nopaque/app/models.py

from datetime import datetime
from flask import current_app
from flask_login import UserMixin, AnonymousUserMixin
from itsdangerous import BadSignature, TimedJSONWebSignatureSerializer
from time import sleep
from werkzeug.security import generate_password_hash, check_password_hash
from werkzeug.utils import secure_filename
from . import db, logger, login_manager
import os
import shutil
import xml.etree.ElementTree as ET


class Permission:
    """
    Defines User permissions as integers by the power of 2. User permission
    can be evaluated using the bitwise operator &. 3 equals to CREATE_JOB and
    DELETE_JOB and so on.
    """
    CREATE_JOB = 1
    DELETE_JOB = 2
    # WRITE = 4
    # MODERATE = 8
    ADMIN = 16


class Role(db.Model):
    """
    Model for the different roles Users can have. Is a one-to-many
    relationship. A Role can be associated with many User rows.
    """
    __tablename__ = 'roles'
    # Primary key
    id = db.Column(db.Integer, primary_key=True)
    default = db.Column(db.Boolean, default=False, index=True)
    name = db.Column(db.String(64), unique=True)
    permissions = db.Column(db.Integer)
    # Relationships
    users = db.relationship('User', backref='role', lazy='dynamic')

    def __init__(self, **kwargs):
        super(Role, self).__init__(**kwargs)
        if self.permissions is None:
            self.permissions = 0

    def __repr__(self):
        """
        String representation of the Role. For human readability.
        """
        return '<Role %r>' % self.name

    def add_permission(self, perm):
        """
        Add new permission to Role. Input is a Permission.
        """
        if not self.has_permission(perm):
            self.permissions += perm

    def remove_permission(self, perm):
        """
        Removes permission from a Role. Input a Permission.
        """
        if self.has_permission(perm):
            self.permissions -= perm

    def reset_permissions(self):
        """
        Resets permissions to zero. Zero equals no permissions at all.
        """
        self.permissions = 0

    def has_permission(self, perm):
        """
        Checks if a Role has a specific Permission. Does this with the bitwise
        operator.
        """
        return self.permissions & perm == perm

    @staticmethod
    def insert_roles():
        """
        Inserts roles into the databes. This has to be executed befor Users are
        added to the database. Otherwiese Users will not have a Role assigned
        to them. Order of the roles dictionary determines the ID of each role.
        User hast the ID 1 and Administrator has the ID 2.
        """
        roles = {'User': [Permission.CREATE_JOB],
                 'Administrator': [Permission.ADMIN, Permission.CREATE_JOB,
                                   Permission.DELETE_JOB]}
        default_role = 'User'
        for r in roles:
            role = Role.query.filter_by(name=r).first()
            if role is None:
                role = Role(name=r)
            role.reset_permissions()
            for perm in roles[r]:
                role.add_permission(perm)
            role.default = (role.name == default_role)
            db.session.add(role)
        db.session.commit()


class User(UserMixin, db.Model):
    """
    Model for Users that are registered to Opaque.
    """
    __tablename__ = 'users'
    # Primary key
    id = db.Column(db.Integer, primary_key=True)
    confirmed = db.Column(db.Boolean, default=False)
    email = db.Column(db.String(254), unique=True, index=True)
    password_hash = db.Column(db.String(128))
    registration_date = db.Column(db.DateTime(), default=datetime.utcnow)
    role_id = db.Column(db.Integer, db.ForeignKey('roles.id'))
    username = db.Column(db.String(64), unique=True, index=True)
    # Relationships
    corpora = db.relationship('Corpus', backref='creator', lazy='dynamic',
                              cascade='save-update, merge, delete')
    jobs = db.relationship('Job', backref='creator', lazy='dynamic',
                           cascade='save-update, merge, delete')
    is_dark = db.Column(db.Boolean, default=False)

    def __repr__(self):
        """
        String representation of the User. For human readability.
        """
        return '<User %r>' % self.username

    def __init__(self, **kwargs):
        super(User, self).__init__(**kwargs)
        if self.role is None:
            if self.email == current_app.config['NOPAQUE_ADMIN']:
                self.role = Role.query.filter_by(name='Administrator').first()
            if self.role is None:
                self.role = Role.query.filter_by(default=True).first()

    def generate_confirmation_token(self, expiration=3600):
        """
        Generates a confirmation token for user confirmation via email.
        """
        s = TimedJSONWebSignatureSerializer(current_app.config['SECRET_KEY'],
                                            expiration)
        return s.dumps({'confirm': self.id}).decode('utf-8')

    def generate_reset_token(self, expiration=3600):
        """
        Generates a reset token for password reset via email.
        """
        s = TimedJSONWebSignatureSerializer(current_app.config['SECRET_KEY'],
                                            expiration)
        return s.dumps({'reset': self.id}).decode('utf-8')

    def confirm(self, token):
        """
        Confirms User if the given token is valid and not expired.
        """
        s = TimedJSONWebSignatureSerializer(current_app.config['SECRET_KEY'])
        try:
            data = s.loads(token.encode('utf-8'))
        except BadSignature:
            return False
        if data.get('confirm') != self.id:
            return False
        self.confirmed = True
        db.session.add(self)
        return True

