[RSPEED-416] Migrate JSON history cache to database solution

.packit.yaml

@@ -13,8 +13,8 @@ jobs:
     project: command-line-assistant
     targets:
       - epel-9-x86_64
-      - epel-10-x86_64
       - epel-9-aarch64
+      - epel-10-x86_64
       - epel-10-aarch64
     actions:
       # do not get the version from a tag (git describe) but from the spec file
@@ -29,8 +29,8 @@ jobs:
     project: command-line-assistant
     targets:
       - epel-9-x86_64
-      - epel-10-x86_64
       - epel-9-aarch64
+      - epel-10-x86_64
       - epel-10-aarch64
     actions:
       # bump spec so we get release starting with 2 and hence all the default branch builds will
@@ -49,8 +49,8 @@ jobs:
     project: command-line-assistant
     targets:
       - epel-9-x86_64
-      - epel-10-x86_64
       - epel-9-aarch64
+      - epel-10-x86_64
       - epel-10-aarch64
     actions:
       # bump spec so we get release starting with 2 and hence all the default branch builds will

.pre-commit-config.yaml

@@ -52,6 +52,7 @@ repos:
         - responses
         - tomli; python_version<"3.11"
         - setuptools
+        - sqlalchemy
 
   - repo: https://github.com/gitleaks/gitleaks
     rev: v8.22.1

command_line_assistant/config/schemas.py

@@ -2,7 +2,67 @@
 
 import dataclasses
 from pathlib import Path
-from typing import Union
+from typing import Optional, Union
+
+
+@dataclasses.dataclass
+class DatabaseSchema:
+    """This class represents the [history.database] section of our config.toml file.
+
+    Attributes:
+        connection (str): The connection string.
+    """
+
+    type: str = "sqlite"  # 'sqlite', 'mysql', 'postgresql', etc.
+    host: Optional[str] = None
+    database: Optional[str] = None
+    port: Optional[int] = None  # Optional for SQLite as it doesn't require host or port
+    user: Optional[str] = None  # Optional for SQLite
+    password: Optional[str] = None  # Optional for SQLite
+    connection_string: Optional[Union[str, Path]] = (
+        None  # Some databases like SQLite can use a file path
+    )
+
+    def __post_init__(self):
+        """Post initialization method to normalize values"""
+        # If the database type is not a supported one, we can just skip it.
+        allowed_databases = ("mysql", "sqlite", "postgresql")
+        if self.type not in allowed_databases:
+            raise ValueError(
+                f"The database type must be one of {','.join(allowed_databases)}, not {self.type}"
+            )
+
+        if self.connection_string:
+            self.connection_string = Path(self.connection_string).expanduser()
+
+        # Post-initialization to set default values for specific db types
+        if self.type == "sqlite" and not self.connection_string:
+            self.connection_string = f"sqlite://{self.database}"
+        elif self.type == "mysql" and not self.port:
+            self.port = 3306  # Default MySQL port
+        elif self.type == "postgresql" and not self.port:
+            self.port = 5432  # Default PostgreSQL port
+
+    def get_connection_url(self) -> str:
+        """
+        Constructs and returns the connection URL or string for the respective database.
+
+        Raises:
+            ValueError: In case the type is not recognized
+
+        Returns:
+            str: The URL formatted connection
+        """
+        connection_urls = {
+            "sqlite": f"sqlite:///{self.connection_string}",
+            "mysql": f"mysql://{self.user}:{self.password}@{self.host}:{self.port}/{self.database}",
+            "postgresql": f"postgresql://{self.user}:{self.password}@{self.host}:{self.port}/{self.database}",
+        }
+
+        if self.type not in connection_urls:
+            raise ValueError(f"Unsupported database type: {self.type}")
+
+        return connection_urls[self.type]
 
