.update(result)
results.append(item)
except concurrent.futures.TimeoutError:
# 处理超时情况
item.update({
"status": "Timeout",
"error": "Task processing timed out in Hugging Face environment"
})
results.append(item)
except Exception as e:
# 处理其他异常
item.update({
"status": "Exception",
"error": f"Error in Hugging Face environment: {str(e)}"
})
results.append(item)
# 清空当前批次
future_to_item = {}
# 短暂休息,让系统喘息
time.sleep(0.5)
return results
except Exception as e:
return {"status": "Exception", "error": f"Evaluation error in Hugging Face environment: {str(e)}"}
def evaluate_single_case(input_data):
"""评估单个代码用例
Args:
input_data: 字典(包含代码信息)
Returns:
dict: 包含评估结果的字典
"""
try:
if not isinstance(input_data, dict):
return {"status": "Exception", "error": "Input item must be a dictionary"}
language = input_data.get('language')
completions = input_data.get('processed_completions', [])
if not completions:
return {"status": "Exception", "error": "No code provided"}
results = []
for comp in completions:
code = input_data.get('prompt') + comp + '\n' + input_data.get('tests')
result = evaluate_code(code, language)
if result["status"] == "OK":
return result
results.append(result)
return results[0]
except Exception as e:
return {"status": "Exception", "error": str(e)}
def evaluate_code(code, language):
"""评估特定语言的代码
Args:
code (str): 要评估的代码
language (str): 编程语言
Returns:
dict: 包含评估结果的字典
"""
try:
# 使用containerized_eval中的eval_string_script函数
result = eval_string_script(language, code)
return result
except Exception as e:
return {"status": "Exception", "error": str(e)}
def get_queue_status():
"""获取当前队列状态
Returns:
dict: 包含队列状态的字典
"""
with task_lock:
queued_tasks = [task for task in active_tasks.values() if task["status"] == "queued"]
processing_tasks = [task for task in active_tasks.values() if task["status"] == "processing"]
# 计算总的预计完成时间
total_estimated_time = timedelta(seconds=0)
for task in active_tasks.values():
if isinstance(task["estimated_completion_time"], timedelta):
total_estimated_time += task["estimated_completion_time"]
# 计算队列中的项目总数
total_items = sum(task["items_count"] for task in active_tasks.values())
return {
"queued_tasks": len(queued_tasks),
"processing_tasks": len(processing_tasks),
"total_tasks": len(active_tasks),
"total_items": total_items,
"estimated_completion_time": str(total_estimated_time),
"active_tasks": [
{
"id": task["id"],
"status": task["status"],
"items_count": task["items_count"],
"submitted_at": task["submitted_at"].strftime("%Y-%m-%d %H:%M:%S"),
"estimated_completion": str(task["estimated_completion_time"])
} for task in active_tasks.values()
],
"recent_completed": [
{
"id": task["id"],
"items_count": task["items_count"],
"submitted_at": task["submitted_at"].strftime("%Y-%m-%d %H:%M:%S"),
"completed_at": task["completed_at"].strftime("%Y-%m-%d %H:%M:%S") if "completed_at" in task else "",
"duration": str(task["completed_at"] - task["submitted_at"]) if "completed_at" in task else ""
} for task in completed_tasks[-5:] # 只显示最近5个完成的任务
],
"poll_counter": status_poll_counter, # 添加轮询计数器
"timestamp": datetime.now().strftime('%Y-%m-%d %H:%M:%S.%f')
}
def render_queue_status():
"""渲染队列状态UI
Returns:
str: HTML格式的队列状态显示
"""
status = get_queue_status()
html = f"""
<div style="font-family: Arial, sans-serif; max-width: 900px; margin: 0 auto; padding: 20px; background: #f9f9f9; border-radius: 10px; box-shadow: 0 4px 12px rgba(0,0,0,0.1);">
<div style="display: flex; justify-content: space-between; margin-bottom: 20px;">
<div style="background: #fff; padding: 15px; border-radius: 8px; width: 30%; box-shadow: 0 2px 6px rgba(0,0,0,0.05);">
