feat: 完善 MVP-2 演示和进度体验

app/services/task_service.py

@@ -8,7 +8,7 @@ from sqlalchemy.orm import Session, sessionmaker
 
 from app.config import get_settings
 from app.db import SessionLocal
-from app.models import Comment, ContentItem, Hotspot, Task
+from app.models import Comment, ContentItem, Hotspot, Report, Task, utc_now
 from app.platforms.base import PlatformAPIError, TikHubClient
 from app.platforms.douyin import DouyinPlatform
 from app.platforms.xiaohongshu import XiaohongshuPlatform
@@ -22,6 +22,18 @@ RESTART_ERROR_MESSAGE = "系统重启，任务被中断"
 task_executor = ThreadPoolExecutor(max_workers=1)
 
 
+STAGE_LABELS = {
+    "queued": "等待启动",
+    "crawl_hotspots": "获取热点中",
+    "search_items": "搜索内容中",
+    "crawl_comments": "抓取评论中",
+    "ai_analysis": "AI 分析中",
+    "generate_reports": "生成报告中",
+    "success": "已完成",
+    "failed": "失败",
+}
+
+
 def has_running_task(session: Session) -> bool:
     return session.scalar(select(Task.id).where(Task.status == "running").limit(1)) is not None
 
@@ -110,6 +122,8 @@ def create_task(session: Session, request: CreateTaskRequest, *, submit_backgrou
         failed_items_count=0,
         analysis_success_rate=0.0,
         analysis_status="normal",
+        current_stage="queued",
+        last_progress_at=utc_now(),
     )
     session.add(task)
     session.commit()
@@ -120,11 +134,38 @@ def create_task(session: Session, request: CreateTaskRequest, *, submit_backgrou
 
 
 def list_tasks(session: Session) -> list[Task]:
-    return list(session.scalars(select(Task).order_by(Task.created_at.desc())))
+    tasks = list(session.scalars(select(Task).order_by(Task.created_at.desc())))
+    for task in tasks:
+        hydrate_task_progress(session, task)
+    return tasks
 
 
 def get_task(session: Session, task_id: str) -> Task | None:
-    return session.get(Task, task_id)
+    task = session.get(Task, task_id)
+    if task is not None:
+        hydrate_task_progress(session, task)
+    return task
+
+
+def hydrate_task_progress(session: Session, task: Task) -> Task:
+    task.comments_count = session.scalar(select(func.count(Comment.id)).where(Comment.task_id == task.id)) or 0
+    task.reports_count = session.scalar(select(func.count(Report.id)).where(Report.task_id == task.id)) or 0
+    task.is_demo = task.id.startswith("demo-")
+    task.last_progress_at = task.last_progress_at or task.created_at
+    stage_code = task.current_stage
+    if not stage_code and task.status == "success":
+        stage_code = "success"
+    elif not stage_code and task.status == "failed":
+        stage_code = "failed"
+    elif not stage_code and task.status == "running":
+        stage_code = "queued"
+    task.current_stage_label = STAGE_LABELS.get(stage_code or "", stage_code or "等待启动")
+    return task
+
+
+def update_task_stage(task: Task, stage: str) -> None:
+    task.current_stage = stage
+    task.last_progress_at = utc_now()
 
 
 def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionmaker = SessionLocal) -> None:
@@ -139,6 +180,8 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
             ai_requester, report_summary_provider = build_ai_dependencies()
 
             try:
+                update_task_stage(task, "crawl_hotspots")
+                session.commit()
                 hotspots = platform.fetch_hotspots(limit=task.hotspot_limit)
             except PlatformAPIError as exc:
                 _mark_task_failed(task, "crawl_hotspots", exc.error_type, str(exc))
@@ -163,6 +206,8 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
                 session.flush()
 
                 try:
+                    update_task_stage(task, "search_items")
+                    session.commit()
                     items = platform.search_items_by_hotspot(hotspot.title, limit=task.item_limit_per_hotspot)
                 except Exception as exc:
                     task.failed_items_count += task.item_limit_per_hotspot
@@ -193,6 +238,8 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
                     session.add(item)
                     session.flush()
                     try:
+                        update_task_stage(task, "crawl_comments")
+                        session.commit()
                         comments = platform.fetch_comments(item.source_item_id, limit=task.comment_limit_per_item)
                         for comment_data in comments:
                             session.add(
@@ -211,6 +258,8 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
                             )
                         session.flush()
                         item_comments = list(session.scalars(select(Comment).where(Comment.content_item_id == item.id)))
+                        update_task_stage(task, "ai_analysis")
+                        session.commit()
                         ai_results = analyze_comments_with_retry(
                             [{"comment_id": comment.id, "content": comment.content} for comment in item_comments],
                             requester=ai_requester,
@@ -231,6 +280,7 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
                             comment.ai_analysis_status = result.ai_analysis_status
                         item.status = "success"
                         task.successful_items_count += 1
+                        update_task_stage(task, "generate_reports")
                         generate_item_report(session, item.id, summary_provider=report_summary_provider)
                     except Exception as exc:
                         item.status = "failed"
@@ -246,6 +296,7 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
                         session.commit()
 
             if task.successful_items_count > 0:
+                update_task_stage(task, "generate_reports")
                 total_comments = session.scalar(select(func.count(Comment.id)).where(Comment.task_id == task.id)) or 0
                 success_comments = sum(1 for comment in task.comments if comment.ai_analysis_status == "success")
                 task.analysis_success_rate, task.analysis_status = calculate_analysis_status(
@@ -255,6 +306,7 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
                 for hotspot in task.hotspots:
                     generate_hotspot_report(session, hotspot.id, summary_provider=report_summary_provider)
                 task.status = "success"
+                update_task_stage(task, "success")
             else:
                 _mark_task_failed(task, task.error_stage or "crawl_items", task.error_type or "no_successful_items", task.error_message or "没有任何内容条目成功")
             session.commit()
@@ -268,6 +320,7 @@ def run_task(task_id: str, *, session_factory: Callable[[], Session] | sessionma
 
 def _mark_task_failed(task: Task, stage: str, error_type: str, message: str) -> None:
     task.status = "failed"
+    update_task_stage(task, "failed")
     task.error_stage = stage
     task.error_type = error_type
     task.error_message = message