diff --git a/backend/app/api/v1/__init__.py b/backend/app/api/v1/__init__.py index f37adeb..c94d15e 100644 --- a/backend/app/api/v1/__init__.py +++ b/backend/app/api/v1/__init__.py @@ -1,7 +1,7 @@ """Versioned API router.""" from fastapi import APIRouter from .arxiv import router as arxiv_router -from app.api.v1.endpoints import health, papers, users, auth, academic, conversations +from app.api.v1.endpoints import health, papers, users, auth, academic, conversations, library api_router = APIRouter() api_router.include_router(health.router, prefix="/health", tags=["health"]) @@ -11,3 +11,4 @@ api_router.include_router(academic.router, prefix="/academic", tags=["academic"]) api_router.include_router(arxiv_router,prefix="/arxiv",tags=["arxiv"]) api_router.include_router(conversations.router, prefix="/conversations", tags=["conversations"]) +api_router.include_router(library.router, prefix="/library", tags=["library"]) diff --git a/backend/app/api/v1/api.py b/backend/app/api/v1/api.py index 4e4cb5e..1bd2f41 100644 --- a/backend/app/api/v1/api.py +++ b/backend/app/api/v1/api.py @@ -1,7 +1,7 @@ """API v1 router aggregation.""" from fastapi import APIRouter -from app.api.v1.endpoints import academic, auth, conversations, papers, users +from app.api.v1.endpoints import academic, auth, conversations, library, papers, users # 创建主路由器 api_router = APIRouter() @@ -25,6 +25,12 @@ tags=["论文检索"] ) +api_router.include_router( + library.router, + prefix="/library", + tags=["我的文库"] +) + api_router.include_router( academic.router, prefix="/academic", @@ -35,4 +41,4 @@ conversations.router, prefix="/conversations", tags=["对话历史"] -) \ No newline at end of file +) diff --git a/backend/app/api/v1/endpoints/auth.py b/backend/app/api/v1/endpoints/auth.py index 029b84b..1e0b886 100644 --- a/backend/app/api/v1/endpoints/auth.py +++ b/backend/app/api/v1/endpoints/auth.py @@ -1,16 +1,33 @@ """认证授权相关API接口""" -from fastapi import APIRouter, Depends, HTTPException, status +from __future__ import annotations + +import json +import secrets +from typing import cast +from urllib.parse import urlencode + +import httpx +from fastapi import APIRouter, Depends, HTTPException, Request, status +from fastapi.responses import RedirectResponse from fastapi.security import OAuth2PasswordRequestForm from sqlalchemy.ext.asyncio import AsyncSession -from typing import cast from app.core.auth import create_access_token -from app.core.security import verify_password +from app.core.config import get_settings +from app.core.security import hash_password, verify_password from app.db import UserRepository from app.db.session import get_db from app.schemas.auth import Token # type: ignore[import-not-found] router = APIRouter() +settings = get_settings() + +GITHUB_AUTHORIZE_URL = "https://github.com/login/oauth/authorize" +GITHUB_TOKEN_URL = "https://github.com/login/oauth/access_token" +GITHUB_USER_API = "https://api.github.com/user" +GITHUB_EMAILS_API = "https://api.github.com/user/emails" +GITHUB_STATE_COOKIE = "github_oauth_state" +GITHUB_STATE_TTL = 600 @router.post( @@ -99,4 +116,189 @@ async def login_for_access_token( ) token = create_access_token(subject=cast(str, getattr(user, "email", ""))) - return Token(access_token=token) \ No newline at end of file + return Token(access_token=token) + + +@router.get("/github/login") +async def github_login(request: Request, next: str | None = None): + client_id = settings.github_client_id + client_secret = settings.github_client_secret + if not client_id or not client_secret: + raise HTTPException(status_code=status.HTTP_503_SERVICE_UNAVAILABLE, detail="GitHub OAuth 未配置") + + nonce = secrets.token_urlsafe(32) + cookie_payload = json.dumps( + { + "nonce": nonce, + "next": _sanitize_next_path(next), + } + ) + + callback_url = str(request.url_for("github_callback")) + params = { + "client_id": client_id, + "redirect_uri": callback_url, + "scope": "read:user user:email", + "state": nonce, + "allow_signup": "true", + } + authorize_url = f"{GITHUB_AUTHORIZE_URL}?{urlencode(params)}" + response = RedirectResponse(authorize_url, status_code=status.HTTP_302_FOUND) + response.set_cookie( + GITHUB_STATE_COOKIE, + cookie_payload, + max_age=GITHUB_STATE_TTL, + httponly=True, + secure=_is_cookie_secure(), + samesite="lax", + ) + return response + + +@router.get("/github/callback", name="github_callback") +async def github_callback( + request: Request, + code: str | None = None, + state: str | None = None, + db: AsyncSession = Depends(get_db), +): + cookie_payload = request.cookies.get(GITHUB_STATE_COOKIE) + if not cookie_payload: + return _oauth_error_redirect("Missing OAuth state cookie.") + + try: + payload = json.loads(cookie_payload) + except json.JSONDecodeError: + return _oauth_error_redirect("Invalid OAuth state payload.") + + expected_state = payload.get("nonce") + next_path = _sanitize_next_path(payload.get("next")) + if not code or not state or expected_state != state: + return _oauth_error_redirect("OAuth state mismatch.", next_path=next_path) + + try: + callback_url = str(request.url_for("github_callback")) + token_data = await _exchange_github_code_for_token(code, callback_url) + access_token = token_data.get("access_token") + if not access_token: + raise RuntimeError("GitHub did not return an access token.") + + github_user, primary_email = await _fetch_github_profile(access_token) + if not primary_email: + raise RuntimeError("未能获取 GitHub 邮箱，请在 GitHub 账户开启公开邮箱或授权 email scope。") + + github_user_id = str(github_user.get("id")) + repo = UserRepository(db) + user = await repo.get_by_oauth_account("github", github_user_id) + + if user is None: + user = await repo.get_by_email(primary_email) + if user: + user = await repo.update( + user, + { + "oauth_provider": "github", + "oauth_account_id": github_user_id, + "avatar_url": user.avatar_url or github_user.get("avatar_url"), + }, + ) + else: + hashed_password = hash_password(secrets.token_urlsafe(32)) + user = await repo.create( + email=primary_email, + hashed_password=hashed_password, + full_name=github_user.get("name") or github_user.get("login"), + avatar_url=github_user.get("avatar_url"), + oauth_provider="github", + oauth_account_id=github_user_id, + ) + else: + updates: dict[str, str | None] = {} + if not user.avatar_url and github_user.get("avatar_url"): + updates["avatar_url"] = github_user.get("avatar_url") + updates["oauth_provider"] = "github" + updates["oauth_account_id"] = github_user_id + user = await repo.update(user, updates) + + await db.commit() + await db.refresh(user) + + token = create_access_token(subject=cast(str, getattr(user, "email", ""))) + redirect_target = _build_frontend_redirect(token=token, next_path=next_path) + response = RedirectResponse(redirect_target, status_code=status.HTTP_302_FOUND) + response.delete_cookie(GITHUB_STATE_COOKIE) + return response + except Exception as exc: # pragma: no cover - defensive + return _oauth_error_redirect(str(exc), next_path=next_path) + + +def _sanitize_next_path(value: str | None) -> str: + if not value or not value.startswith("/"): + return "/" + return value + + +def _is_cookie_secure() -> bool: + return settings.environment not in {"local", "development"} + + +async def _exchange_github_code_for_token(code: str, redirect_uri: str) -> dict[str, str]: + client_id = settings.github_client_id + client_secret = settings.github_client_secret + if not client_id or not client_secret: + raise RuntimeError("GitHub OAuth 未配置。") + + data = { + "client_id": client_id, + "client_secret": client_secret, + "code": code, + "redirect_uri": redirect_uri, + } + + async with httpx.AsyncClient(timeout=15.0, headers={"Accept": "application/json"}) as client: + response = await client.post(GITHUB_TOKEN_URL, data=data) + response.raise_for_status() + return response.json() + + +async def _fetch_github_profile(access_token: str) -> tuple[dict[str, str], str | None]: + headers = { + "Accept": "application/json", + "Authorization": f"Bearer {access_token}", + } + async with httpx.AsyncClient(timeout=15.0) as client: + user_resp = await client.get(GITHUB_USER_API, headers=headers) + user_resp.raise_for_status() + user_data = user_resp.json() + email = user_data.get("email") + if not email: + emails_resp = await client.get(GITHUB_EMAILS_API, headers=headers) + emails_resp.raise_for_status() + emails = emails_resp.json() + email = next( + ( + item.get("email") + for item in emails + if item.get("primary") and item.get("verified") + ), + None, + ) or next((item.get("email") for item in emails if item.get("verified")), None) + return user_data, email + + +def _build_frontend_redirect(*, token: str | None, next_path: str) -> str: + params = {"next": _sanitize_next_path(next_path)} + if token: + params["token"] = token + params["provider"] = "github" + return f"{settings.frontend_oauth_redirect_url}?{urlencode(params)}" + + +def _oauth_error_redirect(message: str, *, next_path: str = "/") -> RedirectResponse: + params = { + "error": message, + "next": _sanitize_next_path(next_path), + } + response = RedirectResponse(f"{settings.frontend_oauth_redirect_url}?{urlencode(params)}", status_code=status.HTTP_302_FOUND) + response.delete_cookie(GITHUB_STATE_COOKIE) + return response diff --git a/backend/app/api/v1/endpoints/library.py b/backend/app/api/v1/endpoints/library.py new file mode 100644 index 0000000..abcdf27 --- /dev/null +++ b/backend/app/api/v1/endpoints/library.py @@ -0,0 +1,369 @@ +"""Endpoints for managing personal library folders.""" +from __future__ import annotations + +import asyncio +import hashlib +from pathlib import Path +from typing import Any +from uuid import uuid4 + +import httpx +from fastapi import APIRouter, Depends, HTTPException, Response, status +from fastapi.responses import FileResponse, StreamingResponse +from sqlalchemy.exc import IntegrityError +from sqlalchemy.ext.asyncio import AsyncSession + +from app.core.config import get_settings +from app.core.auth import decode_access_token +from app.db.repository import LibraryFolderRepository, UploadedPaperRepository, UserRepository +from app.db.session import get_db +from app.dependencies.auth import get_current_user, get_current_user_optional +from app.schemas import library as library_schema +from app.schemas import upload as upload_schema +from app.schemas.auth import TokenPayload + +settings = get_settings() +UPLOAD_DIR = settings.media_path / "uploads" +UPLOAD_DIR.mkdir(parents=True, exist_ok=True) + +router = APIRouter() + + +@router.get( + "/folders", + response_model=library_schema.LibraryFolderListResponse, + summary="列出我的文库文件夹", +) +async def list_library_folders( + current_user=Depends(get_current_user), + db: AsyncSession = Depends(get_db), +) -> library_schema.LibraryFolderListResponse: + folder_repo = LibraryFolderRepository(db) + uploaded_repo = UploadedPaperRepository(db) + folders = await folder_repo.list_for_user(current_user.id) + counts = await uploaded_repo.count_by_folder(current_user.id) + folder_items: list[library_schema.LibraryFolderRead] = [] + for folder in folders: + base = library_schema.LibraryFolderRead.model_validate(folder, from_attributes=True) + folder_items.append(base.model_copy(update={"paper_count": counts.get(folder.id, 0)})) + return library_schema.LibraryFolderListResponse( + folders=folder_items, + unfiled_count=counts.get(None, 0), + ) + + +@router.post( + "/folders", + response_model=library_schema.LibraryFolderRead, + status_code=status.HTTP_201_CREATED, + summary="新增文库文件夹", +) +async def create_library_folder( + payload: library_schema.LibraryFolderCreate, + current_user=Depends(get_current_user), + db: AsyncSession = Depends(get_db), +) -> library_schema.LibraryFolderRead: + repo = LibraryFolderRepository(db) + try: + folder = await repo.create( + user_id=current_user.id, + name=payload.name.strip(), + color=payload.color, + ) + await db.commit() + except IntegrityError: + await db.rollback() + raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="文件夹名称已存在") + + await db.refresh(folder) + return library_schema.LibraryFolderRead.model_validate(folder, from_attributes=True) + + +@router.patch( + "/folders/{folder_id}", + response_model=library_schema.LibraryFolderRead, + summary="更新文件夹信息", +) +async def rename_library_folder( + folder_id: int, + payload: library_schema.LibraryFolderUpdate, + current_user=Depends(get_current_user), + db: AsyncSession = Depends(get_db), +) -> library_schema.LibraryFolderRead: + repo = LibraryFolderRepository(db) + folder = await repo.get_for_user(folder_id, current_user.id) + if folder is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到该文件夹") + + try: + updated = await repo.rename(folder, name=payload.name.strip(), color=payload.color) + await db.commit() + except IntegrityError: + await db.rollback() + raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="文件夹名称已存在") + + await db.refresh(updated) + uploaded_repo = UploadedPaperRepository(db) + counts = await uploaded_repo.count_by_folder(current_user.id) + base = library_schema.LibraryFolderRead.model_validate(updated, from_attributes=True) + return base.model_copy(update={"paper_count": counts.get(folder_id, 0)}) + + +@router.delete( + "/folders/{folder_id}", + summary="删除文件夹", +) +async def delete_library_folder( + folder_id: int, + current_user=Depends(get_current_user), + db: AsyncSession = Depends(get_db), +) -> None: + repo = LibraryFolderRepository(db) + folder = await repo.get_for_user(folder_id, current_user.id) + if folder is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到该文件夹") + + uploaded_repo = UploadedPaperRepository(db) + total = await uploaded_repo.count_unique_for_user(current_user.id, folder_id=folder_id) + if total: + raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="请先移动或删除文件夹中的文献") + + await repo.delete(folder) + await