<h3 style="margin: 0 0 10px; color: #444; font-size: 16px;">排队任务</h3>
<p style="font-size: 28px; font-weight: bold; margin: 0; color: #3498db;">{status['queued_tasks']}</p>
</div>
<div style="background: #fff; padding: 15px; border-radius: 8px; width: 30%; box-shadow: 0 2px 6px rgba(0,0,0,0.05);">
<h3 style="margin: 0 0 10px; color: #444; font-size: 16px;">处理中任务</h3>
<p style="font-size: 28px; font-weight: bold; margin: 0; color: #e74c3c;">{status['processing_tasks']}</p>
</div>
<div style="background: #fff; padding: 15px; border-radius: 8px; width: 30%; box-shadow: 0 2px 6px rgba(0,0,0,0.05);">
<h3 style="margin: 0 0 10px; color: #444; font-size: 16px;">总项目数</h3>
<p style="font-size: 28px; font-weight: bold; margin: 0; color: #2ecc71;">{status['total_items']}</p>
</div>
</div>
<div style="background: #fff; padding: 20px; border-radius: 8px; margin-bottom: 20px; box-shadow: 0 2px 6px rgba(0,0,0,0.05);">
<h2 style="margin: 0 0 15px; color: #333; font-size: 18px;">预计完成时间</h2>
<div style="display: flex; align-items: center;">
<div style="font-size: 24px; font-weight: bold; color: #9b59b6;">{status['estimated_completion_time']}</div>
<div style="margin-left: 10px; color: #777; font-size: 14px;">(小时:分钟:秒)</div>
</div>
</div>
<div style="background: #fff; padding: 20px; border-radius: 8px; margin-bottom: 20px; box-shadow: 0 2px 6px rgba(0,0,0,0.05);">
<h2 style="margin: 0 0 15px; color: #333; font-size: 18px;">活跃任务</h2>
<table style="width: 100%; border-collapse: collapse;">
<thead>
<tr style="background: #f2f2f2;">
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">任务ID</th>
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">状态</th>
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">项目数</th>
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">提交时间</th>
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">预计完成</th>
</tr>
</thead>
<tbody>
"""
if status['active_tasks']:
for task in status['active_tasks']:
status_color = "#3498db" if task['status'] == "queued" else "#e74c3c"
html += f"""
<tr>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['id']}</td>
<td style="padding: 12px; border-bottom: 1px solid #eee;">
<span style="padding: 4px 8px; border-radius: 4px; font-size: 12px; background: {status_color}; color: white;">
{task['status']}
</span>
</td>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['items_count']}</td>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['submitted_at']}</td>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['estimated_completion']}</td>
</tr>
"""
else:
html += f"""
<tr>
<td colspan="5" style="padding: 15px; text-align: center; color: #777;">当前没有活跃任务</td>
</tr>
"""
html += """
</tbody>
</table>
</div>
<div style="background: #fff; padding: 20px; border-radius: 8px; box-shadow: 0 2px 6px rgba(0,0,0,0.05);">
<h2 style="margin: 0 0 15px; color: #333; font-size: 18px;">最近完成的任务</h2>
<table style="width: 100%; border-collapse: collapse;">
<thead>
<tr style="background: #f2f2f2;">
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">任务ID</th>
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">项目数</th>
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">提交时间</th>
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">完成时间</th>
<th style="padding: 12px; text-align: left; border-bottom: 1px solid #ddd;">持续时间</th>
</tr>
</thead>
<tbody>
"""
if status['recent_completed']:
for task in status['recent_completed']:
html += f"""
