def update_existing_table(config_file, table_name, directory_name, data_file,
                          alias, silent, integrity_cancel):
    """

    :param alias:
    :param silent:
    :param config_file:
    :param table_name:
    :param directory_name:
    :param data_file:
    :return:
    """
    _cfg, _tbl, _al, _sil = config_file, table_name, alias, silent
    config, config_file = ConfigManager.confirm_config_set(config_file)
    if alias != "None":
        table_name = ConfigManager.get_name_by_alias(alias, config)
        if table_name is None:
            print(
                "!! Table does not exist! Run CREATE to add to existing database, or INIT to create in new database !!"
            )
            exit(1)
    if table_name != "None" and table_name not in config.keys():
        print(
            "!! Table does not exist! Run CREATE to add to existing database, or INIT to create in new database !!"
        )
        exit(1)
    cfg = ConfigManager(config, table_name)
    if not silent:
        _update_display_message_prelude(
            config[ConfigKeys.DATABASES][ConfigKeys.db_name],
            config[ConfigKeys.DATABASES][ConfigKeys.rel_work_dir], table_name,
            directory_name, data_file, alias)
    if data_file != "None":
        data_to_add = CountTable(data_file)
    else:
        data_to_add = None
    if directory_name != "None":
        genomic_files_to_add = (_f for _f in os.listdir(directory_name))
    else:
        genomic_files_to_add = ()
    if alias == "None":
        new_attrs = ClassManager.populate_data_to_existing_table(
            table_name, data_to_add, cfg, genomic_files_to_add, directory_name,
            silent, alias)
    else:
        new_attrs = ClassManager.populate_data_to_existing_table(
            table_name, data_to_add, cfg, genomic_files_to_add, directory_name,
            silent)
    ClassManager.write_class(new_attrs, cfg.classes_file)
    if not silent:
        _update_table_display_message_epilogue()
    if not integrity_cancel:
        integrity_check(_cfg, _tbl, _al, _sil)
Example #2
0
def load_table_metadata(config, tbl_name):
    cfg = ConfigManager(config, tbl_name)
    engine = BaseData.get_engine(cfg.db_dir, cfg.db_name + ".db")
    sess = BaseData.get_session_from_engine(engine)
    TableClass = ClassManager.get_class_orm(tbl_name, engine)
    UserClass = type(tbl_name, (Record, ), {})
    # Map to SQL orm
    mapper(UserClass, TableClass)
    # Display queried info for single table and break
    return sess, UserClass, cfg
Example #3
0
def remove_columns_from_table(config_file, table_name, list_file, alias,
                              silent, integrity_cancel):
    """

    :param silent:
    :param config_file:
    :param table_name:
    :param list_file:
    :param alias:
    :return:
    """
    _cfg, _tbl, _sil, _al = config_file, table_name, silent, alias
    config, config_file = ConfigManager.confirm_config_set(config_file)
    if alias != "None":
        table_name = ConfigManager.get_name_by_alias(alias, config)
        assert table_name is not None, TableNameAssertString.TABLE_NOT_FOUND
    if list_file == "None":
        raise ListFileNotProvidedError
    cfg = ConfigManager(config, table_name)
    columns_to_remove = set(
        line.rstrip("\r\n") for line in open(list_file, "r"))
    if not silent:
        _remove_columns_display_message_prelude(
            config[ConfigKeys.DATABASES][ConfigKeys.db_name],
            config[ConfigKeys.DATABASES][ConfigKeys.rel_work_dir], table_name,
            alias, columns_to_remove)

    engine = BaseData.get_engine(cfg.db_dir, cfg.db_name + ".db")
    sess = BaseData.get_session_from_engine(engine)
    TableClass = ClassManager.get_class_orm(table_name, engine)
    update_manager = UpdateManager(cfg, ClassManager.get_class_as_dict(cfg),
                                   sess)
    table_copy_csv = update_manager.create_table_copy(
        datetime.today().strftime("%Y%m%d"), TableClass, silent)
    new_attrs = {
        key: value
        for key, value in ClassManager.correct_dict(
            ClassManager.get_class_as_dict(cfg)).items()
        if key not in columns_to_remove
    }
    UpdatedDBClass, metadata = ClassManager.generate_class(
        cfg.table_name, new_attrs, cfg.db_dir, cfg.db_name, cfg.table_dir)
    ClassManager.write_class(new_attrs, cfg.classes_file)
    cfg.update_config_file(table_name)
    update_manager.delete_old_table_and_populate(engine, TableClass,
                                                 UpdatedDBClass,
                                                 table_copy_csv, table_name,
                                                 sess, silent)
    if not silent:
        _remove_columns_display_message_epilogue()
    if not integrity_cancel:
        integrity_check(_cfg, _tbl, _al, _sil)
Example #4
0
def delete_from_table(config_file, table_name, list_file, alias, silent, integrity_cancel):
    """