db.commit() + return Response(status_code=status.HTTP_204_NO_CONTENT) + + +@router.patch( + "/uploads/{paper_id}/folder", + response_model=upload_schema.UploadedPaperRead, + summary="更新文献所在文件夹", +) +async def assign_uploaded_paper_folder( + paper_id: int, + payload: library_schema.LibraryFolderAssignment, + current_user=Depends(get_current_user), + db: AsyncSession = Depends(get_db), +) -> upload_schema.UploadedPaperRead: + folder_repo = LibraryFolderRepository(db) + uploaded_repo = UploadedPaperRepository(db) + record = await uploaded_repo.get_for_user(paper_id, current_user.id) + if record is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到对应的上传文件") + + target_folder_id: int | None = None + if payload.folder_id is not None: + folder = await folder_repo.get_for_user(payload.folder_id, current_user.id) + if folder is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="目标文件夹不存在") + target_folder_id = folder.id + + await uploaded_repo.update_folder(record, target_folder_id) + await db.commit() + await db.refresh(record) + return upload_schema.UploadedPaperRead.model_validate(record, from_attributes=True) + + +@router.get( + "/uploads/{paper_id}/download", + summary="下载上传的 PDF 原文件", +) +async def download_uploaded_paper( + paper_id: int, + access_token: str | None = None, + current_user=Depends(get_current_user_optional), + db: AsyncSession = Depends(get_db), +) -> Any: + user = current_user + if user is None: + if not access_token: + raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="请先登录") + user = await _get_user_from_token(access_token, db) + + uploaded_repo = UploadedPaperRepository(db) + record = await uploaded_repo.get_for_user(paper_id, user.id) + if record is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到对应的上传文件") + + file_url = record.file_url or "" + filename = record.original_filename or f"paper_{paper_id}.pdf" + + if file_url.startswith(("http://", "https://")): + try: + async with httpx.AsyncClient(timeout=30.0, follow_redirects=True) as client: + response = await client.get(file_url) + response.raise_for_status() + content = response.content + except httpx.HTTPError as exc: + raise HTTPException(status_code=status.HTTP_502_BAD_GATEWAY, detail="无法读取远程 PDF") from exc + + headers = {"Content-Disposition": f'inline; filename="{filename}"'} + return StreamingResponse( + iter([content]), + media_type=record.content_type or "application/pdf", + headers=headers, + ) + + local_path = _resolve_local_media_path(file_url) + if local_path is None or not local_path.exists(): + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="PDF 文件已丢失，请重新上传") + + headers = {"Content-Disposition": f'inline; filename="{filename}"'} + return FileResponse( + local_path, + media_type=record.content_type or "application/pdf", + filename=filename, + headers=headers, + ) + + +@router.head( + "/uploads/{paper_id}/download", + summary="HEAD 检查上传 PDF", +) +async def head_uploaded_paper( + paper_id: int, + access_token: str | None = None, + current_user=Depends(get_current_user_optional), + db: AsyncSession = Depends(get_db), +) -> Response: + user = current_user + if user is None: + if not access_token: + raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="请先登录") + user = await _get_user_from_token(access_token, db) + + uploaded_repo = UploadedPaperRepository(db) + record = await uploaded_repo.get_for_user(paper_id, user.id) + if record is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到对应的上传文件") + + file_url = record.file_url or "" + if file_url.startswith(("http://", "https://")): + return Response(status_code=status.HTTP_200_OK) + + local_path = _resolve_local_media_path(file_url) + if local_path is None or not local_path.exists(): + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="PDF 文件已丢失，请重新上传") + + return Response(status_code=status.HTTP_200_OK) + + +@router.post( + "/uploads/{paper_id}/ensure-local", + response_model=upload_schema.UploadedPaperRead, + summary="下载远程 PDF 到本地", +) +async def ensure_uploaded_paper_local( + paper_id: int, + payload: library_schema.LibraryDownloadRequest, + current_user=Depends(get_current_user), + db: AsyncSession = Depends(get_db), +) -> upload_schema.UploadedPaperRead: + uploaded_repo = UploadedPaperRepository(db) + record = await uploaded_repo.get_for_user(paper_id, current_user.id) + if record is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到对应的上传文件") + + candidate_url = payload.pdf_url or record.file_url + + local_path = _resolve_local_media_path(record.file_url or "") + if candidate_url is None or candidate_url.startswith("#"): + if local_path and local_path.exists(): + return upload_schema.UploadedPaperRead.model_validate(record, from_attributes=True) + raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="缺少可下载的 PDF 链接") + + if not candidate_url.startswith(("http://", "https://")): + if local_path and local_path.exists(): + return upload_schema.UploadedPaperRead.model_validate(record, from_attributes=True) + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="本地 PDF 文件已丢失，请提供远程链接") + + content = await _download_pdf_from_url(candidate_url) + stored_filename, file_url, file_size, file_hash = await _save_pdf_bytes(current_user.id, content) + + await uploaded_repo.update_file_fields( + record, + stored_filename=stored_filename, + file_url=file_url, + file_size=file_size, + file_hash=file_hash, + content_type="application/pdf", + ) + await db.commit() + await db.refresh(record) + + return upload_schema.UploadedPaperRead.model_validate(record, from_attributes=True) + + +@router.delete( + "/uploads/{paper_id}", + summary="删除上传的文档", +) +async def delete_uploaded_paper( + paper_id: int, + current_user=Depends(get_current_user), + db: AsyncSession = Depends(get_db), +) -> Response: + uploaded_repo = UploadedPaperRepository(db) + record = await uploaded_repo.get_for_user(paper_id, current_user.id) + if record is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到对应的上传文件") + + file_url = record.file_url or "" + local_path = _resolve_local_media_path(file_url) + if local_path is not None and local_path.exists(): + try: + local_path.unlink() + except OSError: + pass + + await uploaded_repo.delete(record) + await db.commit() + return Response(status_code=status.HTTP_204_NO_CONTENT) + + +def _resolve_local_media_path(file_url: str) -> Path | None: + if not file_url or file_url.startswith(("http://", "https://")): + return None + + relative_path = file_url.lstrip("/") + if relative_path.startswith("media/"): + relative_path = relative_path[len("media/") :] + candidate = (settings.media_path / relative_path).resolve() + media_root = settings.media_path.resolve() + if not str(candidate).startswith(str(media_root)): + return None + return candidate + + +async def _get_user_from_token(token: str, db: AsyncSession): + try: + payload_dict = decode_access_token(token) + payload = TokenPayload.model_validate(payload_dict) + except Exception as exc: # pragma: no cover - defensive + raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="无效的访问令牌") from exc + + if not payload.sub: + raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="无效的访问令牌") + + repo = UserRepository(db) + user = await repo.get_by_email(payload.sub) + if user is None: + raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="用户不存在") + return user + + +async def _download_pdf_from_url(url: str) -> bytes: + async with httpx.AsyncClient(timeout=30.0, follow_redirects=True) as client: + response = await client.get(url) + response.raise_for_status() + return response.content + + +async def _save_pdf_bytes(user_id: int, content: bytes) -> tuple[str, str, int, str]: + stored_filename = f"paper_{user_id}_{uuid4().hex}.pdf" + destination = UPLOAD_DIR / stored_filename + await asyncio.to_thread(destination.write_bytes, content) + file_url = f"/media/uploads/{stored_filename}" + file_size = len(content) + file_hash = hashlib.sha256(content).hexdigest() + return stored_filename, file_url, file_size, file_hash diff --git a/backend/app/api/v1/endpoints/papers.py b/backend/app/api/v1/endpoints/papers.py index e092b60..d32c409 100644 --- a/backend/app/api/v1/endpoints/papers.py +++ b/backend/app/api/v1/endpoints/papers.py @@ -18,18 +18,20 @@ import logging import math from contextlib import suppress +from datetime import datetime from pathlib import Path from typing import TYPE_CHECKING, Any, Final, cast from uuid import uuid4 import httpx -from fastapi import APIRouter, Depends, File, HTTPException, Query, UploadFile, status +from fastapi import APIRouter, Depends, File, Form, HTTPException, Query, UploadFile, status from sqlalchemy.ext.asyncio import AsyncSession from pydantic import BaseModel, Field from app.core.config import get_settings from app.db.paper_repository import PaperRepository from app.db.repository import ( + LibraryFolderRepository, MineruParseJobRepository, ParsedPaperCacheRepository, UploadedPaperRepository, @@ -44,6 +46,7 @@ from app.schemas.annotation import AnnotationApplyResponse, AnnotationCreate from app.schemas.conversation import ConversationCreate, ConversationMessageCreate from app.services.annotations import apply_annotation +from app.services.pdf_metadata import extract_pdf_metadata_async from app.services.mineru_cli import parse_pdf from app.services.ai.llm_client import DeepSeekClient from app.workers.tasks import run_mineru_parse @@ -73,6 +76,26 @@ async def search_papers( raise HTTPException(status_code=status.HTTP_501_NOT_IMPLEMENTED, detail="Search service not yet available") +def _resolve_folder_filter(folder_id: int | None) -> tuple[int | None, bool]: + if folder_id == 0: + return None, True + return folder_id, False + + +async def _ensure_folder_access( + folder_repo: LibraryFolderRepository, + *, + folder_id: int | None, + user_id: int, +) -> int | None: + if folder_id is None: + return None + folder = await folder_repo.get_for_user(folder_id, user_id) + if folder is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到指定的文库文件夹") + return folder.id + + @router.get( "/uploads", response_model=upload_schema.UploadedPaperListResponse, @@ -84,9 +107,15 @@ async def list_uploaded_papers( db: AsyncSession = Depends(get_db), page: int = Query(1, ge=1, description="页码，从 1 开始"), page_size: int = Query(12, ge=1, le=60, description="每页数量，最多 60 条"), + folder_id: int | None = Query(None, ge=0, description="文件夹 ID（0 表示未分类，仅显示未分配文件夹的文献）"), ) -> upload_schema.UploadedPaperListResponse: repo = UploadedPaperRepository(db) - total = await repo.count_for_user(current_user.id) + normalized_folder_id, unfiled_only = _resolve_folder_filter(folder_id) + total = await repo.count_unique_for_user( + current_user.id, + folder_id=normalized_folder_id, + unfiled_only=unfiled_only, + ) total_pages = math.ceil(total / page_size) if total else 0 if total_pages and page > total_pages: @@ -95,10 +124,12 @@ async def list_uploaded_papers( page = 1 offset = max(0, (page - 1) * page_size) - records = await repo.list_for_user_paginated( + records = await repo.list_unique_for_user_paginated( current_user.id, offset=offset, limit=page_size, + folder_id=normalized_folder_id, + unfiled_only=unfiled_only, ) items = [ upload_schema.UploadedPaperRead.model_validate(record, from_attributes=True) @@ -122,6 +153,7 @@ async def list_uploaded_papers( ) async def upload_paper( file: UploadFile = File(..., description="需要上传的 PDF 文件"), + folder_id: int | None = Form(None, description="文件夹 ID，不填则保存在未分类"), current_user=Depends(get_current_user), db: AsyncSession = Depends(get_db), ) -> upload_schema.UploadedPaperRead: @@ -151,7 +183,19 @@ async def upload_paper( relative_url = f"/media/uploads/{stored_filename}" + folder_repo = LibraryFolderRepository(db) repo = UploadedPaperRepository(db) + resolved_folder_id = await _ensure_folder_access( + folder_repo, + folder_id=folder_id if folder_id and folder_id > 0 else None, + user_id=current_user.id, + ) + metadata_json: dict | None = None + try: + metadata_json = await extract_pdf_metadata_async(destination, sanitized_name) + except Exception as exc: # pragma: no cover - best effort only + logger.warning("Failed to extract metadata for uploaded PDF: %s", exc) + metadata_json = None try: record = await repo.create( @@ -162,6 +206,8 @@ async def upload_paper( file_size=len(cleaned_bytes), file_url=relative_url, file_hash=file_hash, + folder_id=resolved_folder_id, + metadata_json=metadata_json, ) await db.commit() except Exception: @@ -175,6 +221,23 @@ async def upload_paper( return upload_schema.UploadedPaperRead.model_validate(record, from_attributes=True) +@router.get( + "/uploads/{paper_id}", + response_model=upload_schema.UploadedPaperRead, + summary="获取单篇上传文档详情", +) +async def get_uploaded_paper_detail( + paper_id: int, + current_user=Depends(get_current_user), + db: AsyncSession = Depends(get_db), +) -> upload_schema.UploadedPaperRead: + repo = UploadedPaperRepository(db) + record = await repo.get_for_user(paper_id, current_user.id) + if record is None: + raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="未找到对应的上传文件") + return upload_schema.UploadedPaperRead.model_validate(record, from_attributes=True) + + @router.get("/{paper_id}", response_model=paper_schema.PaperRead) async def get_paper(paper_id: int, db: AsyncSession = Depends(get_db)) -> paper_schema.PaperRead: """Fetch a single paper by identifier.""" @@ -182,12 +245,12 @@ async def get_paper(paper_id: int, db: AsyncSession = Depends(get_db)) -> paper_ raise HTTPException(status_code=status.HTTP_501_NOT_IMPLEMENTED, detail="Paper retrieval not yet available") -@router.post("/import", response_model=paper_schema.PaperRead, status_code=status.HTTP_201_CREATED) +@router.post("/import", response_model=paper_schema.PaperImportResponse, status_code=status.HTTP_201_CREATED) async def import_paper( request: paper_schema.PaperImportRequest, current_user=Depends(get_current_user), db: AsyncSession = Depends(get_db) -) -> paper_schema.PaperRead: +) -> paper_schema.PaperImportResponse: """Import a recommended paper into the local library and optionally download its PDF. Behavior: @@ -198,7 +261,14 @@ async def import_paper( logger.info(f"Importing paper for user {current_user.id}: {request.title}") + folder_repo = LibraryFolderRepository(db) + try: + resolved_folder_id = await _ensure_folder_access( + folder_repo, + folder_id=request.folder_id if request.folder_id and request.folder_id > 0 else None, + user_id=current_user.id, + ) repo = PaperRepository(db) paper = await repo.create( title=request.title, @@ -227,6 +297,10 @@ async def import_paper( downloaded_url = None download_error = None + metadata_json: dict | None = None + + safe_title = "".join(c for c in (request.title or "paper")[:50] if c.isalnum() or c in (' ', '-', '_')).strip() + if pdf_url: # Normalize common arXiv abstract links to direct PDF links if "arxiv.org/abs/" in pdf_url and "arxiv.org/pdf/" not in pdf_url: @@ -262,6 +336,11 @@ async def import_paper( downloaded_filename = filename downloaded_content = content downloaded_url = f"/media/papers/{filename}" + try: + metadata_json = await extract_pdf_metadata_async(dest, safe_title) + except Exception as exc: # pragma: no cover + logger.warning("Failed to parse metadata for imported paper %s: %s", paper.id, exc) + metadata_json = None # attach file path to metadata paper.metadata_json = (paper.metadata_json or {}) @@ -279,14 +358,23 @@ async def import_paper( paper.metadata_json = (paper.metadata_json or {}) paper.metadata_json["download_error"] = download_error + if metadata_json is None: + metadata_json = { + "title": request.title, + "summary": request.abstract, + "topics": [], + "extracted_at": datetime.utcnow().isoformat(), + } + # Always save to UploadedPaper table, regardless of download success uploaded_repo = UploadedPaperRepository(db) - safe_title = "".join(c for c in (request.title or "paper")[:50] if c.isalnum() or c in (' ', '-', '_')).strip() + + uploaded_record: UploadedPaper | None = None if downloaded_filename and downloaded_content: # Downloaded successfully - save with local file reference logger.info(f"Saving to UploadedPaper with downloaded file: {downloaded_filename}") - await uploaded_repo.create( + uploaded_record = await uploaded_repo.create( user_id=current_user.id, stored_filename=downloaded_filename, original_filename=f"{safe_title}.pdf", @@ -294,13 +382,15 @@ async def import_paper( file_size=len(downloaded_content), file_url=downloaded_url, file_hash=_calculate_file_hash(downloaded_content), + folder_id=resolved_folder_id, + metadata_json=metadata_json, ) else: # No download or download failed - save metadata only with external URL logger.info(f"Saving to UploadedPaper without local file (pdf_url={pdf_url})") # Create a placeholder entry - use original pdf_url as file_url if available fallback_url = pdf_url if pdf_url else f"#{paper.id}" - await uploaded_repo.create( + uploaded_record = await uploaded_repo.create( user_id=current_user.id, stored_filename=f"pending_{paper.id}.pdf", original_filename=f"{safe_title}.pdf", @@ -308,15 +398,24 @@ async def import_paper( file_size=0, # 0 indicates no local file file_url=fallback_url, file_hash=None, + folder_id=resolved_folder_id, + metadata_json=metadata_json, ) # Commit all changes await db.commit() await db.refresh(paper) + if uploaded_record is not None: + await db.refresh(uploaded_record) logger.info(f"Successfully imported paper id={paper.id} for user {current_user.id}") - return paper + return paper_schema.PaperImportResponse( + paper=paper_schema.PaperRead.model_validate(paper, from_attributes=True), + uploaded=upload_schema.UploadedPaperRead.model_validate(uploaded_record, from_attributes=True) + if uploaded_record + else None, + ) except Exception as exc: await db.rollback() diff --git a/backend/app/core/config.py b/backend/app/core/config.py index 5290773..223e8ff 100644 --- a/backend/app/core/config.py +++ b/backend/app/core/config.py @@ -51,7 +51,11 @@ class Settings(BaseSettings): OPENALEX_TRUST_ENV: bool = True OPENALEX_MAILTO: str = "insightreading@example.com" - + github_client_id: str | None = None + github_client_secret: str | None = None + frontend_app_url: str = "http://localhost:3000" + frontend_oauth_redirect_path: str = "/auth/sso" + DEEPSEEK_API_KEY: str = "sk-7b3e938c92fb4a5ca601a35becdee4e2" DEEPSEEK_MODEL: str = "deepseek-chat" DEEPSEEK_BASE_URL: str = "https://api.deepseek.com/v1" @@ -111,6 +115,14 @@ def media_path(self) -> Path: return Path(override) return Path(self.media_root) + @property + def frontend_oauth_redirect_url(self) -> str: + base = (self.frontend_app_url or "http://localhost:3000").rstrip("/") + path = self.frontend_oauth_redirect_path or "/auth/sso" + if not path.startswith("/"): + path = f"/{path}" + return f"{base}{path}" + @lru_cache def get_settings() -> Settings: diff --git a/backend/app/db/repository.py b/backend/app/db/repository.py index 7cdcf24..7d6ed41 100644 --- a/backend/app/db/repository.py +++ b/backend/app/db/repository.py @@ -5,8 +5,10 @@ from sqlalchemy import func, select from sqlalchemy.ext.asyncio import AsyncSession +from sqlalchemy.orm import selectinload from app.models.mineru_parse_job import MineruParseJob +from app.models.library_folder import LibraryFolder from app.models.parsed_paper_cache import ParsedPaperCache from app.models.uploaded_paper import UploadedPaper from app.models.user import User @@ -23,8 +25,24 @@ async def get_by_email(self, email: str) -> User | None: result = await self._session.execute(stmt) return result.scalar_one_or_none() - async def create(self, *, email: str, hashed_password: str, full_name: str | None) -> User: - user = User(email=email, hashed_password=hashed_password, full_name=full_name) + async def create( + self, + *, + email: str, + hashed_password: str, + full_name: str | None, + avatar_url: str | None = None, + oauth_provider: str | None = None, + oauth_account_id: str | None = None, + ) -> User: + user = User( + email=email, + hashed_password=hashed_password, + full_name=full_name, + avatar_url=avatar_url, + oauth_provider=oauth_provider, + oauth_account_id=oauth_account_id, + ) self._session.add(user) await self._session.flush() return user @@ -36,6 +54,55 @@ async def update(self, user: User, updates: Mapping[str, Any]) -> User: await self._session.flush() return user + async def get_by_oauth_account(self, provider: str, account_id: str) -> User | None: + stmt = select(User).where( + User.oauth_provider == provider, + User.oauth_account_id == account_id, + ) + result = await self._session.execute(stmt) + return result.scalar_one_or_none() + + +class LibraryFolderRepository: + """Persistence helpers for organizing library folders.""" + + def __init__(self, session: AsyncSession) -> None: + self._session = session + + async def list_for_user(self, user_id: int) -> list[LibraryFolder]: + stmt = ( + select(LibraryFolder) + .where(LibraryFolder.user_id == user_id) + .order_by(LibraryFolder.created_at.asc(), LibraryFolder.id.asc()) + ) + result = await self._session.execute(stmt) + return list(result.scalars().all()) + + async def get_for_user(self, folder_id: int, user_id: int) -> LibraryFolder | None: + stmt = ( + select(LibraryFolder) + .where(LibraryFolder.id == folder_id, LibraryFolder.user_id == user_id) + ) + result = await self._session.execute(stmt) + return result.scalar_one_or_none() + + async def create(self, *, user_id: int, name: str, color: str | None = None) -> LibraryFolder: + folder = LibraryFolder(user_id=user_id, name=name, color=color) + self._session.add(folder) + await self._session.flush() + return folder + + async def delete(self, folder: LibraryFolder) -> None: + await self._session.delete(folder) + await self._session.flush() + + async def rename(self, folder: LibraryFolder, *, name: str, color: str | None = None) -> LibraryFolder: + folder.name = name + if color is not None: + folder.color = color + await self._session.flush() + return folder + class UploadedPaperRepository: """Database operations for UploadedPaper model.""" @@ -43,6 +110,36 @@ class UploadedPaperRepository: def __init__(self, session: AsyncSession) -> None: self._session = session + def _deduped_uploads_subquery( + self, + user_id: int, + *, + folder_id: int | None = None, + unfiled_only: bool = False, + per_folder_partition: bool = False, + ): + dedupe_key = func.coalesce(UploadedPaper.file_hash, UploadedPaper.original_filename, UploadedPaper.stored_filename) + partition_columns: list[Any] = [dedupe_key] + if per_folder_partition: + partition_columns.append(UploadedPaper.folder_id) + row_number = func.row_number().over( + partition_by=partition_columns, + order_by=(UploadedPaper.uploaded_at.desc(), UploadedPaper.id.desc()), + ) + stmt = select( + UploadedPaper.id.label("id"), + UploadedPaper.uploaded_at.label("uploaded_at"), + UploadedPaper.folder_id.label("folder_id"), + row_number.label("rn"), + ).where(UploadedPaper.user_id == user_id) + + if unfiled_only: + stmt = stmt.where(UploadedPaper.folder_id.is_(None)) + elif folder_id is not None: + stmt = stmt.where(UploadedPaper.folder_id == folder_id) + + return stmt.subquery() + async def count_for_user(self, user_id: int) -> int: stmt = select(func.count()).select_from(UploadedPaper).where(UploadedPaper.user_id == user_id) result = await self._session.execute(stmt) @@ -62,11 +159,59 @@ async def list_for_user_paginated(self, user_id: int, *, offset: int, limit: int async def get_for_user(self, paper_id: int, user_id: int) -> UploadedPaper | None: stmt = ( select(UploadedPaper) + .options(selectinload(UploadedPaper.folder)) .where(UploadedPaper.id == paper_id, UploadedPaper.user_id == user_id) ) result = await self._session.execute(stmt) return result.scalar_one_or_none() + async def count_unique_for_user( + self, + user_id: int, + *, + folder_id: int | None = None, + unfiled_only: bool = False, + ) -> int: + deduped = self._deduped_uploads_subquery(user_id, folder_id=folder_id, unfiled_only=unfiled_only) + stmt = select(func.count()).select_from(deduped).where(deduped.c.rn == 1) + result = await self._session.execute(stmt) + return int(result.scalar_one()) + + async def list_unique_for_user_paginated( + self, + user_id: int, + *, + offset: int, + limit: int, + folder_id: int | None = None, + unfiled_only: bool = False, + ) -> list[UploadedPaper]: + deduped = self._deduped_uploads_subquery(user_id, folder_id=folder_id, unfiled_only=unfiled_only) + stmt = ( + select(UploadedPaper) + .options(selectinload(UploadedPaper.folder)) + .join(deduped, UploadedPaper.id == deduped.c.id) + .where(deduped.c.rn == 1) + .order_by(UploadedPaper.uploaded_at.desc(), UploadedPaper.id.desc()) + .offset(offset) + .limit(limit) + ) + result = await self._session.execute(stmt) + return list(result.scalars().unique().all()) + + async def count_by_folder(self, user_id: int) -> dict[int | None, int]: + deduped = self._deduped_uploads_subquery(user_id, per_folder_partition=True) + stmt = ( + select(deduped.c.folder_id, func.count().label("total")) + .where(deduped.c.rn == 1) + .group_by(deduped.c.folder_id) + ) + result = await self._session.execute(stmt) + counts: dict[int | None, int] = {} + for folder_id, total in result.all(): + counts[folder_id] = int(total) + return counts + async def create( self, *, @@ -77,6 +222,8 @@ async def create( file_size: int, file_url: str, file_hash: str | None, + folder_id: int | None = None, + metadata_json: dict | None = None, ) -> UploadedPaper: record = UploadedPaper( user_id=user_id, @@ -86,11 +233,46 @@ async def create( file_size=file_size, file_url=file_url, file_hash=file_hash, + folder_id=folder_id, + metadata_json=metadata_json, ) self._session.add(record) await self._session.flush() return record + async def update_folder(self, record: UploadedPaper, folder_id: int | None) -> UploadedPaper: + record.folder_id = folder_id + await self._session.flush() + return record + + async def update_metadata(self, record: UploadedPaper, metadata_json: dict | None) -> UploadedPaper: + record.metadata_json = metadata_json + await self._session.flush() + return record + + async def update_file_fields( + self, + record: UploadedPaper, + *, + stored_filename: str, + file_url: str, + file_size: int, + file_hash: str | None, + content_type: str | None = None, + ) -> UploadedPaper: + record.stored_filename = stored_filename + record.file_url = file_url + record.file_size = file_size + record.file_hash = file_hash + if content_type is not None: + record.content_type = content_type + await self._session.flush() + return record + + async def delete(self, record: UploadedPaper) -> None: + await self._session.delete(record) + await self._session.flush() + class ParsedPaperCacheRepository: """Cache MinerU parse outputs by file hash.""" diff --git a/backend/app/models/__init__.py b/backend/app/models/__init__.py index 36d03da..7ed654a 100644 --- a/backend/app/models/__init__.py +++ b/backend/app/models/__init__.py @@ -2,6 +2,7 @@ # Importing these modules registers the SQLAlchemy models. from app.models import conversation # noqa: F401 +from app.models import library_folder # noqa: F401 from app.models import mineru_parse_job # noqa: F401 from app.models import parsed_paper_cache # noqa: F401 from app.models import paper # noqa: F401 diff --git a/backend/app/models/library_folder.py b/backend/app/models/library_folder.py new file mode 100644 index 0000000..d9e1a0c --- /dev/null +++ b/backend/app/models/library_folder.py @@ -0,0 +1,23 @@ +"""Model representing custom folders inside a user's personal library.""" +from datetime import datetime + +from sqlalchemy import Column, DateTime, ForeignKey, Integer, String +from