<tr>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['id']}</td>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['items_count']}</td>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['submitted_at']}</td>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['completed_at']}</td>
<td style="padding: 12px; border-bottom: 1px solid #eee;">{task['duration']}</td>
</tr>
"""
else:
html += f"""
<tr>
<td colspan="5" style="padding: 15px; text-align: center; color: #777;">暂无已完成任务</td>
</tr>
"""
html += """
</tbody>
</table>
</div>
</div>
"""
return html
def refresh_ui():
"""定期刷新UI函数,确保显示最新状态"""
global status_poll_counter
print(f"UI刷新被调用: {datetime.now().strftime('%H:%M:%S')} [状态计数: {status_poll_counter}]")
return render_queue_status(), status_poll_counter
def get_status_for_polling():
"""为AJAX轮询提供的状态获取函数,简化返回数据量"""
global status_poll_counter
# 返回最新状态和计数器值,用于客户端判断是否有更新
return {
"status_html": render_queue_status(),
"poll_counter": status_poll_counter,
"timestamp": datetime.now().strftime('%Y-%m-%d %H:%M:%S.%f')
}
def submit_json_data(json_data):
"""提交JSON数据处理接口"""
try:
# 解析JSON字符串
if isinstance(json_data, str):
data = json.loads(json_data)
else:
data = json_data
# 为调试记录提交来源
print(f"接收到任务提交: {len(data) if isinstance(data, list) else 'Not a list'} 项")
print(f"当前队列状态 (提交前): 活跃任务 {len(active_tasks)}, 已完成任务 {len(completed_tasks)}")
result = submit_task(data)
response = json.dumps(result, ensure_ascii=False, indent=2)
# 强制触发UI更新
trigger_ui_update()
print(f"任务提交后触发UI更新: {last_update_time}")
print(f"当前队列状态 (提交后): 活跃任务 {len(active_tasks)}, 已完成任务 {len(completed_tasks)}")
# 返回任务提交结果和最新的队列状态
status_html = render_queue_status()
return response, status_html
except Exception as e:
print(f"提交任务出错: {str(e)}")
return json.dumps({"status": "error", "message": str(e)}, ensure_ascii=False, indent=2), render_queue_status()
# API端点函数,用于获取最新队列状态
def api_get_queue_status():
"""API端点,获取最新队列状态"""
status = get_queue_status()
status["html"] = render_queue_status() # 添加HTML渲染版本
return status
# 后台UI更新线程
def background_ui_updater():
"""每隔一段时间检查任务状态,如有变化则触发更新"""
global last_background_update
print("后台UI更新线程已启动")
while True:
try:
# 检查队列状态
current_time = datetime.now()
time_since_last_update = (current_time - last_background_update).total_seconds()
# 如果30秒内没有更新,且有活跃任务,触发更新
with task_lock:
has_active_tasks = len(active_tasks) > 0
if time_since_last_update > 30 and has_active_tasks:
print(f"后台更新: 有{len(active_tasks)}个活跃任务,{time_since_last_update:.1f}秒未更新")
trigger_ui_update()
# 如果2分钟内没有更新,无论如何都触发一次更新保持UI活跃
if time_since_last_update > 120:
print(f"后台更新: 保活触发,{time_since_last_update:.1f}秒未更新")
trigger_ui_update()
except Exception as e:
print(f"后台更新出错: {str(e)}")
# 每5秒检查一次
time.sleep(5)
# 创建Gradio接口
with gr.Blocks(title="代码评估服务", theme=gr.themes.Soft()) as demo:
gr.Markdown("""
# 代码评估服务
### 支持多种编程语言的代码评估服务
""")
# 添加隐藏的API处理组件
with gr.Row(visible=False):
api_input = gr.JSON()
api_output = gr.JSON()
# 设置API触发器
def api_trigger(data):
"""处理API请求的函数"""
print(f"通过API接收到请求: {len(data) if isinstance(data, list) else 'Not a list'}")
try:
result = submit_task(data)
# 强制触发UI更新
trigger_ui_update()
return result
except Exception as e:
print(f"API处理出错: {str(e)}")
return {"status": "error", "message": str(e)}
api_input.change(fn=api_trigger, inputs=api_input, outputs=api_output)
with gr.Tab("任务队列状态"):
status_html = gr.HTML(render_queue_status)
poll_counter = gr.Number(value=0, visible=False) # 隐藏的计数器,用于触发更新
refresh_button = gr.Button("刷新状态")
status_text = gr.Markdown("上次更新时间: 未更新")
# 根据Gradio版本使用不同的事件注册方法
if 'gradio_version' not in locals():
import gradio
gradio_version = getattr(gradio, "__version__", "unknown")
if gradio_version.startswith("3."):