 
 @dataclasses.dataclass
@@ -83,13 +143,15 @@
     """
 
     enabled: bool = True
-    file: Union[str, Path] = Path(  # type: ignore
-        "/var/lib/command-line-assistant/history.json"
-    )
+    database: DatabaseSchema = dataclasses.field(default_factory=DatabaseSchema)
 
     def __post_init__(self):
         """Post initialization method to normalize values"""
-        self.file: Path = Path(self.file).expanduser()
+
+        # # Database may be present in the config.toml. If it is not, we odn't do
+        # # anything and go with defaults.
+        if isinstance(self.database, dict):
+            self.database = DatabaseSchema(**self.database)
 
 
 @dataclasses.dataclass
@@ -108,8 +170,8 @@
 
     def __post_init__(self) -> None:
         """Post initialization method to normalize values"""
-        self.cert_file = Path(self.cert_file)
-        self.key_file = Path(self.key_file)
+        self.cert_file = Path(self.cert_file).expanduser()
+        self.key_file = Path(self.key_file).expanduser()
 
 
 @dataclasses.dataclass
@@ -122,7 +184,7 @@
     """
 
     endpoint: str = "http://0.0.0.0:8080"
-    auth: Union[dict, AuthSchema] = dataclasses.field(default_factory=AuthSchema)
+    auth: AuthSchema = dataclasses.field(default_factory=AuthSchema)
 
     def __post_init__(self):
         """Post initialization method to normalize values"""

command_line_assistant/daemon/database/__init__.py

@@ -0,0 +1 @@
+"""Main module to hold anything related to databases."""

command_line_assistant/daemon/database/manager.py

@@ -0,0 +1,170 @@
+"""Database module to handle SQLAlchemy connections and interactions."""
+
+import logging
+import uuid
+from contextlib import contextmanager
+from typing import Generator, Optional, TypeVar
+
+from sqlalchemy import create_engine
+from sqlalchemy.engine import Engine
+from sqlalchemy.orm import Session, sessionmaker
+from sqlalchemy.pool import StaticPool
+
+from command_line_assistant.config import Config
+from command_line_assistant.daemon.database.models.base import BaseModel
+
+logger = logging.getLogger(__name__)
+
+
+# Type variable for ORM models
+T = TypeVar("T")
+
+
+class DatabaseError(Exception):
+    """Base exception for database errors."""
+
+
+class ConnectionError(DatabaseError):
+    """Exception raised when connection fails."""
+
+
+class QueryError(DatabaseError):
+    """Exception raised when query fails."""
+
+
+class DatabaseManager:
+    """Class to handle database operations using SQLAlchemy."""
+
+    def __init__(self, config: Config, echo: bool = False) -> None:
+        """Initialize database connection.
+
+        Args:
+            database (Path): Path to the SQLite database file
+            echo (bool): Enable SQL query logging if True
+        """
+        self._config = config
+        self._engine: Engine = self._create_engine(echo)
+        self._session_factory = sessionmaker(bind=self._engine)
+
+    def _create_engine(self, echo: bool) -> Engine:
+        """Create SQLAlchemy engine with proper settings.
+
+        Args:
+            echo (bool): Enable SQL query logging if True
+
+        Returns:
+            Engine: Configured SQLAlchemy engine
+
+        Raises:
+            ConnectionError: When invalid database settings are provided
+        """
+        try:
+            connection_url = self._config.history.database.get_connection_url()
+
+            # SQLite-specific settings
+            connect_args = {}
+            if self._config.history.database.type == "sqlite":
+                connect_args["check_same_thread"] = False
+                return create_engine(
+                    connection_url,
+                    echo=echo,
+                    poolclass=StaticPool,
+                    connect_args=connect_args,
+                )
+
+            # For other databases, use standard pooling
+            return create_engine(
+                connection_url,
+                echo=echo,
+                pool_pre_ping=True,
+                pool_size=5,
+                max_overflow=10,
+            )
+        except Exception as e:
+            logger.error("Failed to create database engine: %s", e)
+            raise ConnectionError(f"Could not create database engine: {e}") from e
+
+    def connect(self) -> None:
+        """Create database tables if they don't exist."""
+        try:
+            BaseModel.metadata.create_all(self._engine)
+        except Exception as e:
+            logger.error("Failed to create database tables: %s", e)
+            raise ConnectionError(f"Could not create tables: {e}") from e
+
+    @contextmanager
+    def session(self) -> Generator[Session, None, None]:
+        """Create a contextual database session.
+
+        Yields:
+            Session: SQLAlchemy session object
+
+        Raises:
+            QueryError: If session operations fail
+        """
+        session = self._session_factory()
+        try:
+            yield session
+            session.commit()
+        except Exception as e:
+            session.rollback()
+            logger.error("Database session error: %s", e)
+            raise QueryError(f"Session error: {e}") from e
+        finally:
+            session.close()
+
+    def add(self, instance: T) -> None:
+        """Add an instance to the database.
+
+        Args:
+            instance (T): SQLAlchemy model instance to add
+
+        Raises:
+            QueryError: If adding fails
+        """
+        try:
+            with self.session() as session:
+                session.add(instance)
+                session.flush()
+        except Exception as e:
+            logger.error("Failed to add instance: %s", e)
+            raise QueryError(f"Failed to add instance: {e}") from e
+
+    def query(self, model: type[T]) -> list[T]:
+        """Query all instances of a model.
+
+        Args:
+            model (type[T]): SQLAlchemy model class to query
+
+        Returns:
+            list[T]: List of model instances
+
+        Raises:
+            QueryError: If query fails
+        """
+        try:
+            with self.session() as session:
+                return session.query(model).all()
+        except Exception as e:
+            logger.error("Failed to query instances: %s", e)
+            raise QueryError(f"Failed to query instances: {e}") from e
+
+    def get(self, model: type[T], id: uuid.UUID) -> Optional[T]:
+        """Get a single instance by ID.
+
+        Args:
+            model (type[T]): SQLAlchemy model class
+            id (uuid.UUID): Instance ID to get
+
+        Returns:
+            Optional[T]: Model instance if found, None otherwise
+
+        Raises:
+            QueryError: If query fails
+        """
+        try:
+            with self.session() as session:
+                return session.query(model).get(id)
+        except Exception as e:
+            logger.error("Failed to get instance: %s", e)
+            raise QueryError(f"Failed to get instance: {e}") from e