sqlalchemy.orm import relationship + +from app.models.base import Base + + +class LibraryFolder(Base): + """User-scoped folder used to organize uploaded papers.""" + + __tablename__ = "library_folders" + + id = Column(Integer, primary_key=True, index=True) + user_id = Column(Integer, ForeignKey("users.id", ondelete="CASCADE"), nullable=False, index=True) + name = Column(String(120), nullable=False) + color = Column(String(32), nullable=True) + created_at = Column(DateTime, default=datetime.utcnow, nullable=False) + updated_at = Column(DateTime, default=datetime.utcnow, onupdate=datetime.utcnow, nullable=False) + + papers = relationship("UploadedPaper", back_populates="folder", cascade="all, delete", passive_deletes=True) + diff --git a/backend/app/models/uploaded_paper.py b/backend/app/models/uploaded_paper.py index 2c81acd..0a5e4e6 100644 --- a/backend/app/models/uploaded_paper.py +++ b/backend/app/models/uploaded_paper.py @@ -5,6 +5,7 @@ from sqlalchemy.orm import relationship from app.models.base import Base +from app.models.types import JSONBCompat class UploadedPaper(Base): @@ -20,6 +21,9 @@ class UploadedPaper(Base): file_size = Column(Integer, nullable=False) file_url = Column(String(512), nullable=False) file_hash = Column(String(128), nullable=True, index=True) + folder_id = Column(Integer, ForeignKey("library_folders.id", ondelete="SET NULL"), nullable=True, index=True) + metadata_json = Column("metadata", JSONBCompat(), nullable=True) uploaded_at = Column(DateTime, nullable=False, default=datetime.utcnow) user = relationship("User", backref="uploaded_papers") + folder = relationship("LibraryFolder", back_populates="papers") diff --git a/backend/app/models/user.py b/backend/app/models/user.py index 3795d7f..a738f4d 100644 --- a/backend/app/models/user.py +++ b/backend/app/models/user.py @@ -19,6 +19,8 @@ class User(Base): avatar_url = Column(String(512), nullable=True) research_interests = Column(Text, nullable=True) is_active = Column(Boolean, default=True, nullable=False) + oauth_provider = Column(String(50), nullable=True) + oauth_account_id = Column(String(255), nullable=True, index=True) created_at = Column(DateTime, default=datetime.utcnow, nullable=False) # 关系 diff --git a/backend/app/schemas/library.py b/backend/app/schemas/library.py new file mode 100644 index 0000000..7724119 --- /dev/null +++ b/backend/app/schemas/library.py @@ -0,0 +1,49 @@ +"""Pydantic schemas for personal library folders.""" +from datetime import datetime + +from pydantic import BaseModel, Field, validator + + +class LibraryFolderBase(BaseModel): + name: str = Field(..., min_length=1, max_length=120) + color: str | None = Field(None, max_length=32) + + +class LibraryFolderCreate(LibraryFolderBase): + pass + + +class LibraryFolderUpdate(LibraryFolderBase): + pass + + +class LibraryFolderSummary(BaseModel): + id: int + name: str + color: str | None = None + + class Config: + from_attributes = True + + +class LibraryFolderRead(LibraryFolderSummary): + paper_count: int = 0 + created_at: datetime + updated_at: datetime + + @validator("paper_count", pre=True, always=True) + def default_count(cls, value: int | None) -> int: # noqa: B902 + return value or 0 + + +class LibraryFolderListResponse(BaseModel): + folders: list[LibraryFolderRead] + unfiled_count: int = 0 + + +class LibraryFolderAssignment(BaseModel): + folder_id: int | None = Field(None, description="Target folder ID；null 表示移动到未分类") + + +class LibraryDownloadRequest(BaseModel): + pdf_url: str | None = Field(None, description="远程 PDF 下载链接，可覆盖已有 file_url") diff --git a/backend/app/schemas/paper.py b/backend/app/schemas/paper.py index a660978..d83dc46 100644 --- a/backend/app/schemas/paper.py +++ b/backend/app/schemas/paper.py @@ -3,6 +3,8 @@ from pydantic import BaseModel +from app.schemas import upload as upload_schema + class PaperBase(BaseModel): title: str @@ -37,3 +39,10 @@ class PaperImportRequest(PaperBase): """ pdf_url: str | None = None + folder_id: int | None = None + + +class PaperImportResponse(BaseModel): + paper: "PaperRead" + uploaded: upload_schema.UploadedPaperRead | None = None + folder_id: int | None = None diff --git a/backend/app/schemas/upload.py b/backend/app/schemas/upload.py index dbe8f08..215dea8 100644 --- a/backend/app/schemas/upload.py +++ b/backend/app/schemas/upload.py @@ -3,6 +3,8 @@ from pydantic import BaseModel +from app.schemas.library import LibraryFolderSummary + class UploadedPaperBase(BaseModel): original_filename: str @@ -10,11 +12,14 @@ class UploadedPaperBase(BaseModel): content_type: str file_size: int file_hash: str | None = None + folder_id: int | None = None + metadata_json: dict | None = None class UploadedPaperRead(UploadedPaperBase): id: int uploaded_at: datetime + folder: LibraryFolderSummary | None = None class Config: from_attributes = True diff --git a/backend/app/services/ai/llm_client.py b/backend/app/services/ai/llm_client.py index 2c01c02..96e591e 100644 --- a/backend/app/services/ai/llm_client.py +++ b/backend/app/services/ai/llm_client.py @@ -8,6 +8,9 @@ from aiohttp import ClientTimeout from app.core.config import settings +import logging + +logger = logging.getLogger(__name__) class DeepSeekClient: @@ -120,7 +123,15 @@ def _build_ssl_context(self) -> ssl.SSLContext: if self.ca_bundle: try: context.load_verify_locations(self.ca_bundle) - except Exception as exc: + except FileNotFoundError as exc: + # Fallback to default cert store when the configured CA bundle does not exist. + # Logging is intentionally simple since DeepSeek usage isn't critical in local dev. + logger.warning( + "DeepSeek CA bundle not found at %s, falling back to default cert store: %s", + self.ca_bundle, + exc, + ) + except Exception as exc: # pragma: no cover - defensive raise RuntimeError( f"Failed to load DeepSeek CA bundle at {self.ca_bundle}: {exc}" ) from exc diff --git a/backend/app/services/pdf_metadata.py b/backend/app/services/pdf_metadata.py new file mode 100644 index 0000000..40c614c --- /dev/null +++ b/backend/app/services/pdf_metadata.py @@ -0,0 +1,130 @@ +"""Lightweight PDF metadata extractor for organizing the personal library.""" +from __future__ import annotations + +import asyncio +from collections import Counter +from datetime import datetime +import re +from pathlib import Path +from typing import Any + +import fitz # PyMuPDF + +_TOKEN_PATTERN = re.compile(r"[\u4e00-\u9fa5]{2,}|[A-Za-z]{4,}") +_STOPWORDS = { + "abstract", + "introduction", + "research", + "paper", + "study", + "analysis", + "system", + "design", + "using", + "based", + "results", + "method", + "methods", + "conclusion", + "data", + "model", + "models", + "framework", + "information", + "对于", + "我们", + "提出", + "研究", + "方案", + "模型", + "数据", + "结果", + "本文", + "以及", + "进行", +} + + +def _summarize_text(text: str, max_length: int = 420) -> str | None: + cleaned = " ".join(part.strip() for part in text.splitlines() if part.strip()) + if not cleaned: + return None + if len(cleaned) <= max_length: + return cleaned + return f"{cleaned[:max_length].rstrip()}…" + + +def _extract_topics(text: str, limit: int = 4) -> list[str]: + if not text: + return [] + tokens = [token.lower() for token in _TOKEN_PATTERN.findall(text)] + filtered = [token for token in tokens if token not in _STOPWORDS and len(token) > 1] + if not filtered: + return [] + counter = Counter(filtered) + topics: list[str] = [] + for token, _ in counter.most_common(limit * 2): + if token not in topics: + topics.append(token) + if len(topics) >= limit: + break + return topics + + +def extract_pdf_metadata(pdf_path: Path, fallback_title: str | None = None) -> dict[str, Any]: + """Parse lightweight metadata such as title/author/summary/topics from a PDF.""" + result: dict[str, Any] = { + "title": fallback_title or pdf_path.stem, + "summary": None, + "topics": [], + "page_count": None, + "author": None, + "subject": None, + "keywords": None, + "extracted_at": datetime.utcnow().isoformat(), + } + + if not pdf_path.exists(): + return result + + doc = fitz.open(pdf_path) + try: + base_meta = doc.metadata or {} + title = (base_meta.get("title") or fallback_title or pdf_path.stem).strip() + author = (base_meta.get("author") or "").strip() or None + subject = (base_meta.get("subject") or "").strip() or None + keywords = (base_meta.get("keywords") or "").strip() or None + + text_parts: list[str] = [] + max_pages = min(doc.page_count, 3) + for page_index in range(max_pages): + try: + page = doc.load_page(page_index) + text_parts.append(page.get_text("text")) + except Exception: + continue + + combined_text = "\n".join(text_parts).strip() + summary = _summarize_text(combined_text) + topics = _extract_topics(combined_text) + + result.update( + { + "title": title or fallback_title or pdf_path.stem, + "author": author, + "subject": subject, + "keywords": keywords, + "summary": summary, + "topics": topics, + "page_count": doc.page_count, + } + ) + finally: + doc.close() + + return result + + +async def extract_pdf_metadata_async(pdf_path: Path, fallback_title: str | None = None) -> dict[str, Any]: + """Async wrapper for `extract_pdf_metadata` to keep FastAPI endpoints non-blocking.""" + return await asyncio.to_thread(extract_pdf_metadata, pdf_path, fallback_title) diff --git a/backend/migrations/versions/20251110_add_library_folders.py b/backend/migrations/versions/20251110_add_library_folders.py new file mode 100644 index 0000000..d5412c2 --- /dev/null +++ b/backend/migrations/versions/20251110_add_library_folders.py @@ -0,0 +1,55 @@ +"""Add library folders and metadata for uploaded papers.""" +from __future__ import annotations + +from alembic import op +import sqlalchemy as sa +from sqlalchemy.dialects import postgresql as psql + + +# revision identifiers, used by Alembic. +revision = "20251110_add_library_folders" +down_revision = "20251106_add_mineru_parse_jobs" +branch_labels = None +depends_on = None + + +def upgrade() -> None: + op.create_table( + "library_folders", + sa.Column("id", sa.Integer(), primary_key=True), + sa.Column("user_id", sa.Integer(), nullable=False), + sa.Column("name", sa.String(length=120), nullable=False), + sa.Column("color", sa.String(length=32), nullable=True), + sa.Column("created_at", sa.DateTime(), nullable=False, server_default=sa.func.now()), + sa.Column("updated_at", sa.DateTime(), nullable=False, server_default=sa.func.now()), + sa.ForeignKeyConstraint(["user_id"], ["users.id"], ondelete="CASCADE"), + ) + op.create_index("ix_library_folders_user_id", "library_folders", ["user_id"]) + op.create_unique_constraint( + "uq_library_folders_user_id_name", + "library_folders", + ["user_id", "name"], + ) + + op.add_column("uploaded_papers", sa.Column("folder_id", sa.Integer(), nullable=True)) + op.add_column("uploaded_papers", sa.Column("metadata", psql.JSONB(astext_type=sa.Text()), nullable=True)) + op.create_foreign_key( + "fk_uploaded_papers_folder_id", + "uploaded_papers", + "library_folders", + ["folder_id"], + ["id"], + ondelete="SET NULL", + ) + op.create_index("ix_uploaded_papers_folder_id", "uploaded_papers", ["folder_id"]) + + +def downgrade() -> None: + op.drop_index("ix_uploaded_papers_folder_id", table_name="uploaded_papers") + op.drop_constraint("fk_uploaded_papers_folder_id", "uploaded_papers", type_="foreignkey") + op.drop_column("uploaded_papers", "metadata") + op.drop_column("uploaded_papers", "folder_id") + + op.drop_constraint("uq_library_folders_user_id_name", "library_folders", type_="unique") + op.drop_index("ix_library_folders_user_id", table_name="library_folders") + op.drop_table("library_folders") diff --git a/backend/migrations/versions/20251112_add_oauth_fields.py b/backend/migrations/versions/20251112_add_oauth_fields.py new file mode 100644 index 0000000..6ed1eab --- /dev/null +++ b/backend/migrations/versions/20251112_add_oauth_fields.py @@ -0,0 +1,29 @@ +"""add oauth provider columns to users""" +from __future__ import annotations + +from alembic import op +import sqlalchemy as sa + + +# revision identifiers, used by Alembic. +revision = "20251112_add_oauth_fields" +down_revision = "20251110_add_library_folders" +branch_labels = None +depends_on = None + + +def upgrade() -> None: + op.add_column("users", sa.Column("oauth_provider", sa.String(length=50), nullable=True)) + op.add_column("users", sa.Column("oauth_account_id", sa.String(length=255), nullable=True)) + op.create_index( + "ix_users_oauth_provider_account", + "users", + ["oauth_provider", "oauth_account_id"], + unique=True, + ) + + +def downgrade() -> None: + op.drop_index("ix_users_oauth_provider_account", table_name="users") + op.drop_column("users", "oauth_account_id") + op.drop_column("users", "oauth_provider") diff --git a/frontend/src/app/auth/sso/page.tsx b/frontend/src/app/auth/sso/page.tsx new file mode 100644 index 0000000..037e2ad --- /dev/null +++ b/frontend/src/app/auth/sso/page.tsx @@ -0,0 +1,48 @@ +"use client"; + +import { useEffect, useState } from "react"; +import { useRouter, useSearchParams } from "next/navigation"; +import { Loader2 } from "lucide-react"; + +import { setAccessToken } from "@/lib/auth"; + +export default function OAuthBridgePage() { + const router = useRouter(); + const searchParams = useSearchParams(); + const [status, setStatus] = useState("正在处理登录，请稍候…"); + + useEffect(() => { + const errorParam = searchParams.get("error"); + const token = searchParams.get("token"); + const nextPath = searchParams.get("next") || "/"; + + if (errorParam) { + setStatus(`登录失败：${errorParam}`); + return; + } + + if (!token) { + setStatus("未收到登录凭据，请重新尝试。"); + return; + } + + setAccessToken(token); + setStatus("登录成功，正在跳转…"); + const timer = window.setTimeout(() => { + router.replace(nextPath); + }, 600); + + return () => window.clearTimeout(timer); + }, [router, searchParams]); + + return ( +
+
+
+
+