    :param silent:
    :param config_file:
    :param table_name:
    :param list_file:
    :param alias:
    :return:
    """
    _cfg, _tbl, _sil, _al = config_file, table_name, silent, alias
    config, config_file = ConfigManager.confirm_config_set(config_file)
    if alias != "None":
        table_name = ConfigManager.get_name_by_alias(alias, config)
        assert table_name is not None, TableNameAssertString.TABLE_NOT_FOUND
    if list_file == "None":
        raise ListFileNotProvidedError
    cfg = ConfigManager(config, table_name)
    ids_to_remove = set(line.rstrip("\r\n") for line in open(list_file, "r"))
    if not silent:
        _delete_records_display_message_prelude(
            config[ConfigKeys.DATABASES][ConfigKeys.db_name],
            config[ConfigKeys.DATABASES][ConfigKeys.rel_work_dir],
            table_name,
            alias,
            ids_to_remove
        )

    engine = BaseData.get_engine(cfg.db_dir, cfg.db_name + ".db")
    sess = BaseData.get_session_from_engine(engine)
    TableClass = ClassManager.get_class_orm(table_name, engine)
    UserClass = type(table_name, (Record,), {})
    mapper(UserClass, TableClass)
    for _id in ids_to_remove:
        print_if_not_silent(silent, " ..Removing record %s" % _id)
        try:
            os.remove(sess.query(UserClass).filter_by(_id=_id).first().full_path())
        except TypeError:
            continue
        sess.delete(sess.query(UserClass).filter_by(_id=_id).first())
    sess.commit()
    if not silent:
        _remove_columns_display_message_epilogue()
    if not integrity_cancel:
        integrity_check(_cfg, _tbl, _al, _sil)
Example #5
0
def integrity_check(config_file, table_name, alias, silent):
    """  Function called from dbdm that checks integrity and issues in project at all levels

    :param config_file:
    :param table_name:
    :param alias:l
    :return:
    """
    config, config_file = ConfigManager.confirm_config_set(config_file)
    if alias != "None":
        table_name = ConfigManager.get_name_by_alias(alias, config)
    if table_name is None or (table_name == "None" and alias == "None"):
        tables_to_search = list(config[ConfigKeys.TABLES_TO_DB].keys())
    else:
        tables_to_search = [
            table_name,
        ]
    py_fixfile_name = "%s.%s.fix" % (datetime.today().strftime("%Y%m%d"),
                                     str(randint(1, 1001)))
    if not silent:
        _integrity_check_display_message_prelude(
            config[ConfigKeys.DATABASES][ConfigKeys.db_name],
            config[ConfigKeys.DATABASES][ConfigKeys.working_dir],
            tables_to_search, py_fixfile_name)
    im = IntegrityManager(config, py_fixfile_name, tables_to_search)
    # TODO: Implement table-level checks
    im.initialize_fix_file()
    im.initial_project_check()
    # im.table_check()
    for table in tables_to_search:
        cfg = ConfigManager(config, table)
        engine = BaseData.get_engine(cfg.db_dir, cfg.db_name + ".db")
        sess = BaseData.get_session_from_engine(engine)
        TableClass = ClassManager.get_class_orm(table, engine)
        UserClass = type(table, (Record, ), {})
        # Map to SQL orm
        mapper(UserClass, TableClass)
        im.record_check(sess, UserClass, table)
    if im.issues_found == 0:
        os.remove(py_fixfile_name)
    if not silent:
        _integrity_check_display_message_epilogue(im.issues_found,
                                                  py_fixfile_name)
    del im
Example #6
0
def remove_table_from_database(config_file, table_name, alias, silent,
                               integrity_cancel):
    """ Function removes a given table from a database