command_line_assistant/daemon/database/models/__init__.py

@@ -0,0 +1 @@
+"""Main module to hold the database models"""

command_line_assistant/daemon/database/models/base.py

@@ -0,0 +1,6 @@
+"""Base module to hold the declarative base for sqlalchemy models"""
+
+from sqlalchemy.ext.declarative import declarative_base
+
+#: The declarative base model for SQLAlchemy models
+BaseModel = declarative_base()

command_line_assistant/daemon/database/models/history.py

@@ -0,0 +1,43 @@
+"""Module containing SQLAlchemy models for the daemon."""
+
+import uuid
+from datetime import datetime
+
+from sqlalchemy import Column, DateTime, ForeignKey, Integer, String
+from sqlalchemy.dialects.postgresql import UUID
+from sqlalchemy.orm import relationship
+
+from command_line_assistant.daemon.database.models.base import BaseModel
+
+
+class HistoryModel(BaseModel):
+    """SQLAlchemy model for history table that maps to HistoryEntry dataclass."""
+
+    __tablename__ = "history"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    timestamp = Column(DateTime, default=datetime.utcnow())
+    deleted_at = Column(DateTime, nullable=True)
+
+    # Relationships
+    interaction_id = Column(
+        UUID(as_uuid=True), ForeignKey("interaction.id"), nullable=False
+    )
+    interaction = relationship("InteractionModel", backref="history")
+
+
+class InteractionModel(BaseModel):
+    """SQLAlchemy model for interaction table."""
+
+    __tablename__ = "interaction"
+
+    id = Column(UUID(as_uuid=True), primary_key=True, default=uuid.uuid4)
+    query_text = Column(String)
+    query_role = Column(String, default="user")
+    response_text = Column(String)
+    response_role = Column(String, default="assistant")
+    response_tokens = Column(Integer, default=0)
+    session_id = Column(UUID(as_uuid=True), nullable=False, default=uuid.uuid4)
+    os_distribution = Column(String, default="RHEL")
+    os_version = Column(String, nullable=False)
+    os_arch = Column(String, nullable=False)