{status}

+
+
+ ); +} diff --git a/frontend/src/app/library/page.tsx b/frontend/src/app/library/page.tsx index e12b206..646fff1 100644 --- a/frontend/src/app/library/page.tsx +++ b/frontend/src/app/library/page.tsx @@ -2,11 +2,19 @@ import Link from "next/link"; import { useCallback, useEffect, useMemo, useState } from "react"; +import { useRouter } from "next/navigation"; import DashboardShell from "@/components/layout/dashboard-shell"; -import { fetchUploadedPapers, type UploadedPaperDTO } from "@/lib/api-client"; +import { + deleteUploadedPaper, + fetchUploadedPapers, + updateUploadedPaperFolder, + type LibraryFolderDTO, + type UploadedPaperDTO, +} from "@/lib/api-client"; import { getAccessToken } from "@/lib/auth"; import { BACKEND_URL } from "@/lib/config"; +import { useLibraryFolders } from "@/hooks/use-library-folders"; function formatTitle(filename: string): string { const withoutExt = filename.replace(/\.[^/.]+$/, ""); @@ -103,8 +111,12 @@ function getAdaptiveTitleClass(title: string): string { return "text-lg sm:text-xl"; } +type FolderFilter = "all" | "unfiled" | number; + +const FOLDER_COLORS = ["#0ea5e9", "#2563eb", "#10b981", "#f97316", "#f43f5e", "#a855f7", "#eab308"]; + export default function LibraryPage() { - const token = getAccessToken(); + const router = useRouter(); const [papers, setPapers] = useState([]); const [isLoading, setIsLoading] = useState(true); const [isRefreshing, setIsRefreshing] = useState(false); @@ -112,17 +124,65 @@ export default function LibraryPage() { const [page, setPage] = useState(1); const [total, setTotal] = useState(0); const [totalPages, setTotalPages] = useState(0); - const PAGE_SIZE = 12; + const [activeFilter, setActiveFilter] = useState("all"); + const [selectedPaper, setSelectedPaper] = useState(null); + const [isFolderModalOpen, setIsFolderModalOpen] = useState(false); + const [newFolderName, setNewFolderName] = useState(""); + const [newFolderColor, setNewFolderColor] = useState(FOLDER_COLORS[0]); + const [folderModalError, setFolderModalError] = useState(null); + const [libraryToken, setLibraryToken] = useState(null); + const [authChecked, setAuthChecked] = useState(false); + const [gridMode, setGridMode] = useState<"lg" | "xl">("xl"); + const [folderActionMessage, setFolderActionMessage] = useState(null); + const [isUpdatingFolder, setIsUpdatingFolder] = useState(false); + const [isDeletingPaper, setIsDeletingPaper] = useState(false); + const [deleteMessage, setDeleteMessage] = useState(null); + + const { + folders, + unfiledCount, + isLoading: isFolderLoading, + error: folderError, + refresh: refreshFolders, + create: createFolder, + } = useLibraryFolders(libraryToken); const downloadBase = useMemo( () => (BACKEND_URL.endsWith("/") ? BACKEND_URL.slice(0, -1) : BACKEND_URL), [], ); + useEffect(() => { + if (typeof window !== "undefined") { + setLibraryToken(getAccessToken()); + } + setAuthChecked(true); + }, []); + + useEffect(() => { + if (typeof window === "undefined") { + return; + } + const detectMode = () => { + setGridMode(window.innerWidth >= 1280 ? "xl" : "lg"); + }; + detectMode(); + window.addEventListener("resize", detectMode); + return () => window.removeEventListener("resize", detectMode); + }, []); + + const pageSize = gridMode === "xl" ? 10 : 8; + + const resolveFolderParam = useCallback((filter: FolderFilter) => { + if (filter === "all") return undefined; + if (filter === "unfiled") return 0; + return filter; + }, []); + const loadPage = useCallback( - async (targetPage: number, { showLoading = true }: { showLoading?: boolean } = {}) => { - const authToken = token; - if (!authToken) { + async (targetPage: number, filter: FolderFilter, { showLoading = true }: { showLoading?: boolean } = {}) => { + const tokenValue = libraryToken; + if (!tokenValue) { return; } @@ -132,14 +192,20 @@ export default function LibraryPage() { setError(null); try { - const data = await fetchUploadedPapers(authToken, { + const folderParam = resolveFolderParam(filter); + const data = await fetchUploadedPapers(tokenValue, { page: targetPage, - pageSize: PAGE_SIZE, + pageSize, + folderId: typeof folderParam === "number" ? folderParam : undefined, }); setPapers(data.items); setPage(data.page); setTotal(data.total); setTotalPages(data.total_pages); + setSelectedPaper((current) => { + if (!current) return null; + return data.items.find((item) => item.id === current.id) ?? null; + }); } catch (err) { const message = err instanceof Error ? err.message : "获取上传记录失败"; setError(message); @@ -149,11 +215,15 @@ export default function LibraryPage() { } } }, - [PAGE_SIZE, token], + [libraryToken, pageSize, resolveFolderParam], ); useEffect(() => { - if (!token) { + if (!authChecked) { + return; + } + + if (!libraryToken) { setIsLoading(false); setPapers([]); setTotal(0); @@ -162,11 +232,20 @@ export default function LibraryPage() { return; } - void loadPage(1); - }, [loadPage, token]); + void loadPage(1, activeFilter); + }, [activeFilter, authChecked, loadPage, libraryToken]); + + useEffect(() => { + setSelectedPaper(null); + }, [activeFilter]); + + useEffect(() => { + setFolderActionMessage(null); + setDeleteMessage(null); + }, [selectedPaper]); async function handleRefresh() { - if (!token) { + if (!libraryToken) { return; } @@ -174,7 +253,8 @@ export default function LibraryPage() { setError(null); try { - await loadPage(page, { showLoading: false }); + await loadPage(page, activeFilter, { showLoading: false }); + await refreshFolders(); } finally { setIsRefreshing(false); } @@ -187,16 +267,151 @@ export default function LibraryPage() { } if (direction === "prev" && page > 1) { - void loadPage(page - 1); + void loadPage(page - 1, activeFilter); return; } if (direction === "next" && totalPageCount > 0 && page < totalPageCount) { - void loadPage(page + 1); + void loadPage(page + 1, activeFilter); + } + } + + const totalLibraryCount = useMemo( + () => folders.reduce((sum, folder) => sum + (folder.paper_count ?? 0), 0) + unfiledCount, + [folders, unfiledCount], + ); + + const folderEntries = useMemo(() => { + const entries: Array<{ id: FolderFilter; name: string; count: number; color?: string | null }> = [ + { id: "all", name: "全部文献", count: totalLibraryCount }, + { id: "unfiled", name: "未分类", count: unfiledCount }, + ...folders.map((folder) => ({ + id: folder.id as FolderFilter, + name: folder.name, + color: folder.color ?? undefined, + count: folder.paper_count ?? 0, + })), + ]; + return entries; + }, [folders, totalLibraryCount, unfiledCount]); + + const activeFolderInfo = useMemo(() => { + if (activeFilter === "all") { + return { + name: "全部文献", + count: totalLibraryCount, + description: "查看你保存的所有论文，自动去重展示。", + }; + } + if (activeFilter === "unfiled") { + return { + name: "未分类", + count: unfiledCount, + description: "尚未归档的论文会暂存于此，方便快速整理。", + }; + } + const folder = folders.find((item) => item.id === activeFilter); + return { + name: folder?.name ?? "文献分类", + count: folder?.paper_count ?? 0, + color: folder?.color, + description: "该分类内的论文独立分页和管理。", + }; + }, [activeFilter, folders, totalLibraryCount, unfiledCount]); + + const handleFolderSelect = (filter: FolderFilter) => { + setActiveFilter(filter); + setPage(1); + void loadPage(1, filter); + }; + + async function handleCreateFolder() { + if (!libraryToken) { + setFolderModalError("请先登录后再创建文件夹"); + return; + } + const trimmed = newFolderName.trim(); + if (!trimmed) { + setFolderModalError("请填写文件夹名称"); + return; + } + + setFolderModalError(null); + try { + await createFolder({ name: trimmed, color: newFolderColor }); + await refreshFolders(); + setIsFolderModalOpen(false); + setNewFolderName(""); + } catch (err) { + const message = err instanceof Error ? err.message : "创建文件夹失败"; + setFolderModalError(message); + } + } + + async function handleAssignFolder(paperId: number, targetFolderId: number | null) { + if (!libraryToken) { + setFolderActionMessage("请先登录后再管理文献"); + return; + } + + setIsUpdatingFolder(true); + setFolderActionMessage(null); + try { + await updateUploadedPaperFolder(libraryToken, paperId, targetFolderId); + await loadPage(page, activeFilter, { showLoading: false }); + await refreshFolders(); + setFolderActionMessage("✓ 已更新文献所在文件夹"); + } catch (err) { + const message = err instanceof Error ? err.message : "更新文献文件夹失败"; + setFolderActionMessage(message); + } finally { + setIsUpdatingFolder(false); + } + } + + async function handleDeletePaper(paperId: number) { + if (!libraryToken) { + setDeleteMessage("请先登录后再删除文献"); + return; + } + + const confirmed = window.confirm("确定要删除这篇文献吗？此操作不可撤销。"); + if (!confirmed) { + return; + } + + setIsDeletingPaper(true); + setDeleteMessage(null); + try { + await deleteUploadedPaper(libraryToken, paperId); + setDeleteMessage("✓ 文献已删除"); + setSelectedPaper(null); + await refreshFolders(); + await loadPage(1, activeFilter); + } catch (err) { + const message = err instanceof Error ? err.message : "删除文献失败"; + setDeleteMessage(message); + } finally { + setIsDeletingPaper(false); } } - if (!token) { + if (!authChecked) { + return ( + +
+

初始化中…

+
+ + ); + } + + if (!libraryToken) { return ( -
-
-
-

上传论文

-

- 最近上传的论文会显示在这里，你可以快速下载或重新整理它们。 -

-
-
+
+
- - {error && ( -
- {error} + {folderError &&

{folderError}

} +
+ {folderEntries.map((entry) => { + const isActive = activeFilter === entry.id; + return ( + + ); + })} + {isFolderLoading && ( +

加载文件夹中…

+ )}
- )} - - {isLoading ? ( -
- {Array.from({ length: 10 }).map((_, index) => ( -
-
-
-
+ + +
+
+
+
+

+ {activeFolderInfo.name} +

+

{activeFolderInfo.description}

+
+
+ + + 上传新论文 +
- ))} -
- ) : papers.length === 0 ? ( -
- 📄 -
-

还没有上传的论文

-

- 试着上传第一篇 PDF 文献，创建你的个人文献库。 -

- - 去上传 - +
+ 共 {activeFolderInfo.count} 篇 + 已为你自动去重显示 +
- ) : ( -
- {papers.map((paper) => { - const title = formatTitle(paper.original_filename); - const downloadUrl = resolveDownloadUrl(downloadBase, paper.file_url); - return ( -
+ {error} +
+ )} + + {papers.length > 0 && ( + handlePageChange("prev")} + onNext={() => handlePageChange("next")} + disablePrev={page <= 1 || total === 0} + disableNext={total === 0 || (totalPages > 0 && page >= totalPages)} + /> + )} + + {isLoading ? ( +
+ {Array.from({ length: pageSize }).map((_, index) => ( +
- {/* PDF 图标区域 */} -
-
- +
+
+
+ ))} +
+ ) : papers.length === 0 ? ( +
+ 📄 +
+

此分类暂无文献

+

+ 上传或保存一篇 PDF，或切换到其它文件夹查看。 +

+
+ + 去上传 + +
+ ) : ( +
+ {papers.map((paper) => { + const title = formatTitle(paper.original_filename); + const downloadUrl = resolveDownloadUrl(downloadBase, paper.file_url); + const isActive = selectedPaper?.id === paper.id; + + return ( +
setSelectedPaper(paper)} + className={`group flex h-full cursor-pointer flex-col rounded-2xl border bg-white/80 p-4 shadow-sm transition hover:-translate-y-0.5 hover:shadow-lg dark:bg-slate-900/70 ${ + isActive ? "border-blue-500 dark:border-blue-400" : "border-slate-200 dark:border-slate-700" + }`} + > +
+
+

+ {getThumbnailLabel(title, "PDF")} +

+
+
+
+

+ {title} +

+

+ {formatFileSize(paper.file_size)} · 上传于 {formatDateTime(paper.uploaded_at)} +

+
+ + event.stopPropagation()} + className="flex flex-1 items-center justify-center rounded-lg border border-slate-200 px-3 py-1.5 font-semibold text-slate-700 transition hover:bg-slate-50 dark:border-slate-600 dark:text-slate-200 dark:hover:bg-slate-700/60" > - - + 下载 +
-
- - {/* 标题 */} -

- {title} -

- - {/* 文件大小 */} -

- {formatFileSize(paper.file_size)} -

- - {/* 底部操作栏 */} -
- - PDF - - - 下载 - -
-
- ); - })} -
- )} + + ); + })} +
+ )} + + setSelectedPaper(null)} + onAssignFolder={handleAssignFolder} + isAssigning={isUpdatingFolder} + actionMessage={folderActionMessage} + onDeletePaper={handleDeletePaper} + isDeleting={isDeletingPaper} + deleteMessage={deleteMessage} + /> +
+
- {papers.length > 0 && ( -
-

- 共 {total} 篇 · 每页 {PAGE_SIZE} 篇 -

-
+ {isFolderModalOpen && ( +
+
+

新建文件夹

+

为你的论文创建一个独立分类，方便管理。

+
+ +
+

颜色标记

+
+ {FOLDER_COLORS.map((color) => ( +
+
+ {folderModalError &&

{folderModalError}

} +
+
- - 第 {page} / {Math.max(totalPages || 1, 1)} 页 -
+
+ )} + + ); +} + +function MetadataPanel({ + paper, + onClose, + folders, + unfiledCount, + onAssignFolder, + isAssigning, + actionMessage, + onDeletePaper, + isDeleting, + deleteMessage, +}: { + paper: UploadedPaperDTO | null; + onClose: () => void; + folders: LibraryFolderDTO[]; + unfiledCount: number; + onAssignFolder: (paperId: number, folderId: number | null) => void; + isAssigning: boolean; + actionMessage: string | null; + onDeletePaper: (paperId: number) => void; + isDeleting: boolean; + deleteMessage: string | null; +}) { + if (!paper) { + return null; + } + + const metadata = (paper.metadata_json ?? {}) as Record; + const title = typeof metadata.title === "string" && metadata.title.trim().length > 0 ? metadata.title : formatTitle(paper.original_filename); + const author = typeof metadata.author === "string" ? metadata.author : null; + const summary = typeof metadata.summary === "string" ? metadata.summary : null; + const topics = Array.isArray(metadata.topics) + ? metadata.topics.filter((topic): topic is string => typeof topic === "string") + : []; + const subject = typeof metadata.subject === "string" ? metadata.subject : null; + const keywords = typeof metadata.keywords === "string" ? metadata.keywords : null; + const pageCount = + typeof metadata.page_count === "number" + ? metadata.page_count + : metadata.page_count && typeof metadata.page_count === "string" + ? Number(metadata.page_count) + : null; + + return ( +
+
+

