Adding Custom Tables to the Database

User can also add custom SQLite tables to the analysis. The process is rather simple and requires the following steps:

Step 1: Define custom table in custom/custom_tables/tables.py

In the custom/custom_tables/tables.py file, enter a custom SQLite table structure following the SQLAlchemy package structure. Below, we provide an example of CustomTable class with the following columns:

• id (Integer): primary key id for entry.
• value1 (Integer): column with an integer value.
• value2 (Integer): column with an integer value.
• value3 (Integer): column with an integer value.
• datetime (DateTime): data and time for the entry.
• analysis_id (Integer, ForeignKey): id from Analysis table.

Here is the code for the table:

class CustomTable(Base):
    __tablename__ = 'custom_table'
    id = Column(Integer, primary_key=True)
    value1 = Column(Integer)
    value2 = Column(Integer)
    value3 = Column(Integer)
    datetime = Column(DateTime, default=datetime.datetime.utcnow)
    analysis_id = Column(Integer, ForeignKey(Analysis.id))
    analysis = relationship(Analysis)

    def __repr__(self):
        return "<CustomTable(id=%d)>" % self.id

The full code for custom/custom_tables/tables.py will look like this:

import datetime
from sqlalchemy import *
from sqlalchemy.ext.declarative import declarative_base
from sqlalchemy.orm import relationship
from aisy.sca_tables import Analysis
from sqlalchemy import create_engine
from sqlalchemy.orm import sessionmaker

Base = declarative_base()


def base():
    return Base


def start_custom_tables(database_name):
    engine = create_engine('sqlite:///{}'.format(database_name), echo=False)
    base().metadata.create_all(engine)


def start_custom_tables_session(database_name):
    engine = create_engine('sqlite:///{}'.format(database_name), echo=False)
    return sessionmaker(bind=engine)()


class CustomTable(Base):
    __tablename__ = 'custom_table'
    id = Column(Integer, primary_key=True)
    value1 = Column(Integer)
    value2 = Column(Integer)
    value3 = Column(Integer)
    datetime = Column(DateTime, default=datetime.datetime.utcnow)
    analysis_id = Column(Integer, ForeignKey(Analysis.id))
    analysis = relationship(Analysis)

    def __repr__(self):
        return "<CustomTable(id=%d)>" % self.id

Step 2: Inserting entries to your custom table

First, in your script file (located at scripts/), create the Aisy() object and set dataset folder, dataset name and database name and run the analysis as usual (look at the imports):

import aisy_sca
from app import *
from custom.custom_models.neural_networks import *
from custom.custom_tables.tables import *

aisy = aisy_sca.Aisy()
aisy.set_resources_root_folder(resources_root_folder)
aisy.set_database_root_folder(databases_root_folder)
aisy.set_datasets_root_folder(datasets_root_folder)
aisy.set_database_name("database_ascad.sqlite")
aisy.set_dataset(datasets_dict["ascad-variable.h5"])
aisy.set_aes_leakage_model(leakage_model="HW", byte=2)
aisy.set_batch_size(400)
aisy.set_epochs(20)
aisy.set_neural_network(mlp)

aisy.run()

Next, start the custom table session:

start_custom_tables(databases_root_folder + "database_ascad.sqlite")
session = start_custom_tables_session(databases_root_folder + "database_ascad.sqlite")

Then, insert your entries to your CustomTable as soon as you define the values to be inserted:

new_insert = CustomTable(value1=10, value2=20, value3=30, analysis_id=aisy.settings["analysis_id"])
session.add(new_insert)
session.commit()

A full script example will look like this:

import aisy_sca
from app import *
from custom.custom_models.neural_networks import *
from custom.custom_tables.tables import *

aisy = aisy_sca.Aisy()
aisy.set_resources_root_folder(resources_root_folder)
aisy.set_database_root_folder(databases_root_folder)
aisy.set_datasets_root_folder(datasets_root_folder)
aisy.set_database_name("database_ascad.sqlite")
aisy.set_dataset(datasets_dict["ascad-variable.h5"])
aisy.set_aes_leakage_model(leakage_model="HW", byte=2)
aisy.set_batch_size(400)
aisy.set_epochs(20)
aisy.set_neural_network(mlp)

aisy.run()

start_custom_tables(databases_root_folder + "database_ascad.sqlite")
session = start_custom_tables_session(databases_root_folder + "database_ascad.sqlite")

new_insert = CustomTable(value1=10, value2=20, value3=30, analysis_id=aisy.get_analysis_id())
session.add(new_insert)
session.commit()