feat(knowledge): 任务队列、独立 worker 与 pgvector 向量索引

知识处理改为数据库任务队列，独立 worker 做切块与向量化，失败可重试。

backend/app/knowledge/pdf_limits.py

@@ -0,0 +1,61 @@
+"""
+Copyright 2024-2026 ChatterMate
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+
+PDFs we will not index, and why.
+
+Bytes don't bound the work: a small file can hold thousands of pages, and a
+PDF is read page by page - every page's text embedded, and a scanned one
+OCR'd at a second or more a page. So the page count is limited too, checked
+before any of it is read.
+"""
+
+from typing import BinaryIO, Optional, Union
+
+from pypdf import PdfReader
+
+from app.core.config import settings
+from app.core.logger import get_logger
+
+logger = get_logger(__name__)
+
+
+class PdfRefused(ValueError):
+    """A PDF over one of our limits. The message is for the user, as it is
+    shown on the failed source."""
+
+
+class PdfTooManyPages(PdfRefused):
+    pass
+
+
+def count_pages(pdf: Union[str, BinaryIO]) -> Optional[int]:
+    """Pages in `pdf` (a path or a file object), or None when it cannot be
+    counted (encrypted, damaged): the reader then fails it its own way."""
+    try:
+        return len(PdfReader(pdf).pages)
+    except Exception as e:
+        logger.warning(f"Could not count the pages of a PDF: {e}")
+        return None
+
+
+def check_page_count(pdf: Union[str, BinaryIO], max_pages: Optional[int] = None) -> None:
+    """Raise PdfTooManyPages when `pdf` has more pages than we index."""
+    limit = max_pages if max_pages is not None else settings.KB_PDF_MAX_PAGES
+    pages = count_pages(pdf)
+    if pages is not None and pages > limit:
+        raise PdfTooManyPages(
+            f"This PDF has {pages} pages; the limit is {limit}. Upload the part "
+            "you need, or split it into smaller files."
+        )