    @staticmethod
    def reset_password(token, new_password):
        """
        Resets password for User if the given token is valid and not expired.
        """
        s = TimedJSONWebSignatureSerializer(current_app.config['SECRET_KEY'])
        try:
            data = s.loads(token.encode('utf-8'))
        except BadSignature:
            return False
        user = User.query.get(data.get('reset'))
        if user is None:
            return False
        user.password = new_password
        db.session.add(user)
        return True

    @property
    def password(self):
        raise AttributeError('password is not a readable attribute')

    @password.setter
    def password(self, password):
        self.password_hash = generate_password_hash(password)

    def verify_password(self, password):
        return check_password_hash(self.password_hash, password)

    def can(self, perm):
        """
        Checks if a User with its current role can doe something. Checks if the
        associated role actually has the needed Permission.
        """
        return self.role is not None and self.role.has_permission(perm)

    def is_administrator(self):
        """
        Checks if User has Admin permissions.
        """
        return self.can(Permission.ADMIN)

    def delete(self):
        """
        Delete the user and its corpora and jobs from database and filesystem.
        """
        for job in self.jobs:
            job.delete()
        for corpus in self.corpora:
            corpus.delete()
        path = os.path.join(current_app.config['NOPAQUE_STORAGE'],
                            str(self.id))
        try:
            shutil.rmtree(path)
        except Exception as e:
            logger.warning(e)
            pass
        db.session.delete(self)
        db.session.commit()


class AnonymousUser(AnonymousUserMixin):
    """
    Model replaces the default AnonymousUser.
    """

    def can(self, permissions):
        return False

    def is_administrator(self):
        return False


class JobInput(db.Model):
    """
    Class to define JobInputs.
    """
    __tablename__ = 'job_inputs'
    # Primary key
    id = db.Column(db.Integer, primary_key=True)
    filename = db.Column(db.String(255))
    dir = db.Column(db.String(255))
    job_id = db.Column(db.Integer, db.ForeignKey('jobs.id'))

    def __repr__(self):
        """
        String representation of the JobInput. For human readability.
        """
        return '<JobInput %r>' % self.filename

    def to_dict(self):
        return {'id': self.id,
                'dir': self.dir,
                'filename': self.filename,
                'job_id': self.job_id}


class JobResult(db.Model):
    """
    Class to define JobResults.
    """
    __tablename__ = 'job_results'
    # Primary key
    id = db.Column(db.Integer, primary_key=True)
    filename = db.Column(db.String(255))
    dir = db.Column(db.String(255))
    job_id = db.Column(db.Integer, db.ForeignKey('jobs.id'))

    def __repr__(self):
        """
        String representation of the JobResult. For human readability.
        """
        return '<JobResult %r>' % self.filename

    def to_dict(self):
        return {'id': self.id,
                'dir': self.dir,
                'filename': self.filename,
                'job_id': self.job_id}


class Job(db.Model):
    """
    Class to define Jobs.
    """
    __tablename__ = 'jobs'
    # Primary key
    id = db.Column(db.Integer, primary_key=True)
    creation_date = db.Column(db.DateTime(), default=datetime.utcnow)
    description = db.Column(db.String(255))
    end_date = db.Column(db.DateTime())
    mem_mb = db.Column(db.Integer)
    n_cores = db.Column(db.Integer)
    secure_filename = db.Column(db.String(32))
    service = db.Column(db.String(64))
    '''
    ' Service specific arguments as string list.
    ' Example: ["-l eng", "--keep-intermediates", "--skip-binarization"]
    '''
    service_args = db.Column(db.String(255))
    service_version = db.Column(db.String(16))
    status = db.Column(db.String(16))
    title = db.Column(db.String(32))
    user_id = db.Column(db.Integer, db.ForeignKey('users.id'))
    # Relationships
    inputs = db.relationship('JobInput', backref='job', lazy='dynamic',
                             cascade='save-update, merge, delete')
    results = db.relationship('JobResult', backref='job', lazy='dynamic',
                              cascade='save-update, merge, delete')

    def __repr__(self):
        """
        String representation of the Job. For human readability.
        """
        return '<Job %r>' % self.title

    def create_secure_filename(self):
        """
        Takes the job.title string nad cratesa a secure filename from this.
        """
        self.secure_filename = secure_filename(self.title)

    def delete(self):
        """
        Delete the job and its inputs and outputs from database and filesystem.
        """
        if self.status != 'complete' and self.status != 'failed':
            self.status = 'canceling'
            db.session.commit()
            while self.status != 'canceled':
                # In case the daemon handled a job in any way
                if self.status != 'canceling':
                    self.status = 'canceling'
                    db.session.commit()
                sleep(1)
                db.session.refresh(self)
        path = os.path.join(current_app.config['NOPAQUE_STORAGE'],
                            str(self.user_id), 'jobs', str(self.id))
        try:
            shutil.rmtree(path)
        except Exception as e:
            ''' TODO: Proper exception handling '''
            logger.warning(e)
            pass
        db.session.delete(self)
        db.session.commit()