# Gradio 3.x 方式
refresh_button.click(fn=refresh_ui, outputs=[status_html, poll_counter], concurrency_limit=2)
else:
# Gradio 4.x 方式 (不使用concurrency_limit参数)
refresh_button.click(
fn=refresh_ui,
outputs=[status_html, poll_counter],
every=10 # 每10秒自动刷新一次
)
# API断点,用于轮询最新状态
status_polling_input = gr.Number(value=0, visible=False, label="轮询触发器")
status_polling_output = gr.JSON(visible=False, label="轮询结果")
status_polling_input.change(
fn=get_status_for_polling,
inputs=[],
outputs=status_polling_output
)
# 使用JavaScript实现自动轮询
polling_js = """
<script>
// 初始化轮询计数器
let lastPollCounter = 0;
let pollingActive = true;
// 轮询函数
async function pollStatus() {
if (!pollingActive) return;
try {
// 查找JSON输出元素 (在Gradio 4.x中,元素ID可能不同)
const jsonElements = document.querySelectorAll('[data-testid="json"]');
const statusElements = document.querySelectorAll('[id*="status_html"]');
// 如果找不到元素,使用普通AJAX请求
if (!jsonElements.length) {
// 通过已有的刷新按钮点击来刷新
const refreshButton = [...document.querySelectorAll('button')].find(btn =>
btn.textContent.includes('刷新状态'));
if (refreshButton) {
refreshButton.click();
console.log("刷新按钮被触发");
}
} else {
// 使用Gradio API调用来获取最新状态
// 构建API请求的路径
const apiBase = window.location.pathname.endsWith('/')
? window.location.pathname
: window.location.pathname + '/';
const response = await fetch(apiBase + 'api/queue/status');
if (response.ok) {
const data = await response.json();
const statusHtml = statusElements[0];
// 如果有状态更新,更新UI
if (statusHtml && data.status) {
// 根据API返回更新页面状态
console.log("服务器状态已更新");
// 触发刷新按钮
const refreshButton = [...document.querySelectorAll('button')].find(btn =>
btn.textContent.includes('刷新状态'));
if (refreshButton) {
refreshButton.click();
}
}
}
}
} catch (err) {
console.error("轮询错误:", err);
}
// 继续轮询
setTimeout(pollStatus, 5000); // 每5秒轮询一次
}
// 页面加载完成后开始轮询
if (document.readyState === 'loading') {
document.addEventListener('DOMContentLoaded', () => setTimeout(pollStatus, 1000));
} else {
setTimeout(pollStatus, 1000);
}
// 页面可见性变化时管理轮询
document.addEventListener('visibilitychange', function() {
pollingActive = document.visibilityState === 'visible';
if (pollingActive) {
// 页面变为可见时立即轮询一次
pollStatus();
}
});
</script>
"""
gr.HTML(polling_js)
# 以下是原来的自动刷新脚本,保留但不使用
auto_refresh_js = """
<script>
// 兼容Gradio 3.x和4.x的自动刷新机制 - 仅作为备用
console.log('自动刷新机制已加载,但已被新的轮询系统替代');
</script>
"""
# 不显示旧的自动刷新脚本
# gr.HTML(auto_refresh_js)
with gr.Tab("提交新任务"):
with gr.Row():
with gr.Column():
json_input = gr.Textbox(
label="输入JSON数据",
placeholder='[{"language": "python", "prompt": "def add(a, b):\\n", "processed_completions": [" return a + b"], "tests": "assert add(1, 2) == 3"}]',
lines=10
)
submit_button = gr.Button("提交任务")
with gr.Column():
result_output = gr.Textbox(label="提交结果", lines=5)
# 根据Gradio版本使用不同的事件注册方法
if gradio_version.startswith("3."):
# Gradio 3.x 方式
submit_button.click(fn=submit_json_data, inputs=json_input, outputs=[result_output, status_html], concurrency_limit=2)
else:
# Gradio 4.x 方式
submit_button.click(fn=submit_json_data, inputs=json_input, outputs=[result_output, status_html])
with gr.Tab("API文档"):
gr.Markdown("""
## API 文档
### 1. 提交任务
**请求:**
```
POST /api/predict
Content-Type: application/json
[
{
"language": "python",
"prompt": "def add(a, b):\\n",
"processed_completions": [" return a + b"],
"tests": "assert add(1, 2) == 3"
}
]
```
**响应:**
```json
{
"status": "success",
"task_id": "task_1"
}
```
### 2. 查询任务状态
**请求:**
```
GET /api/status
```
**响应:**
```json
{
"queued_tasks": 1,
"processing_tasks": 2,
"total_tasks": 3,
"total_items": 15,
"estimated_completion_time": "0:05:30",
"active_tasks": [...],
"recent_completed": [...]
