Skip to content

Feature/necessary typing #7

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Apr 11, 2022
Merged

Feature/necessary typing #7

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e0e95f4
Added some necessary type definitions and improved parsing globals.
Hadatko Apr 7, 2022
c96e7ba
C language is case sensitive.
Hadatko Apr 8, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 14 additions & 0 deletions dissect/cstruct/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,8 +11,15 @@
RawType,
BaseType,
StructureType,
StructureFieldType,
EnumType,
ArrayType,
PackedType,
CharType,
WcharType,
UnionType,
VoidType,
typeAll,
Copy link
Collaborator

@qcabrol qcabrol Apr 8, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

typeAll ?

Copy link
Member Author

@Hadatko Hadatko Apr 9, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

typeAll => i created few predefined types to simpliefied typing. This one i exposed.

Copy link
Member Author

@Hadatko Hadatko Apr 9, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

image

Error,
ParserError,
CompilerError,
Expand All @@ -33,8 +40,15 @@
"RawType",
"BaseType",
"StructureType",
"StructureFieldType",
"EnumType",
"ArrayType",
"PackedType",
"CharType",
"WcharType",
"UnionType",
"VoidType",
"typeAll",
"Error",
"ParserError",
"CompilerError",
Expand Down
124 changes: 86 additions & 38 deletions dissect/cstruct/cstruct.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,6 +33,7 @@
from io import BytesIO
from collections import OrderedDict
import json
import typing

try:
from builtins import bytes as newbytes
Expand Down Expand Up @@ -88,6 +89,7 @@ def __repr__(self):
return '<StructureType {name} +compiled>'
"""

#common regex values
COMMENT_MULTILINE = r'/\*[^*]*\*+(?:[^/*][^*]*\*+)*/'
COMMENT_INLINE = r'//[^\n]*'
COMMENT_MULTILINE_REPEATED = r'(^[ ]*('+COMMENT_INLINE+r'|'+COMMENT_MULTILINE+r'([ ]*('+COMMENT_INLINE+r'|'+COMMENT_MULTILINE+r'))*)[ \t\r]*\n?)*^[ ]*('+COMMENT_INLINE+r'|(?P<commentBlock>'+COMMENT_MULTILINE+r'))+'
Expand All @@ -97,6 +99,12 @@ def __repr__(self):
#print(f"COMMENT_REGEX_START:{COMMENT_REGEX_START}")
#print(f"COMMENT_REGEX_END:{COMMENT_REGEX_END}")

#types
typeCommentAttributes=dict[str,str]
typeValuesDetails =dict[str,dict[str,typing.Union[str,int,typeCommentAttributes]]]
typeAll = typing.Union[str,'PackedType','CharType','WcharType','BytesInteger','VoidType','ArrayType','StructureType','EnumType']
typeTypedefs = dict[str, typeAll]

class Error(Exception):
pass

Expand Down Expand Up @@ -139,7 +147,7 @@ def __init__(self, endian='<', pointer='uint64'):

self.consts = {}
self.lookups = {}
self.typedefs = {
self.typedefs:typeTypedefs = {
'byte': 'int8',
'ubyte': 'uint8',
'uchar': 'uint8',
Expand Down Expand Up @@ -186,7 +194,7 @@ def __init__(self, endian='<', pointer='uint64'):

self.pointer = self.resolve(pointer)

def addtype(self, name, t, replace=False):
def addtype(self, name:str, t:typeAll, replace=False):
"""Add a type or type reference.

Args:
Expand All @@ -197,8 +205,7 @@ def addtype(self, name, t, replace=False):
Raises:
ValueError: If the type already exists.
"""
name = name.lower()
if not replace and name.lower() in self.typedefs:
if not replace and name in self.typedefs:
raise ValueError("Duplicate type: %s" % name)

self.typedefs[name] = t
Expand All @@ -223,7 +230,7 @@ def load(self, s, deftype=None, **kwargs):

parser.parse(s)

def loadfile(self, s, deftype=None, **kwargs):
def loadfile(self, s:str, deftype:typing.Optional[int]=None, **kwargs:typing.Union['cstruct',bool]):
"""Load structure definitions from a file.

The given path will be read and parsed using the .load() function.
Expand Down Expand Up @@ -268,25 +275,61 @@ def resolve(self, name):
return t

for i in xrange(10):
if t.lower() not in self.typedefs:
if t not in self.typedefs:
raise ResolveError("Unknown type %s" % name)

t = self.typedefs[t.lower()]
t = self[t]

if not isinstance(t, str):
return t

raise ResolveError("Recursion limit exceeded while resolving type %s" % name)

def __getattr__(self, attr):
if attr.lower() in self.typedefs:
return self.typedefs[attr.lower()]
if attr in self.typedefs:
return self.typedefs[attr]

if attr in self.consts:
return self.consts[attr]

for enum in self.getEnumTypes():
for enumMember in enum.values.items():
if enumMember[0] == attr:
return enumMember[1]


raise AttributeError("Invalid Attribute: %s" % attr)

def __getitem__(self, item):
return self.__getattr__(item)

def keys(self):
return self.typedefs.keys()

def values(self):
return self.typedefs.values()

def items(self):
return self.typedefs.items()

def getEnumTypes(self):
retVal: list[EnumType] = []
allTypes = self.typedefs
for currentType in allTypes.values():
if isinstance(currentType, EnumType):
retVal.append(currentType)

return retVal

def getStructTypes(self):
retVal: list[StructureType] = []
allTypes = self.typedefs
for currentType in allTypes.values():
if isinstance(currentType, StructureType):
retVal.append(currentType)