    :param silent:
    :param config_file:
    :param table_name:
    :param alias:
    :return:
    """
    _cfg, _tbl, _sil, _al = config_file, table_name, silent, alias
    config, config_file = ConfigManager.confirm_config_set(config_file)
    if alias != "None":
        table_name = ConfigManager.get_name_by_alias(alias, config)
        assert table_name is not None, TableNameAssertString.TABLE_NOT_FOUND
    if not silent:
        _remove_table_display_message_prelude(
            config[ConfigKeys.DATABASES][ConfigKeys.db_name],
            config[ConfigKeys.DATABASES][ConfigKeys.rel_work_dir], table_name,
            alias)

    cfg = ConfigManager(config, table_name)
    engine = BaseData.get_engine(cfg.db_dir, cfg.db_name + ".db")
    sess = BaseData.get_session_from_engine(engine)
    TableClass = ClassManager.get_class_orm(table_name, engine)
    UserClass = type(table_name, (Record, ), {})
    mapper(UserClass, TableClass)
    all_records = sess.query(UserClass).all()
    for record in all_records:
        if record:
            print_if_not_silent(silent, " ..Removing record %s" % record._id)
            try:
                os.remove(record.full_path())
            except OSError:
                continue
    TableClass.drop(engine)
    cfg.remove_table_from_config_file(table_name)
    os.remove(cfg.classes_file)
    shutil.rmtree(cfg.table_dir)
    if not silent:
        _remove_columns_display_message_epilogue()
    if not integrity_cancel:
        integrity_check(_cfg, _tbl, _al, _sil)
Example #7
0
def get_table(config_path, table_name=None, alias=None):
    """ Primary method for users importing

    :param config_path:
    :param table_name:
    :param alias:
    :return:
    """
    # Load config data from file
    cfg, config_path = ConfigManager.confirm_config_set(config_path)
    if alias:
        table_name = ConfigManager.get_name_by_alias(alias, cfg)
    config = ConfigManager(cfg, table_name)
    # Get session to return
    engine = BaseData.get_engine(config.db_dir, config.db_name + ".db")
    sess = BaseData.get_session_from_engine(engine)
    # Generate table class and name based on presence of alias
    TableClass = ClassManager.get_class_orm(table_name, engine)
    UserClass = type(table_name, (Record, ), {})
    # Map to SQL orm
    mapper(UserClass, TableClass)
    return RecordList(sess, UserClass, config)
Example #8
0
def create_database(db_name, table_name, directory_name, data_file, alias,
                    silent, integrity_cancel):
    """ Function called from dbdm initializes project/module