    def to_dict(self):
        return {'id': self.id,
                'creation_date': self.creation_date.timestamp(),
                'description': self.description,
                'end_date': (self.end_date.timestamp() if self.end_date else
                             None),
                'inputs': [input.to_dict() for input in self.inputs],
                'mem_mb': self.mem_mb,
                'n_cores': self.n_cores,
                'results': [result.to_dict() for result in self.results],
                'service': self.service,
                'service_args': self.service_args,
                'service_version': self.service_version,
                'status': self.status,
                'title': self.title,
                'user_id': self.user_id}


class CorpusFile(db.Model):
    """
    Class to define Files.
    """
    __tablename__ = 'corpus_files'
    # Primary key
    id = db.Column(db.Integer, primary_key=True)
    address = db.Column(db.String(255))
    author = db.Column(db.String(255))
    booktitle = db.Column(db.String(255))
    chapter = db.Column(db.String(255))
    dir = db.Column(db.String(255))
    editor = db.Column(db.String(255))
    filename = db.Column(db.String(255))
    institution = db.Column(db.String(255))
    journal = db.Column(db.String(255))
    pages = db.Column(db.String(255))
    publisher = db.Column(db.String(255))
    publishing_year = db.Column(db.Integer)
    school = db.Column(db.String(255))
    title = db.Column(db.String(255))
    corpus_id = db.Column(db.Integer, db.ForeignKey('corpora.id'))

    def delete(self):
        path = os.path.join(current_app.config['NOPAQUE_STORAGE'],
                            self.dir, self.filename)
        try:
            os.remove(path)
        except Exception as e:
            ''' TODO: Proper exception handling '''
            logger.warning(e)
            pass
        self.corpus.status = 'unprepared'
        db.session.delete(self)
        db.session.commit()

    def insert_metadata(self):
        file = os.path.join(current_app.config['NOPAQUE_STORAGE'],
                            self.dir, self.filename)
        element_tree = ET.parse(file)
        text_node = element_tree.find('text')
        text_node.set('address', self.address if self.address else "NULL")
        text_node.set('author', self.author)
        text_node.set('booktitle', self.booktitle if self.booktitle else "NULL")
        text_node.set('chapter', self.chapter if self.chapter else "NULL")
        text_node.set('editor', self.editor if self.editor else "NULL")
        text_node.set('institution', self.institution if self.institution else "NULL")
        text_node.set('journal', self.journal if self.journal else "NULL")
        text_node.set('pages', self.pages if self.pages else "NULL")
        text_node.set('publisher', self.publisher if self.publisher else "NULL")
        text_node.set('publishing_year', str(self.publishing_year))
        text_node.set('school', self.school if self.school else "NULL")
        text_node.set('title', self.title)
        element_tree.write(file)
        self.corpus.status = 'unprepared'
        db.session.commit()

    def to_dict(self):
        return {'id': self.id,
                'address': self.address,
                'author': self.author,
                'booktitle': self.booktitle,
                'chapter': self.chapter,
                'dir': self.dir,
                'editor': self.editor,
                'filename': self.filename,
                'institution': self.institution,
                'journal': self.journal,
                'pages': self.pages,
                'publisher': self.publisher,
                'publishing_year': self.publishing_year,
                'school': self.school,
                'title': self.title,
                'corpus_id': self.corpus_id}


class Corpus(db.Model):
    """
    Class to define a corpus.
    """
    __tablename__ = 'corpora'
    # Primary key
    id = db.Column(db.Integer, primary_key=True)
    creation_date = db.Column(db.DateTime(), default=datetime.utcnow)
    description = db.Column(db.String(255))
    status = db.Column(db.String(16))
    title = db.Column(db.String(32))
    user_id = db.Column(db.Integer, db.ForeignKey('users.id'))
    analysis_container_ip = db.Column(db.String(16))
    analysis_container_name = db.Column(db.String(32))
    # Relationships
    files = db.relationship('CorpusFile', backref='corpus', lazy='dynamic',
                            cascade='save-update, merge, delete')

    def __repr__(self):
        """
        String representation of the corpus. For human readability.
        """
        return '<Corpus %r>' % self.title

    def to_dict(self):
        return {'id': self.id,
                'creation_date': self.creation_date.timestamp(),
                'description': self.description,
                'files': [file.to_dict() for file in self.files],
                'status': self.status,
                'title': self.title,
                'user_id': self.user_id}

    def delete(self):
        for corpus_file in self.files:
            corpus_file.delete()
        path = os.path.join(current_app.config['NOPAQUE_STORAGE'],
                            str(self.user_id), 'corpora', str(self.id))
        try:
            shutil.rmtree(path)
        except Exception as e:
            ''' TODO: Proper exception handling '''
            logger.warning(e)
            pass
        db.session.delete(self)
        db.session.commit()

    def prepare(self):
        pass


'''
' Flask-Login is told to use the application’s custom anonymous user by setting
' its class in the login_manager.anonymous_user attribute.
'''
login_manager.anonymous_user = AnonymousUser


@login_manager.user_loader
def load_user(user_id):
    return User.query.get(int(user_id))