return retVal


class Parser(object):
"""Base class for definition parsers.
Expand Down Expand Up @@ -428,7 +471,7 @@ def _parse_fields_struct(self, s):
COMMENT_REGEX_START+r'(?P<type>[^\s]+)\s+(?P<name>[^\s\[:]+)(\s*:\s*(?P<bits>\d+))?(\[(?P<count>[^;\n]*)\])?;'+COMMENT_REGEX_END,
s, re.MULTILINE
)
r = []
r:list[StructureFieldType] = []
for f in fields:
d = f.groupdict()
if d['type'].startswith('//'):
Expand Down Expand Up @@ -458,13 +501,13 @@ def _parse_fields_struct(self, s):
d['name'] = d['name'][1:]
type_ = Pointer(self.cstruct, type_)

field = StructureField(d['name'], type_, int(d['bits']) if d['bits'] else None, commentAttributes=commentAttributes)
field = StructureFieldType(d['name'], type_, int(d['bits']) if d['bits'] else None, commentAttributes=commentAttributes)
r.append(field)

return r

def parse_comment_block(self,s):
commentAttributes={}
commentAttributes:typeCommentAttributes={}

#parse the comment header
if s is not None and s.startswith('/*'):
Expand Down Expand Up @@ -622,7 +665,7 @@ class Expression(object):
('<<', lambda a, b: a << b),
]

def __init__(self, cstruct, expr):
def __init__(self, cstruct:cstruct, expr):
self.cstruct = cstruct
self.expr = expr

Expand Down Expand Up @@ -667,6 +710,11 @@ def evaluate_part(self, e, v):
if e in self.cstruct.consts:
return self.cstruct.consts[e]

for enum in self.cstruct.getEnumTypes():
for enumMember in enum.values.items():
if enumMember[0] == e:
return enumMember[1]

return int(e)

def __repr__(self):
Expand Down Expand Up @@ -778,9 +826,9 @@ def __call__(self, *args, **kwargs):
class RawType(BaseType):
"""Base class for raw types that have a name and size."""

def __init__(self, cstruct, name=None, size=0):
self.name = name
self.size = size
def __init__(self, cstruct, name:typing.Union[str,None]=None, size:int=0):
self.name:typing.Union[str,None] = name
self.size:int = size
super(RawType, self).__init__(cstruct)

def __len__(self):
Expand All @@ -796,12 +844,12 @@ def __repr__(self):
class StructureType(BaseType):
"""Type class for structures."""

def __init__(self, cstruct, name, fields=None, commentAttributes={}):
self.name = name
self.size = None
def __init__(self, cstruct, name:str, fields:list['StructureFieldType']=[], commentAttributes:typeCommentAttributes={}):
self.name:str = name
self.size:typing.Union[int,None] = None
self.lookup = OrderedDict()
self.fields = fields if fields else []
self.commentAttributes = commentAttributes
self.fields:list[StructureFieldType] = fields
self.commentAttributes:typeCommentAttributes = commentAttributes


for f in self.fields:
Expand Down Expand Up @@ -917,15 +965,15 @@ def _write(self, stream, data):

return num

def add_fields(self, name, type_, offset=None, commentAttributes={}):
def add_fields(self, name, type_, offset=None, commentAttributes:typeCommentAttributes={}):
"""Add a field to this structure.

Args:
name: The field name.
type_: The field type.
offset: The field offset.
"""
field = StructureField(name, type_, offset=offset, commentAttributes=commentAttributes)
field = StructureFieldType(name, type_, offset=offset, commentAttributes=commentAttributes)
self.fields.append(field)
self.lookup[name] = field
self.size = None
Expand Down Expand Up @@ -1023,15 +1071,15 @@ def reset(self):
self._remaining = 0


class StructureField(object):
class StructureFieldType(object):
"""Holds a structure field."""

def __init__(self, name, type_, bits=None, offset=None, commentAttributes={}):
self.name = name
self.type = type_
def __init__(self, name:str, type_:typeAll, bits=None, offset:typing.Union[int,None]=None, commentAttributes:typeCommentAttributes={}):
self.name:str = name
self.type:typeAll = type_
self.bits = bits
self.offset = offset
self.commentAttributes = commentAttributes
self.offset:typing.Union[int,None] = offset
self.commentAttributes:typeCommentAttributes = commentAttributes

def __repr__(self):
return '<Field {} {} {}>'.format(self.name, self.type, self.commentAttributes)
Expand All @@ -1048,8 +1096,8 @@ class ArrayType(BaseType):
x[expr] -> expr -> dynamic length.
"""

def __init__(self, cstruct, type_, count):
self.type = type_
def __init__(self, cstruct, type_:typeAll, count):
self.type:typeAll = type_
self.count = count
self.dynamic = isinstance(self.count, Expression) or self.count is None

Expand Down Expand Up @@ -1352,11 +1400,11 @@ class EnumType(RawType):
};
"""

def __init__(self, cstruct, name, type_, values, valuesDetails, commentAttributes={}):
self.type = type_
self.values = values
self.valuesDetails = valuesDetails
self.commentAttributes = commentAttributes
def __init__(self, cstruct, name:str, type_:typeAll, values:dict[str,int], valuesDetails:typeValuesDetails, commentAttributes:typeCommentAttributes={}):
self.type:typeAll = type_
self.values:dict[str,int] = values
self.valuesDetails:typeValuesDetails = valuesDetails
self.commentAttributes:typeCommentAttributes = commentAttributes
self.reverse = {}

for k, v in values.items():
Expand Down Expand Up @@ -1449,10 +1497,10 @@ def __call__(self):
return self.value


class Union(RawType):
class UnionType(RawType):
def __init__(self, cstruct):
self.cstruct = cstruct
super(Union, self).__init__(cstruct)
super(UnionType, self).__init__(cstruct)

def _read(self, stream):
raise NotImplementedError()
Expand Down