文档元信息

+ +
+

{title}

+ {author &&

作者：{author}

} +
+

文件大小：{formatFileSize(paper.file_size)}

+ {pageCount &&

页数：{pageCount}

} +

上传时间：{formatDateTime(paper.uploaded_at)}

+ {subject &&

主题：{subject}

} + {keywords &&

关键词：{keywords}

} +
+ {summary && ( +
+

摘要

+

{summary}

+
+ )} + {topics.length > 0 && ( +
+ {topics.map((topic) => ( + + #{topic} + + ))} +
+ )} + +
+ ); +} + +function FolderManagementSection({ + paper, + folders, + unfiledCount, + onAssignFolder, + isAssigning, + actionMessage, + onDeletePaper, + isDeleting, + deleteMessage, +}: { + paper: UploadedPaperDTO; + folders: LibraryFolderDTO[]; + unfiledCount: number; + onAssignFolder: (paperId: number, folderId: number | null) => void; + isAssigning: boolean; + actionMessage: string | null; + onDeletePaper: (paperId: number) => void; + isDeleting: boolean; + deleteMessage: string | null; +}) { + const [selectedValue, setSelectedValue] = useState("unfiled"); + + useEffect(() => { + if (paper.folder_id) { + setSelectedValue(paper.folder_id.toString()); + } else { + setSelectedValue("unfiled"); + } + }, [paper]); + + const folderOptions = useMemo( + () => [ + { value: "unfiled", label: `未分类（${unfiledCount}）` }, + ...folders.map((folder) => ({ + value: folder.id.toString(), + label: `${folder.name}（${folder.paper_count ?? 0}）`, + })), + ], + [folders, unfiledCount], + ); + + return ( +
+

文献管理

+ + + {actionMessage &&

{actionMessage}

} + + {deleteMessage &&

{deleteMessage}

} +
+ ); +} + +function PageTurner({ + page, + totalPages, + pageSize, + onPrev, + onNext, + disablePrev, + disableNext, +}: { + page: number; + totalPages: number; + pageSize?: number; + onPrev: () => void; + onNext: () => void; + disablePrev?: boolean; + disableNext?: boolean; +}) { + if (totalPages <= 1) { + return null; + } + + return ( +
+ +
+
+ 第 + {page} + / {totalPages} 页 +
+ {pageSize && ( + 每页 {pageSize} 篇（2 行排布） )}
- + +
); } diff --git a/frontend/src/app/smart-reading/page.tsx b/frontend/src/app/smart-reading/page.tsx index 7e8ef03..1fbdd60 100644 --- a/frontend/src/app/smart-reading/page.tsx +++ b/frontend/src/app/smart-reading/page.tsx @@ -1,6 +1,6 @@ "use client"; -import { useRouter } from "next/navigation"; +import { useRouter, useSearchParams } from "next/navigation"; import { useEffect, useState, useRef, useCallback } from "react"; import { UploadCloud, BookOpenCheck, XCircle, AlertTriangle } from "lucide-react"; @@ -8,7 +8,7 @@ import DashboardShell from "@/components/layout/dashboard-shell"; import AgentChatPanel from "@/components/ai-agent/AgentChatPanel"; import { PDFViewer } from "@/components/pdf-reader"; import { getAccessToken } from "@/lib/auth"; -import { uploadPaper, type UploadedPaperDTO } from "@/lib/api-client"; +import { uploadPaper, fetchUploadedPaper, type UploadedPaperDTO } from "@/lib/api-client"; import { BACKEND_URL } from "@/lib/config"; import type { PDFParseProgress, PDFParseResult, MinerUParseResult } from "@/types/pdf-reader"; @@ -20,9 +20,9 @@ import type { PDFParseProgress, PDFParseResult, MinerUParseResult } from "@/type */ export default function SmartReadingPage() { const router = useRouter(); + const searchParams = useSearchParams(); const fileInputRef = useRef(null); - const [, setToken] = useState(null); const [selectedPdfUrl, setSelectedPdfUrl] = useState(null); const [uploadedPaper, setUploadedPaper] = useState(null); const [isUploading, setIsUploading] = useState(false); @@ -43,11 +43,38 @@ export default function SmartReadingPage() { const accessToken = getAccessToken(); if (!accessToken) { router.push("/auth"); - return; } - setToken(accessToken); }, [router]); + const preloadLibraryPaper = useCallback( + async (paperId: number) => { + const authToken = getAccessToken(); + if (!authToken) { + router.push("/auth"); + return; + } + + setUploadError(null); + setParseError(null); + setParseProgress(null); + setParsedDocument(null); + setIsParsing(false); + setMineruResult(null); + + try { + const paper = await fetchUploadedPaper(authToken, paperId); + setUploadedPaper(paper); + const downloadUrl = new URL(`/api/v1/library/uploads/${paperId}/download`, BACKEND_URL); + downloadUrl.searchParams.set("access_token", authToken); + setSelectedPdfUrl(downloadUrl.toString()); + } catch (err) { + const message = err instanceof Error ? err.message : "加载文库文档失败"; + setUploadError(message); + } + }, + [router], + ); + // 处理PDF文件上传 /** * Validate and upload a PDF, then wire resulting metadata into the viewer. @@ -78,17 +105,17 @@ export default function SmartReadingPage() { throw new Error("用户未登录"); } - const uploadedPaper = await uploadPaper(token, file); + const uploadedPaper = await uploadPaper(token, file); - // 构建PDF访问URL，确保使用混合运行时提供的后端地址 - const pdfUrl = new URL(uploadedPaper.file_url, BACKEND_URL).toString(); - setSelectedPdfUrl(pdfUrl); - setUploadedPaper(uploadedPaper); - setParseProgress(null); - setParseError(null); - setParsedDocument(null); - setMineruResult(null); - setIsParsing(false); + // 构建PDF访问URL，确保使用混合运行时提供的后端地址 + const pdfUrl = new URL(uploadedPaper.file_url, BACKEND_URL).toString(); + setSelectedPdfUrl(pdfUrl); + setUploadedPaper(uploadedPaper); + setParseProgress(null); + setParseError(null); + setParsedDocument(null); + setMineruResult(null); + setIsParsing(false); console.log("PDF上传成功:", uploadedPaper); } catch (error) { @@ -100,6 +127,18 @@ export default function SmartReadingPage() { } }; + useEffect(() => { + const paperIdParam = searchParams.get("paperId"); + if (!paperIdParam) { + return; + } + const numericId = Number(paperIdParam); + if (Number.isNaN(numericId)) { + return; + } + void preloadLibraryPaper(numericId); + }, [preloadLibraryPaper, searchParams]); + const handlePdfError = (error: string) => { console.error("PDF加载错误:", error); setParseError(error); diff --git a/frontend/src/app/watch/page.tsx b/frontend/src/app/watch/page.tsx new file mode 100644 index 0000000..f933ec1 --- /dev/null +++ b/frontend/src/app/watch/page.tsx @@ -0,0 +1,593 @@ +"use client"; + +import { useMemo, useRef, useState } from "react"; +import { useRouter } from "next/navigation"; +import type { LucideIcon } from "lucide-react"; +import { + ArrowRight, + BadgeHelp, + ChevronLeft, + ChevronRight, + Clapperboard, + Film, + Play, + Sparkles, + Video, + Wand2, +} from "lucide-react"; + +import DashboardShell from "@/components/layout/dashboard-shell"; + +type SpotlightPaper = { + id: string; + title: string; + summary: string; + tags: string[]; + vibe: string; + duration: string; + cover: string; +}; + +type CinemaRow = { + title: string; + description: string; + accent: string; + icon: LucideIcon; + papers: Array<{ + id: string; + title: string; + highlight: string; + tags: string[]; + duration: string; + }>; +}; + +const HERO_BADGES: Array<{ title: string; detail: string; icon: LucideIcon }> = [ + { + title: "章节分镜", + detail: "一键拉出 8 个镜头 + 导航字幕", + icon: Clapperboard, + }, + { + title: "AI 配音", + detail: "公式、图表自动补位旁白", + icon: Sparkles, + }, + { + title: "沉浸注释", + detail: "重点段落自动贴边批注", + icon: Wand2, + }, +]; + +const HERO_STATS = [ + { label: "实时镜头生成", value: "2.3s", description: "平均首镜头延迟" }, + { label: "沉浸场次", value: "1,280+", description: "本周开启" }, + { label: "长文档通过率", value: "98%", description: "≤200 页 PDF" }, +]; + +const SPOTLIGHT_PAPERS: SpotlightPaper[] = [ + { + id: "1706.03762", + title: "Attention Is All You Need", + summary: "用 18 个镜头复盘 Transformer 的诞生，搭配自注意力动效与结构拆解。", + tags: ["NLP", "Transformer"], + vibe: "分镜式推演", + duration: "12 分钟", + cover: "https://images.unsplash.com/photo-1520607162513-77705c0f0d4a?auto=format&fit=crop&w=1600&q=80", + }, + { + id: "2302.13971", + title: "Segment Anything", + summary: "逐帧对齐 SAM 掩膜生成过程，辅助动态图层说明 prompt 影响。", + tags: ["CV", "图像分割"], + vibe: "逐帧多镜头", + duration: "10 分钟", + cover: "https://images.unsplash.com/photo-1498050108023-c5249f4df085?auto=format&fit=crop&w=1600&q=80", + }, + { + id: "2310.02992", + title: "Gemini: A Family of LMs", + summary: "多模态长镜头，展示音频 / 图像提示链路与跨模态推理策略。", + tags: ["多模态", "LLM"], + vibe: "全景式讲解", + duration: "15 分钟", + cover: "https://images.unsplash.com/photo-1518770660439-4636190af475?auto=format&fit=crop&w=1600&q=80", + }, + { + id: "2402.03300", + title: "Sora Technical Report", + summary: "用胶片质感呈现生成视频模型的 3 段训练周期与失真修复策略。", + tags: ["生成式视频", "Diffusion"], + vibe: "电影化拆解", + duration: "9 分钟", + cover: "https://images.unsplash.com/photo-1521737604893-d14cc237f11d?auto=format&fit=crop&w=1600&q=80", + }, + { + id: "2201.11903", + title: "Chain-of-Thought Prompting", + summary: "节奏较慢的旁白视角，结合板书式字幕，帮助理解 COT 提示结构。", + tags: ["Prompting", "推理"], + vibe: "慢速讲堂", + duration: "11 分钟", + cover: "https://images.unsplash.com/photo-1454496522488-7a8e488e8606?auto=format&fit=crop&w=1600&q=80", + }, +]; + +const CINEMA_ROWS: CinemaRow[] = [ + { + title: "多模态放映厅", + description: "图像 + 文本 + 音频联动，逐层拆解感知节点。", + accent: "from-sky-500/20 via-blue-500/10 to-transparent", + icon: Video, + papers: [ + { + id: "2304.08485", + title: "LLaVA", + highlight: "图像问答与字幕联合导览", + tags: ["多模态", "指令调优"], + duration: "14 分钟", + }, + { + id: "2305.11644", + title: "AudioPaLM", + highlight: "语音 + 文本统一对齐之旅", + tags: ["语音", "统一建模"], + duration: "16 分钟", + }, + ], + }, + { + title: "LLM 编剧场", + description: "聚焦推理、效率与工具编排的长镜头解构。", + accent: "from-fuchsia-500/15 via-purple-500/10 to-transparent", + icon: Film, + papers: [ + { + id: "2306.11644", + title: "Toolformer", + highlight: "让模型自己挑镜头的工具调度", + tags: ["Agent", "工具调用"], + duration: "13 分钟", + }, + { + id: "2304.03442", + title: "QLoRA", + highlight: "低比特场景的压感式呈现", + tags: ["高效微调", "量化"], + duration: "9 分钟", + }, + ], + }, + { + title: "视觉工作流", + description: "把渲染、生成和编辑的关键帧堆叠成可浏览时间线。", + accent: "from-amber-500/15 via-orange-500/10 to-transparent", + icon: Clapperboard, + papers: [ + { + id: "2312.02144", + title: "Gaussian Splatting 2.0", + highlight: "实时 3D 场景与光追调度", + tags: ["3D", "NeRF"], + duration: "8 分钟", + }, + { + id: "2403.13035", + title: "DreamCraft3D", + highlight: "从草图到立体的逐帧生成流程", + tags: ["3D 生成", "Diffusion"], + duration: "12 分钟", + }, + ], + }, +]; + +const FALLBACK_ARXIV_ID = "1706.03762"; + +export default function WatchPage() { + return ( + +
+ + + +
+ + ); +} + +function HeroSection() { + const router = useRouter(); + const [input, setInput] = useState(""); + const [error, setError] = useState(null); + const [isLaunching, setIsLaunching] = useState(false); + + function handleSubmit(event: React.FormEvent) { + event.preventDefault(); + const normalized = normalizeArxivInput(input); + if (!normalized) { + setError("请输入有效的 arXiv 链接或编号"); + return; + } + setError(null); + setIsLaunching(true); + window.open(`https://arxiv.org/abs/${normalized}`, "_blank", "noopener,noreferrer"); + setTimeout(() => { + setIsLaunching(false); + }, 600); + } + + return ( +
+
+
+
+
+
+ +
+
+

+ + 即时把 PDF 变成电影感播放 +

+

+ 输入任意 arXiv 链接，
3 秒钟开播一场“论文放映” +

+

+ InsightReading 的“看论文”模式会自动解析章节结构、生成镜头字幕、搭配旁白，把冗长 PDF 拆成一场更容易吸收的沉浸式放映。 +

+ +
+
+ +
+ setInput(event.target.value)} + className="w-full bg-transparent placeholder:text-slate-400 focus:outline-none" + disabled={isLaunching} + /> +
+
+ + +
+
+ {error ? ( +

{error}

+ ) : ( +

+ 没有链接？先体验经典案例： + +

+ )} +
+ +
+ {HERO_BADGES.map((badge) => { + const Icon = badge.icon; + return ( +
+
+ +
+
+