    :param silent:
    :param alias:
    :param db_name: (str)   Name of db
    :param table_name: (str)    Record that will be created
    :param directory_name: (str)    Directory with files to add
    :param data_file: (str)     File with metadata for storing in database
    :return:
    """
    # Confirm working dir does not exist and that directory with genomes does exist
    assert db_name != "None", CreateDBAssertString.WORKING_DB_NOT_SET
    assert table_name != "None", CreateDBAssertString.TABLE_NAME_NOT_SET
    assert os.path.isdir(
        db_name) is False, CreateDBAssertString.WORKING_DIR_EXISTS
    if directory_name != "None":
        assert os.path.isdir(
            directory_name), CreateDBAssertString.SEQUENCE_DIR_NOT_EXISTS
    table_name = table_name.lower()
    if not silent:
        _initialization_display_message_prelude(db_name, db_name, table_name,
                                                directory_name, data_file,
                                                alias)
    # Gather files to commit and initial data to store for each file
    print_if_not_silent(silent, "Beginning process...")
    print_if_not_silent(silent, " Getting files from %s" % directory_name)
    if directory_name != "None":
        genomic_files_to_add = (_f for _f in os.listdir(directory_name))
    else:
        genomic_files_to_add = ()
    data_types = {}
    initial_data = []
    if data_file is not "None":
        _initial_data = CountTable(data_file)
        # Ignore first column name, typically announcing "Name" or "Genome ID"
        # Get names and types of each column in dict
        # Key is header name (which will be used as name of column in database)
        # Value is type of data (int, float, str) that is found for a random genome id's data value
        # TODO: Data type determination requires a uniformity from the .tsv file data. Consider a workaround
        data_types = TypeMapper.get_translated_types(
            _initial_data, TypeMapper.py_type_to_string)
        initial_data.append(_initial_data)
    # Create working directories
    print_if_not_silent(silent,
                        " Creating directories at database root %s" % db_name)
    classes_dir, config_dir, db_dir, table_dir = _create_all_directories(
        db_name, table_name)
    # Create database file
    print_if_not_silent(silent, " Creating database file in %s" % db_dir)
    touch(os.path.join(db_dir, db_name + ".db"))
    # Write configuration info
    config_file = db_name + ".ini"
    print_if_not_silent(
        silent, " Writing database configuration to %s" %
        os.path.join(config_dir, config_file))
    config = Config()
    abs_path_working_dir = os.path.abspath(db_name)
    db_name = os.path.basename(db_name)
    config[ConfigKeys.DATABASES] = {
        ConfigKeys.db_name:
        db_name,
        ConfigKeys.working_dir:
        abs_path_working_dir,
        ConfigKeys.rel_work_dir:
        db_name,
        ConfigKeys.migrations_dir:
        os.path.join(abs_path_working_dir, Directories.MIGRATIONS),
        ConfigKeys.config_dir:
        os.path.join(abs_path_working_dir, Directories.CONFIG),
        ConfigKeys.db_dir:
        os.path.join(abs_path_working_dir, Directories.DATABASE),
        ConfigKeys.rel_db_dir:
        os.path.join(db_name, Directories.DATABASE),
    }
    config[table_name] = {
        ConfigKeys.rel_classes_dir: os.path.join(db_name, Directories.CLASSES),
    }
    config[ConfigKeys.TABLES_TO_DB] = {
        table_name: db_name,
    }
    config[ConfigKeys.TABLES_TO_ALIAS] = {
        "{}|{}".format(alias, table_name): table_name,
    }
    with open(os.path.join(config_dir, config_file), "w") as W:
        config.write(W)
    # Create table
    print_if_not_silent(
        silent, "Creating new table %s at %s" %
        (table_name, os.path.join(db_dir, db_name)))
    os.makedirs(table_dir)
    ClassManager.create_initial_table_in_db(db_name,
                                            db_name,
                                            table_name,
                                            data_types,
                                            silent,
                                            initial=False)
    # Populate table with data from file and genomes
    # Get config file - confirms that it was written correctly
    cfg = ConfigManager(config, table_name)
    if data_file is not "None":
        for _data in initial_data:
            ClassManager.populate_data_to_existing_table(
                table_name, _data, cfg, genomic_files_to_add, directory_name,
                silent)
    if not silent:
        _initialization_display_message_epilogue()
    if not integrity_cancel:
        integrity_check(db_name, table_name, "None", silent)
def create_table_in_existing_database(config_file, table_name, directory_name, data_file, alias, silent, integrity_cancel):
    """

