Source code for plaso.parsers.winreg_plugins.dtfabric_plugin

# -*- coding: utf-8 -*-
"""Shared functionality for dtFabric-based data format Registry plugins."""

from __future__ import unicode_literals

import abc
import os

from dtfabric import errors as dtfabric_errors
from dtfabric.runtime import fabric as dtfabric_fabric

from plaso.lib import errors
from plaso.parsers.winreg_plugins import interface


[docs]class DtFabricBaseWindowsRegistryPlugin(interface.WindowsRegistryPlugin): """Shared functionality for dtFabric-based data format Registry plugins. A dtFabric-based data format Windows Registry parser plugin defines its data format structures in dtFabric definition file, for example "dtfabric.yaml": name: int32 type: integer description: 32-bit signed integer type attributes: format: signed size: 4 units: bytes --- name: point3d aliases: [POINT] type: structure description: Point in 3 dimensional space. attributes: byte_order: little-endian members: - name: x aliases: [XCOORD] data_type: int32 - name: y data_type: int32 - name: z data_type: int32 The path to the definition file is defined in the class constant "_DEFINITION_FILE" and will be read on class instantiation. The definition files contains data type definitions such as "int32" and "point3d" in the previous example. A data type map can be used to create a Python object that represent the data type definition mapped to a byte stream, for example if we have the following byte stream: 01 00 00 00 02 00 00 00 03 00 00 00 The corresponding "point3d" Python object would be: point3d(x=1, y=2, z=3) A parser that wants to implement a dtFabric-based data format parser needs to: * define a definition file and override _DEFINITION_FILE; * implement the ParseFileObject method. The _GetDataTypeMap method of this class can be used to retrieve data type maps from the "fabric", which is the collection of the data type definitions in definition file. Data type maps are cached for reuse. The _ReadStructure method of this class can be used to read structure data from a file-like object and create a Python object using a data type map. """ # The dtFabric definition file, which must be overwritten by a subclass. _DEFINITION_FILE = None # Preserve the absolute path value of __file__ in case it is changed # at run-time. _DEFINITION_FILES_PATH = os.path.dirname(__file__) def __init__(self): """Initializes a dtFabric-based data format Registry plugin.""" super(DtFabricBaseWindowsRegistryPlugin, self).__init__() self._data_type_maps = {} self._fabric = self._ReadDefinitionFile(self._DEFINITION_FILE) def _GetDataTypeMap(self, name): """Retrieves a data type map defined by the definition file. The data type maps are cached for reuse. Args: name (str): name of the data type as defined by the definition file. Returns: dtfabric.DataTypeMap: data type map which contains a data type definition, such as a structure, that can be mapped onto binary data. """ data_type_map = self._data_type_maps.get(name, None) if not data_type_map: data_type_map = self._fabric.CreateDataTypeMap(name) self._data_type_maps[name] = data_type_map return data_type_map def _ReadDefinitionFile(self, filename): """Reads a dtFabric definition file. Args: filename (str): name of the dtFabric definition file. Returns: dtfabric.DataTypeFabric: data type fabric which contains the data format data type maps of the data type definition, such as a structure, that can be mapped onto binary data or None if no filename is provided. """ if not filename: return None path = os.path.join(self._DEFINITION_FILES_PATH, filename) with open(path, 'rb') as file_object: definition = file_object.read() return dtfabric_fabric.DataTypeFabric(yaml_definition=definition) def _ReadStructureFromByteStream( self, byte_stream, file_offset, data_type_map, context=None): """Reads a structure from a byte stream. Args: byte_stream (bytes): byte stream. file_offset (int): offset of the structure data relative to the start of the file-like object. data_type_map (dtfabric.DataTypeMap): data type map of the structure. context (Optional[dtfabric.DataTypeMapContext]): data type map context. The context is used within dtFabric to hold state about how to map the data type definition onto the byte stream. In this class it is used to determine the size of variable size data type definitions. Returns: object: structure values object. Raises: ParseError: if the structure cannot be read. ValueError: if file-like object or data type map is missing. """ if not byte_stream: raise ValueError('Missing byte stream.') if not data_type_map: raise ValueError('Missing data type map.') try: return data_type_map.MapByteStream(byte_stream, context=context) except (dtfabric_errors.ByteStreamTooSmallError, dtfabric_errors.MappingError) as exception: raise errors.ParseError(( 'Unable to map {0:s} data at offset: 0x{1:08x} with error: ' '{2!s}').format(data_type_map.name or '', file_offset, exception)) # pylint 1.9.3 wants a docstring for kwargs, but this is not useful to add. # pylint: disable=missing-param-doc
[docs] @abc.abstractmethod def ExtractEvents(self, parser_mediator, registry_key, **kwargs): """Extracts events from a Windows Registry key. Args: parser_mediator (ParserMediator): mediates interactions between parsers and other components, such as storage and dfvfs. registry_key (dfwinreg.WinRegistryKey): Windows Registry key.
"""