{badge.title}

+

{badge.detail}

+
+
+ ); + })} +
+
+ +
+

+ IMMERSIVE LIVEBOARD +

+

放映中的体验

+
+ {HERO_STATS.map((stat) => ( +
+

{stat.label}

+

{stat.value}

+

{stat.description}

+
+ ))} +
+
+

支持来源

+

+ arXiv / Semantic Scholar / OpenAlex / 自建文库。自动去重，统一字幕模版。 +

+
+
+
+
+ ); +} + +function PaperShowcaseCarousel() { + const containerRef = useRef(null); + const [activeIndex, setActiveIndex] = useState(0); + + const cards = useMemo(() => SPOTLIGHT_PAPERS, []); + + function scroll(direction: "prev" | "next") { + const node = containerRef.current; + if (!node) return; + const delta = direction === "next" ? node.clientWidth * 0.9 : -node.clientWidth * 0.9; + node.scrollBy({ left: delta, behavior: "smooth" }); + } + + return ( +
+
+
+

Spotlight

+

热门放映场次

+

+ 精选近两周播放率最高的论文场次，已为每篇文档准备好字幕、旁白和章节要点。左右滑动即可预览每个镜头的气质。 +

+
+
+ + +
+
+ +
{ + const node = containerRef.current; + if (!node) return; + const midpoint = node.scrollLeft + node.clientWidth / 2; + let closestIndex = 0; + let closestDistance = Number.POSITIVE_INFINITY; + node.childNodes.forEach((child, index) => { + if (!(child instanceof HTMLElement)) { + return; + } + const childCenter = child.offsetLeft + child.clientWidth / 2; + const distance = Math.abs(childCenter - midpoint); + if (distance < closestDistance) { + closestDistance = distance; + closestIndex = index; + } + }); + setActiveIndex(closestIndex); + }} + className="flex snap-x snap-mandatory gap-6 overflow-x-auto pb-4" + > + {cards.map((paper, index) => { + const isActive = index === activeIndex; + return ( +
+
+
+
+
+
+ + {paper.vibe} +
+

片长 {paper.duration}

+
+
+
+
+

ArXiv · {paper.id}

+

{paper.title}

+

{paper.summary}

+
+
+ {paper.tags.map((tag) => ( + + {tag} + + ))} +
+ +
+
+ ); + })} +
+
+ ); +} + +function CinemaReelGrid() { + return ( +
+
+
+

Immersive Tracks

+

主题影厅

+

+ 不同研究方向配不同镜头语言：多模态看特效，LLM 看脚本，视觉模型看调色。挑一个影厅，直接点播对应的论文合集。 +

+
+ +
+ +
+ {CINEMA_ROWS.map((row) => { + const Icon = row.icon; + return ( +
+
+
+
+
+

+ + Movie Track +

+

{row.title}

+

{row.description}

+
+ +
+
+ {row.papers.map((paper) => ( +
+

+ #{paper.id} +

+

{paper.title}

+

{paper.highlight}

+
+ {paper.tags.map((tag) => ( + + {tag} + + ))} +
+
+ 片长 {paper.duration} + +
+
+ ))} +
+
+
+ ); + })} +
+
+ ); +} + +function normalizeArxivInput(raw: string): string | null { + if (!raw) { + return null; + } + const trimmed = raw.trim(); + if (!trimmed) { + return null; + } + + const sanitized = trimmed.replace(/\.pdf$/i, ""); + const match = sanitized.match( + /(?:arxiv(?:\.org)?\/(?:abs|pdf)\/|arxiv:)?([A-Za-z-]+\/\d{7}|\d{4}\.\d{4,5})(v\d+)?/i, + ); + if (!match) { + return null; + } + + const core = match[1]?.toLowerCase(); + if (!core) { + return null; + } + + const version = match[2]?.toLowerCase() ?? ""; + return `${core}${version}`; +} diff --git a/frontend/src/components/Recommendations.tsx b/frontend/src/components/Recommendations.tsx index e3b6afe..a13859a 100644 --- a/frontend/src/components/Recommendations.tsx +++ b/frontend/src/components/Recommendations.tsx @@ -2,7 +2,6 @@ import { useCallback, useEffect, useMemo, useState } from "react"; -import { BACKEND_URL } from "@/lib/config"; import { getAccessToken } from "@/lib/auth"; import { fetchCurrentUser, @@ -12,6 +11,8 @@ import { } from "@/lib/api-client"; import { searchPapers } from "@/lib/api/academic"; import type { Paper as AcademicPaper } from "@/lib/types/academic"; +import { useLibraryFolders } from "@/hooks/use-library-folders"; +import { buildBackendUrl } from "@/lib/url"; interface AuthorDetail { name?: string | null; @@ -49,6 +50,14 @@ const DEFAULT_TAGS = ["Machine Learning", "Natural Language Processing"]; const MAX_SAVED_LIBRARY_PAGES = 5; const UPLOADS_PAGE_SIZE = 60; const TITLE_CHAR_PATTERN = /[\p{L}\p{N}]/gu; +type FolderSelection = number | "unfiled"; + +interface PaperImportResponse { + paper: { + id: number; + }; + uploaded?: UploadedPaperDTO | null; +} function sanitizeTitleKey(value?: string | null): string | null { if (!value) { @@ -233,27 +242,6 @@ function mapAcademicPaper(paper: AcademicPaper): Paper { }; } -/** 构造后端请求的完整 URL（自动处理开发/生产环境） */ -function buildBackendUrl(path: string): string { - let base = BACKEND_URL?.trim() || ""; - - // If a public/browser-facing backend URL isn't provided via env (NEXT_PUBLIC_BACKEND_URL), - // fall back to using the current origin in the browser so fetches target the same host/port - // the page was served from. On the server (SSR/build) the value from `config.ts` will - // provide the internal backend address (e.g. `http://backend:8000`) when available. - if (typeof window !== "undefined") { - if (!base) { - const { protocol, hostname, port } = window.location; - base = `${protocol}//${hostname}${port ? `:${port}` : ""}`; - } - } - - // 确保 path 拼接时没有多余斜杠 - base = base.endsWith("/") ? base.slice(0, -1) : base; - path = path.startsWith("/") ? path : `/${path}`; - return `${base}${path}`; -} - async function fetchRecommendationPapers(query: string, sources: string[]) { const activeSources = sources.length ? sources : ["arxiv"]; const response = await searchPapers({ @@ -270,6 +258,7 @@ async function fetchRecommendationPapers(query: string, sources: string[]) { } export default function Recommendations() { + const token = getAccessToken(); const [tagInput, setTagInput] = useState(""); const [tags, setTags] = useState(DEFAULT_TAGS); const [papers, setPapers] = useState([]); @@ -288,6 +277,19 @@ export default function Recommendations() { const [savedPaperKeys, setSavedPaperKeys] = useState>(new Set()); const [savedPapersLoading, setSavedPapersLoading] = useState(false); const [savedPapersError, setSavedPapersError] = useState(null); + const { + folders, + isLoading: foldersLoading, + refresh: refreshFolders, + create: createFolder, + } = useLibraryFolders(token); + const [folderPickerPaper, setFolderPickerPaper] = useState(null); + const [selectedFolderId, setSelectedFolderId] = useState("unfiled"); + const [folderPickerError, setFolderPickerError] = useState(null); + const [isSavingToLibrary, setIsSavingToLibrary] = useState(false); + const [newFolderInline, setNewFolderInline] = useState(""); + const [isCreatingFolderInline, setIsCreatingFolderInline] = useState(false); + const [inlineFolderError, setInlineFolderError] = useState(null); const visiblePapers = useMemo(() => { if (!hideSavedPapers) { @@ -302,6 +304,12 @@ export default function Recommendations() { }); }, [hideSavedPapers, papers, savedPaperKeys]); + useEffect(() => { + if (folderPickerPaper) { + setSelectedFolderId(folders[0]?.id ?? "unfiled"); + } + }, [folderPickerPaper, folders]); + const hiddenCount = hideSavedPapers ? papers.length - visiblePapers.length : 0; const activeTagLabel = activeTag ?? tags[0] ?? "未设置"; const visibleCountLabel = @@ -397,6 +405,117 @@ export default function Recommendations() { } } + const handleOpenFolderPicker = useCallback( + (paper: Paper) => { + setFolderPickerPaper(paper); + setFolderPickerError(null); + setSelectedFolderId(folders[0]?.id ?? "unfiled"); + setInlineFolderError(null); + setNewFolderInline(""); + }, + [folders], + ); + + const handleConfirmSave = useCallback(async () => { + if (!folderPickerPaper) { + return; + } + const authToken = getAccessToken(); + if (!authToken) { + setFolderPickerError("请先登录后再保存论文"); + return; + } + + setIsSavingToLibrary(true); + setFolderPickerError(null); + + try { + let pdfToSend = folderPickerPaper.link ?? folderPickerPaper.url ?? ""; + if (pdfToSend.includes("arxiv.org/abs/")) { + pdfToSend = pdfToSend.replace("/abs/", "/pdf/"); + if (!pdfToSend.endsWith(".pdf")) { + pdfToSend = `${pdfToSend}.pdf`; + } + } + + const payload = { + title: folderPickerPaper.title, + abstract: folderPickerPaper.summary ?? folderPickerPaper.abstract, + pdf_url: pdfToSend || undefined, + folder_id: typeof selectedFolderId === "number" ? selectedFolderId : undefined, + }; + + const response = await fetch(buildBackendUrl("/api/v1/papers/import"), { + method: "POST", + headers: { + "Content-Type": "application/json", + Authorization: `Bearer ${authToken}`, + }, + body: JSON.stringify(payload), + }); + + if (!response.ok) { + const detail = (await response.json().catch(() => ({}))) as { detail?: string }; + throw new Error(detail.detail ?? "保存论文失败"); + } + + const importData = (await response.json().catch(() => null)) as PaperImportResponse | null; + const uploadedId = importData?.uploaded?.id; + + if (uploadedId) { + try { + await fetch(buildBackendUrl(`/api/v1/library/uploads/${uploadedId}/ensure-local`), { + method: "POST", + headers: { + "Content-Type": "application/json", + Authorization: `Bearer ${authToken}`, + }, + body: JSON.stringify({ pdf_url: pdfToSend || undefined }), + }); + } catch (downloadErr) { + console.error("Force download failed:", downloadErr); + setFolderPickerError("论文已保存，但下载 PDF 失败，请稍后在文库中重试。"); + } + } + + await loadSavedPaperKeys(); + await refreshFolders(); + setFolderPickerPaper(null); + setSelectedFolderId("unfiled"); + alert('✓ 已保存到你的文献库！\n可以在"我的文库"页面查看。'); + } catch (err) { + const message = err instanceof Error ? err.message : "保存论文失败，请重试"; + setFolderPickerError(message); + } finally { + setIsSavingToLibrary(false); + } + }, [folderPickerPaper, loadSavedPaperKeys, refreshFolders, selectedFolderId]); + + const handleCreateFolderInline = useCallback(async () => { + if (!newFolderInline.trim()) { + setInlineFolderError("请输入文件夹名称"); + return; + } + if (!token) { + setInlineFolderError("请先登录后再创建文件夹"); + return; + } + + setInlineFolderError(null); + setIsCreatingFolderInline(true); + try { + const created = await createFolder({ name: newFolderInline.trim() }); + await refreshFolders(); + setNewFolderInline(""); + setSelectedFolderId(created.id); + } catch (err) { + const message = err instanceof Error ? err.message : "创建文件夹失败"; + setInlineFolderError(message); + } finally { + setIsCreatingFolderInline(false); + } + }, [createFolder, newFolderInline, refreshFolders, token]); + function handleTagUpdate() { const newTag = tagInput.trim(); if (newTag) { @@ -730,165 +849,7 @@ export default function Recommendations() {
+ {folderPickerPaper && ( +
+
+

保存到我的文库

+

请选择要保存的文件夹，或创建一个新的分类。

+

+ {folderPickerPaper.title} +

+
+ + {folders.map((folder) => ( + + ))} + {foldersLoading && ( +

文件夹加载中…

+ )} +
+
+

快速创建文件夹

+
+ setNewFolderInline(event.target.value)} + placeholder="输入文件夹名称" + className="flex-1 rounded-lg border border-slate-200 px-3 py-2 text-sm text-slate-800 focus:border-blue-500 focus:outline-none focus:ring-2 focus:ring-blue-100 dark:border-slate-600 dark:bg-slate-800 dark:text-slate-100 dark:focus:ring-blue-500/30" + /> + +
+ {inlineFolderError &&

{inlineFolderError}

} +
+ {folderPickerError && ( +

{folderPickerError}

+ )} +
+ + +
+
+
+ )}
); } diff --git a/frontend/src/components/auth-form.tsx b/frontend/src/components/auth-form.tsx index 185e156..2f3f95c 100644 --- a/frontend/src/components/auth-form.tsx +++ b/frontend/src/components/auth-form.tsx @@ -2,6 +2,7 @@ import { useEffect, useMemo, useState, type FormEvent } from "react"; import { useRouter } from "next/navigation"; +import { Github } from "lucide-react"; import { fetchCurrentUser, @@ -12,6 +13,7 @@ import { type UserDTO, } from "@/lib/api-client"; import { clearAccessToken, getAccessToken, setAccessToken } from "@/lib/auth"; +import { buildBackendUrl } from "@/lib/url"; type Mode = "login" | "register"; @@ -126,6 +128,16 @@ export function AuthForm() { setSuccess("已安全退出。"); } + function handleGitHubLogin() { + if (typeof window === "undefined") { + return; + } + const nextPath = window.location.pathname || "/"; + const loginUrl = new URL(buildBackendUrl("/api/v1/auth/github/login")); + loginUrl.searchParams.set("next", nextPath); + window.location.href = loginUrl.toString(); + } + return (
@@ -223,6 +235,19 @@ export function AuthForm() { +
+

