Add MinerU document ingestion support

Normalize uploaded documents into structured markdown, add clearer parser
errors for missing dependencies, and cover the ingestion flow with
backend tests. This also replaces deprecated UTC timestamp helpers in
the touched backend paths so the knowledge pipeline stays warning-free.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend/app/models/document.py

@@ -1,4 +1,4 @@
-from sqlalchemy import Column, String, Integer, Text, ForeignKey, Boolean
+from sqlalchemy import Column, String, Integer, Text, ForeignKey, Boolean, DateTime
 from sqlalchemy.orm import relationship
 from app.models.base import BaseModel
 
@@ -16,6 +16,13 @@ class Document(BaseModel):
     summary = Column(Text, nullable=True)
     chunk_count = Column(Integer, default=0)
     is_indexed = Column(Boolean, default=False)
+    ingestion_status = Column(String(50), default="uploaded", nullable=False)
+    ingestion_error = Column(Text, nullable=True)
+    indexed_at = Column(DateTime, nullable=True)
+    parser_version = Column(String(50), nullable=True)
+    index_version = Column(String(50), nullable=True)
+    normalized_content = Column(Text, nullable=True)
+    normalized_format = Column(String(50), nullable=True)
 
     chunks = relationship("DocumentChunk", back_populates="document", cascade="all, delete-orphan")