}
```
""")
# 这里不再添加状态API端点,避免与FastAPI冲突
# demo.queue(api_open=True).add_api_route("/api/queue/status", api_get_queue_status, methods=["GET"])
if __name__ == "__main__":
# 检测Gradio版本以适配不同版本的API
import gradio
gradio_version = getattr(gradio, "__version__", "unknown")
print(f"当前Gradio版本: {gradio_version}")
# 启动后台UI更新线程
bg_thread = threading.Thread(target=background_ui_updater, daemon=True)
bg_thread.start()
print("后台UI更新线程已启动")
try:
# 设置标准日志记录,确保足够的调试信息
import logging
logging.basicConfig(level=logging.INFO,
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
logger = logging.getLogger("CodeEvalService")
logger.info(f"启动代码评估服务,Gradio版本: {gradio_version}")
# 尝试使用兼容所有版本的参数启动
launch_kwargs = {
"server_name": "0.0.0.0",
"server_port": int(os.environ.get("PORT", 7860)),
"share": False,
}
# Gradio 4.x专用的FastAPI初始化
if not gradio_version.startswith("3."):
# 针对Gradio 4的API配置
import fastapi
from fastapi import FastAPI
# 创建FastAPI应用
app = FastAPI(title="代码评估服务API")
# 添加队列状态API端点
@app.get("/api/queue/status")
def get_queue_status_api():
return api_get_queue_status()
# 添加任务提交API端点
@app.post("/api/submit_task")
async def submit_task_api(data: list):
try:
result = submit_task(data)
return result
except Exception as e:
return {"status": "error", "message": str(e)}
# 添加/evaluate API端点
@app.post("/evaluate")
async def evaluate_api(data: list):
try:
result = evaluate(data)
return result
except Exception as e:
return {"status": "error", "message": str(e)}
# 启动应用到FastAPI
demo.launch(
**launch_kwargs,
app=app,
max_threads=5,
)
else:
# Gradio 3.x的方式
# 设置队列,并添加API支持
try:
demo.queue(api_open=True, max_size=30)
# 添加/evaluate API端点
demo.add_api_route("/evaluate", evaluate, methods=["POST"])
except Exception as e:
logger.warning(f"配置队列时出错: {e}")
# 启动应用
demo.launch(
**launch_kwargs,
debug=False,
show_api=True,
max_threads=5,
concurrency_limit=2
)
except Exception as e:
print(f"启动时发生错误: {e}")
import traceback
traceback.print_exc()
# 尝试最简配置启动
try:
print("使用最小配置重试...")
demo.launch(server_name="0.0.0.0", server_port=int(os.environ.get("PORT", 7860)))
except Exception as e2:
print(f"最小配置启动也失败: {e2}")
traceback.print_exc()
# 终极回退方案:创建最简单的接口并启动
try:
print("尝试创建备用界面...")
import gradio as gr
def simple_evaluate(json_data):
try:
print(f"备用界面收到请求: {json_data[:100] if isinstance(json_data, str) else 'Not a string'}")
data = json.loads(json_data) if isinstance(json_data, str) else json_data
result = submit_task(data)
return json.dumps(result, ensure_ascii=False)
except Exception as e:
return {"error": str(e)}
backup_demo = gr.Interface(
fn=simple_evaluate,
inputs=gr.Textbox(label="JSON输入"),
outputs=gr.Textbox(label="结果"),
title="代码评估服务 (备用界面)",
description="原界面启动失败,这是简化版本。提交格式: [{\"language\": \"python\", \"prompt\": \"def add(a, b):\\n\", \"processed_completions\": [\" return a + b\"], \"tests\": \"assert add(1, 2) == 3\"}]"
)
backup_demo.launch(server_name="0.0.0.0", server_port=int(os.environ.get("PORT", 7860)))
except Exception as e3:
print(f"备用界面也启动失败: {e3}")
traceback.print_exc()