或使用第三方账号

+ +
+ {currentUser && (
diff --git a/frontend/src/components/layout/dashboard-shell.tsx b/frontend/src/components/layout/dashboard-shell.tsx index 472889c..1ad6230 100644 --- a/frontend/src/components/layout/dashboard-shell.tsx +++ b/frontend/src/components/layout/dashboard-shell.tsx @@ -69,6 +69,7 @@ interface NavItem { label: string; icon: React.ReactNode; children?: string[]; + path?: string; } const NAV_ITEMS: NavItem[] = [ @@ -96,7 +97,8 @@ const NAV_ITEMS: NavItem[] = [ - ) + ), + path: "/watch", }, { label: "设置", @@ -128,6 +130,7 @@ const PATHNAME_TO_CHILD: Record = { "/upload": "上传文件", "/smart-reading": "智能阅读", "/library": "我的文库", + "/watch": "看论文", }; interface DashboardShellProps { @@ -572,6 +575,9 @@ export default function DashboardShell({ } // 设置该项为激活状态 setActiveNavItem(item.label); + if (!hasChildren && item.path) { + router.push(item.path); + } return; } @@ -579,6 +585,9 @@ export default function DashboardShell({ setActiveNavItem(item.label); if (!hasChildren) { + if (item.path) { + router.push(item.path); + } return; } // 切换子菜单的开关状态 diff --git a/frontend/src/components/pdf-reader/PDFJSViewer.tsx b/frontend/src/components/pdf-reader/PDFJSViewer.tsx index 88bb07f..e42319e 100644 --- a/frontend/src/components/pdf-reader/PDFJSViewer.tsx +++ b/frontend/src/components/pdf-reader/PDFJSViewer.tsx @@ -330,7 +330,7 @@ export default function PDFJSViewer({ }`} title={showOverlay ? "隐藏智能选择" : "显示智能选择"} > - {showOverlay ? "🎯 智能选择" : "⭕ 普通模式"} + {showOverlay ? "智能选择" : "普通模式"}
@@ -513,9 +513,7 @@ function PDFPageCanvas({ if (textLayer) { textLayer.innerHTML = ""; - // Set the required CSS variable for textLayer - textLayer.style.setProperty('--scale-factor', viewport.scale.toString()); - + textLayer.style.setProperty("--scale-factor", `${viewport.scale}`); try { const textContent = await page.getTextContent(); if (isStale()) { diff --git a/frontend/src/hooks/use-library-folders.ts b/frontend/src/hooks/use-library-folders.ts new file mode 100644 index 0000000..e3efcca --- /dev/null +++ b/frontend/src/hooks/use-library-folders.ts @@ -0,0 +1,114 @@ +import { useCallback, useEffect, useMemo, useState } from "react"; + +import { + createLibraryFolder, + deleteLibraryFolder, + fetchLibraryFolders, + type LibraryFolderDTO, + type LibraryFolderListResponse, + type LibraryFolderPayload, + updateLibraryFolder, +} from "@/lib/api-client"; + +interface UseLibraryFoldersResult { + folders: LibraryFolderDTO[]; + unfiledCount: number; + isLoading: boolean; + error: string | null; + refresh: () => Promise; + create: (payload: LibraryFolderPayload) => Promise; + rename: (folderId: number, payload: LibraryFolderPayload) => Promise; + remove: (folderId: number) => Promise; +} + +export function useLibraryFolders(token: string | null): UseLibraryFoldersResult { + const [folders, setFolders] = useState([]); + const [unfiledCount, setUnfiledCount] = useState(0); + const [isLoading, setIsLoading] = useState(false); + const [error, setError] = useState(null); + + const hasToken = useMemo(() => Boolean(token), [token]); + + const handleResponse = useCallback((data: LibraryFolderListResponse) => { + setFolders( + data.folders.map((folder) => ({ + ...folder, + paper_count: folder.paper_count ?? 0, + })), + ); + setUnfiledCount(data.unfiled_count ?? 0); + }, []); + + const refresh = useCallback(async () => { + if (!hasToken || !token) { + setFolders([]); + setUnfiledCount(0); + setError(null); + return; + } + + setIsLoading(true); + setError(null); + try { + const data = await fetchLibraryFolders(token); + handleResponse(data); + } catch (err) { + const message = err instanceof Error ? err.message : "加载文件夹失败"; + setError(message); + } finally { + setIsLoading(false); + } + }, [handleResponse, hasToken, token]); + + useEffect(() => { + void refresh(); + }, [refresh]); + + const create = useCallback( + async (payload: LibraryFolderPayload) => { + if (!token) { + throw new Error("当前未登录，无法创建文件夹"); + } + const folder = await createLibraryFolder(token, payload); + setFolders((prev) => [...prev, { ...folder, paper_count: folder.paper_count ?? 0 }]); + return folder; + }, + [token], + ); + + const rename = useCallback( + async (folderId: number, payload: LibraryFolderPayload) => { + if (!token) { + throw new Error("当前未登录，无法更新文件夹"); + } + const updated = await updateLibraryFolder(token, folderId, payload); + setFolders((prev) => + prev.map((folder) => (folder.id === folderId ? { ...updated, paper_count: updated.paper_count ?? folder.paper_count ?? 0 } : folder)), + ); + return updated; + }, + [token], + ); + + const remove = useCallback( + async (folderId: number) => { + if (!token) { + throw new Error("当前未登录，无法删除文件夹"); + } + await deleteLibraryFolder(token, folderId); + setFolders((prev) => prev.filter((folder) => folder.id !== folderId)); + }, + [token], + ); + + return { + folders, + unfiledCount, + isLoading, + error, + refresh, + create, + rename, + remove, + }; +} diff --git a/frontend/src/lib/api-client.ts b/frontend/src/lib/api-client.ts index 6eead2a..ff80d63 100644 --- a/frontend/src/lib/api-client.ts +++ b/frontend/src/lib/api-client.ts @@ -43,6 +43,25 @@ export interface PasswordChangePayload { new_password: string; } +export interface LibraryFolderDTO { + id: number; + name: string; + color?: string | null; + paper_count?: number; + created_at: string; + updated_at: string; +} + +export interface LibraryFolderListResponse { + folders: LibraryFolderDTO[]; + unfiled_count: number; +} + +export interface LibraryFolderPayload { + name: string; + color?: string | null; +} + export interface UploadedPaperDTO { id: number; original_filename: string; @@ -50,6 +69,9 @@ export interface UploadedPaperDTO { content_type: string; file_size: number; file_hash?: string | null; + folder_id?: number | null; + metadata_json?: Record | null; + folder?: LibraryFolderDTO | null; uploaded_at: string; } @@ -61,6 +83,10 @@ export interface UploadedPaperListResponse { total_pages: number; } +export interface LibraryFolderAssignmentPayload { + folder_id?: number | null; +} + export type AnnotationType = "highlight" | "note"; export interface AnnotationRect { @@ -263,9 +289,16 @@ export async function deleteAccount(token: string): Promise<{ message: string }> return response.json(); } -export async function uploadPaper(token: string, file: File): Promise { +export async function uploadPaper( + token: string, + file: File, + options?: { folderId?: number | null }, +): Promise { const formData = new FormData(); formData.append("file", file); + if (options?.folderId && options.folderId > 0) { + formData.append("folder_id", options.folderId.toString()); + } const response = await fetch(`${BACKEND_URL}/api/v1/papers/upload`, { method: "POST", @@ -285,7 +318,7 @@ export async function uploadPaper(token: string, file: File): Promise { const searchParams = new URLSearchParams(); if (params?.page && params.page > 0) { @@ -294,6 +327,9 @@ export async function fetchUploadedPapers( if (params?.pageSize && params.pageSize > 0) { searchParams.set("page_size", params.pageSize.toString()); } + if (typeof params?.folderId === "number") { + searchParams.set("folder_id", params.folderId.toString()); + } const queryString = searchParams.toString(); const url = queryString.length > 0 @@ -315,6 +351,131 @@ export async function fetchUploadedPapers( return response.json(); } +export async function fetchUploadedPaper(token: string, paperId: number): Promise { + const response = await fetch(`${BACKEND_URL}/api/v1/papers/uploads/${paperId}`, { + headers: { + Authorization: `Bearer ${token}`, + }, + cache: "no-store", + }); + + if (!response.ok) { + const detail = (await response.json().catch(() => ({}))) as { detail?: string }; + throw new Error(detail.detail ?? "获取文档详情失败"); + } + + return response.json(); +} + +export async function fetchLibraryFolders(token: string): Promise { + const response = await fetch(`${BACKEND_URL}/api/v1/library/folders`, { + headers: { + Authorization: `Bearer ${token}`, + }, + cache: "no-store", + }); + + if (!response.ok) { + const detail = (await response.json().catch(() => ({}))) as { detail?: string }; + throw new Error(detail.detail ?? "获取文库文件夹失败"); + } + + return response.json(); +} + +export async function createLibraryFolder( + token: string, + payload: LibraryFolderPayload, +): Promise { + const response = await fetch(`${BACKEND_URL}/api/v1/library/folders`, { + method: "POST", + headers: { + Authorization: `Bearer ${token}`, + "Content-Type": "application/json", + }, + body: JSON.stringify(payload), + }); + + if (!response.ok) { + const detail = (await response.json().catch(() => ({}))) as { detail?: string }; + throw new Error(detail.detail ?? "创建文件夹失败"); + } + + return response.json(); +} + +export async function updateLibraryFolder( + token: string, + folderId: number, + payload: LibraryFolderPayload, +): Promise { + const response = await fetch(`${BACKEND_URL}/api/v1/library/folders/${folderId}`, { + method: "PATCH", + headers: { + Authorization: `Bearer ${token}`, + "Content-Type": "application/json", + }, + body: JSON.stringify(payload), + }); + + if (!response.ok) { + const detail = (await response.json().catch(() => ({}))) as { detail?: string }; + throw new Error(detail.detail ?? "更新文件夹失败"); + } + + return response.json(); +} + +export async function deleteLibraryFolder(token: string, folderId: number): Promise { + const response = await fetch(`${BACKEND_URL}/api/v1/library/folders/${folderId}`, { + method: "DELETE", + headers: { + Authorization: `Bearer ${token}`, + }, + }); + + if (!response.ok) { + const detail = (await response.json().catch(() => ({}))) as { detail?: string }; + throw new Error(detail.detail ?? "删除文件夹失败"); + } +} + +export async function updateUploadedPaperFolder( + token: string, + paperId: number, + folderId: number | null, +): Promise { + const response = await fetch(`${BACKEND_URL}/api/v1/library/uploads/${paperId}/folder`, { + method: "PATCH", + headers: { + Authorization: `Bearer ${token}`, + "Content-Type": "application/json", + }, + body: JSON.stringify({ folder_id: folderId }), + }); + + if (!response.ok) { + const detail = (await response.json().catch(() => ({}))) as { detail?: string }; + throw new Error(detail.detail ?? "更新文献所在文件夹失败"); + } + + return response.json(); +} + +export async function deleteUploadedPaper(token: string, paperId: number): Promise { + const response = await fetch(`${BACKEND_URL}/api/v1/library/uploads/${paperId}`, { + method: "DELETE", + headers: { + Authorization: `Bearer ${token}`, + }, + }); + + if (!response.ok) { + const detail = (await response.json().catch(() => ({}))) as { detail?: string }; + throw new Error(detail.detail ?? "删除文献失败"); + } +} + export async function createPaperAnnotation( token: string, paperId: number, diff --git a/frontend/src/lib/api/academic.ts b/frontend/src/lib/api/academic.ts index 32d397e..e48a784 100644 --- a/frontend/src/lib/api/academic.ts +++ b/frontend/src/lib/api/academic.ts @@ -1,8 +1,9 @@ -import { BACKEND_URL } from '../config'; import { ChatMessage, IntelligentSearchResponse, Paper } from '../types/academic'; +import { buildBackendUrl } from '@/lib/url'; -const API_BASE = BACKEND_URL ?? 'http://localhost:8000'; -const API_V1_BASE = `${API_BASE}/api/v1`; +const API_PREFIX = "/api/v1"; + +const buildApiUrl = (path: string) => buildBackendUrl(`${API_PREFIX}${path}`); export type SearchMode = 'standard' | 'ai'; @@ -31,10 +32,11 @@ export interface SearchResponse { export async function searchPapers(params: SearchParams): Promise { try { - console.log('Fetching from:', `${API_V1_BASE}/academic/search`); + const url = buildApiUrl("/academic/search"); + console.log('Fetching from:', url); console.log('Request params:', params); - const response = await fetch(`${API_V1_BASE}/academic/search`, { + const response = await fetch(url, { method: 'POST', headers: { 'Content-Type': 'application/json', @@ -72,6 +74,7 @@ export async function intelligentSearch( token?: string | null, ): Promise { try { + const url = buildApiUrl("/academic/intelligent-search"); const headers: Record = { 'Content-Type': 'application/json', }; @@ -80,19 +83,31 @@ export async function intelligentSearch( headers.Authorization = `Bearer ${token}`; } - const response = await fetch(`${API_V1_BASE}/academic/intelligent-search`, { + const response = await fetch(url, { method: 'POST', headers, body: JSON.stringify(params), }); if (!response.ok) { - const errorText = await response.text(); - console.error('Intelligent search failed:', errorText); - throw new Error(`智能搜索失败: ${response.status} ${response.statusText}`); + let errorPayload: unknown = null; + let errorText = ""; + try { + errorPayload = await response.json(); + } catch { + errorText = await response.text(); + } + console.error("Intelligent search failed:", errorPayload || errorText); + + const detail: string | undefined = + typeof errorPayload === "object" && errorPayload !== null && "detail" in errorPayload + ? (errorPayload as { detail?: string }).detail + : undefined; + + throw new Error(detail ?? `智能搜索失败: ${response.status} ${response.statusText}`); } - return await response.json(); + return await response.json(); } catch (error) { console.error('智能搜索请求失败:', error); throw error; diff --git a/frontend/src/lib/url.ts b/frontend/src/lib/url.ts new file mode 100644 index 0000000..e3b81f0 --- /dev/null +++ b/frontend/src/lib/url.ts @@ -0,0 +1,18 @@ +import { BACKEND_URL } from "@/lib/config"; + +export function buildBackendUrl(path: string): string { + const normalizedPath = path.startsWith("/") ? path : `/${path}`; + const base = (BACKEND_URL ?? "").trim().replace(/\/$/, ""); + + if (base.length > 0) { + return `${base}${normalizedPath}`; + } + + if (typeof window !== "undefined") { + const { protocol, hostname, port } = window.location; + const origin = `${protocol}//${hostname}${port ? `:${port}` : ""}`; + return `${origin}${normalizedPath}`; + } + + return normalizedPath; +}