    :param silent:
    :param config_file:
    :param table_name:
    :param directory_name:
    :param data_file:
    :param alias:
    :return:
    """
    assert table_name != "None", CreateDBAssertString.TABLE_NAME_NOT_SET
    assert config_file is not None, ConfigAssertString.CONFIG_FILE_NOT_PASSED
    _cfg, _tbl, _sil, _al = config_file, table_name, silent, alias
    config, config_file = ConfigManager.confirm_config_set(config_file)
    if table_name in config.keys():
        print("!! Record exists, exiting. To update table, use UPDATE !!")
        exit(1)
    if not silent:
        _create_table_display_message_prelude(config[ConfigKeys.DATABASES][ConfigKeys.db_name],
                                              config[ConfigKeys.DATABASES][ConfigKeys.rel_work_dir],
                                              table_name,
                                              directory_name,
                                              data_file,
                                              alias)
    data_types = {}
    initial_data = []
    if data_file is not "None":
        _initial_data = CountTable(data_file)
        # Ignore first column name, typically announcing "Name" or "Genome ID"
        # Get names and types of each column in dict
        # Key is header name (which will be used as name of column in database)
        # Value is type of data (int, float, str) that is found for a random genome id's data value
        # TODO: Data type determination requires a uniformity from the .tsv file data. Consider a workaround
        data_types = TypeMapper.get_translated_types(_initial_data, TypeMapper.py_type_to_string)
        initial_data.append(_initial_data)
    # Gather bio data from folder
    if directory_name != "None":
        genomic_files_to_add = (_f for _f in os.listdir(directory_name))
    else:
        genomic_files_to_add = ()
    # Create new table directories
    _create_all_directories(config[ConfigKeys.DATABASES][ConfigKeys.working_dir], table_name)
    # Update config object with new data
    table_name = table_name.lower()
    config[table_name] = {
        ConfigKeys.rel_classes_dir: os.path.join(config[ConfigKeys.DATABASES][ConfigKeys.rel_work_dir],
                                                 Directories.CLASSES),
    }
    config.set(ConfigKeys.TABLES_TO_DB, table_name, config[ConfigKeys.DATABASES][ConfigKeys.db_name])
    config.set(ConfigKeys.TABLES_TO_ALIAS, "{}|{}".format(alias, table_name), table_name)
    # Write new config file
    with open(config_file, "w") as W:
        config.write(W)
    # Update ConfigManager object
    cfg = ConfigManager(config, table_name)
    # Create new table and populate with new data
    ClassManager.create_initial_table_in_db(cfg.db_name, cfg.working_dir, table_name, data_types, silent, initial=False)
    if data_file is not "None":
        for _data in initial_data:
            ClassManager.populate_data_to_existing_table(table_name, _data, cfg,
                                                         genomic_files_to_add, directory_name, silent)
    if not silent:
        _create_table_display_message_epilogue()
    if not integrity_cancel:
        integrity_check(_cfg, _tbl, _al, _sil)
Example #10
0
def move_project(config_file, path, integrity_cancel, silent):
    """

    :param config_file:
    :param path:
    :param integrity_cancel:
    :param silent:
    :return:
    """
    assert path != 'None', "Path (-p) does not exist, exiting"
    current_path = os.path.abspath(
        os.path.abspath(os.path.relpath(config_file)))
    path = os.path.abspath(os.path.relpath(path))
    config, config_file = ConfigManager.confirm_config_set(config_file)
    old_path = config[ConfigKeys.DATABASES][ConfigKeys.working_dir]
    assert os.path.dirname(old_path) != os.path.abspath(path), \
        "Project exists in directory, cancelling"
    project_name = config[ConfigKeys.DATABASES][ConfigKeys.rel_work_dir]
    if not silent:
        _move_project_display_message_prelude(project_name, old_path, path),
    # Move directory from old location to new location
    shutil.move(current_path, path)
    # Update config file with new location
    abs_path_working_dir = os.path.abspath(os.path.join(path, project_name))
    db_name = project_name
    config[ConfigKeys.DATABASES] = {
        ConfigKeys.db_name:
        db_name,
        ConfigKeys.working_dir:
        abs_path_working_dir,
        ConfigKeys.rel_work_dir:
        db_name,
        ConfigKeys.migrations_dir:
        os.path.join(abs_path_working_dir, Directories.MIGRATIONS),
        ConfigKeys.config_dir:
        os.path.join(abs_path_working_dir, Directories.CONFIG),
        ConfigKeys.db_dir:
        os.path.join(abs_path_working_dir, Directories.DATABASE),
        ConfigKeys.rel_db_dir:
        os.path.join(db_name, Directories.DATABASE),
    }
    with open(
            os.path.join(path, project_name, Directories.CONFIG,
                         os.path.basename(config_file)), "w") as W:
        config.write(W)
    # Update location for each record in each table
    tables_in_database = config[ConfigKeys.TABLES_TO_DB].keys()
    engine = BaseData.get_engine(
        os.path.join(config[ConfigKeys.DATABASES][ConfigKeys.working_dir],
                     Directories.DATABASE),
        config[ConfigKeys.DATABASES][ConfigKeys.rel_work_dir] + ".db")
    sess = BaseData.get_session_from_engine(engine)
    for tbl_name in tables_in_database:

        TableClass = ClassManager.get_class_orm(tbl_name, engine)
        UserClass = type(tbl_name, (Record, ), {})
        mapper(UserClass, TableClass)
        for record in sess.query(UserClass).all():
            record.location = os.path.join(abs_path_working_dir,
                                           Directories.DATABASE, tbl_name)
        sess.commit()
        if not integrity_cancel:
            integrity_check(abs_path_working_dir, tbl_name, "None", silent)
    if not silent:
        _move_project_display_message_epilogue()