def replace(self, pattern, repl, flags=0, count=0):
"""
Replaces a regular expression pattern with repl in both the headers
and the body of the message. Encoded body will be decoded
before replacement, and re-encoded afterwards.
Returns:
The number of replacements made.
"""
if isinstance(pattern, str):
pattern = strutils.escaped_str_to_bytes(pattern)
if isinstance(repl, str):
repl = strutils.escaped_str_to_bytes(repl)
replacements = 0
if self.content:
self.content, replacements = re.subn(pattern,
repl,
self.content,
flags=flags,
count=count)
replacements += self.headers.replace(pattern,
repl,
flags=flags,
count=count)
return replacements
def replace(self, pattern, repl, flags=0, count=0):
"""
Replaces a regular expression pattern with repl in each "name: value"
header line.
Returns:
The number of replacements made.
"""
if isinstance(pattern, str):
pattern = strutils.escaped_str_to_bytes(pattern)
if isinstance(repl, str):
repl = strutils.escaped_str_to_bytes(repl)
pattern = re.compile(pattern, flags)
replacements = 0
flag_count = count > 0
fields = []
for name, value in self.fields:
line, n = pattern.subn(repl, name + b": " + value, count=count)
try:
name, value = line.split(b": ", 1)
except ValueError:
# We get a ValueError if the replacement removed the ": "
# There's not much we can do about this, so we just keep the header as-is.
pass
else:
replacements += n
if flag_count:
count -= n
if count == 0:
break
fields.append((name, value))
self.fields = tuple(fields)
return replacements
def replace(self, pattern, repl, flags=0, count=0):
"""
Replaces a regular expression pattern with repl in each "name: value"
header line.
Returns:
The number of replacements made.
"""
if isinstance(pattern, str):
pattern = strutils.escaped_str_to_bytes(pattern)
if isinstance(repl, str):
repl = strutils.escaped_str_to_bytes(repl)
pattern = re.compile(pattern, flags)
replacements = 0
flag_count = count > 0
fields = []
for name, value in self.fields:
line, n = pattern.subn(repl, name + b": " + value, count=count)
try:
name, value = line.split(b": ", 1)
except ValueError:
# We get a ValueError if the replacement removed the ": "
# There's not much we can do about this, so we just keep the header as-is.
pass
else:
replacements += n
if flag_count:
count -= n
if count == 0:
break
fields.append((name, value))
self.fields = tuple(fields)
return replacements
def _extract_modifier_config(modifier: bytes, ctx: dict) -> tuple:
modifier_obj = json.loads(modifier)
regex = _replace_modifier_values(modifier_obj['regex'], ctx)
if regex is None:
return None, None
replacement = _replace_modifier_values(modifier_obj['replacement'], ctx)
if replacement is None:
return None, None
regex = re.compile(strutils.escaped_str_to_bytes(regex))
replacement = strutils.escaped_str_to_bytes(replacement)
return regex, replacement
def parse_modify_spec(option) -> ModifySpec:
"""
The form for the modify_* options is as follows:
* modify_headers: [/flow-filter]/header-name/[@]header-value
* modify_body: [/flow-filter]/search-regex/[@]replace
The @ allows to provide a file path that is used to read the respective option.
Both ModifyHeaders and ModifyBody use ModifySpec to represent a single rule.
The first character specifies the separator. Example:
:~q:foo:bar
If only two clauses are specified, the flow filter is set to
match universally (i.e. ".*"). Example:
/foo/bar
Clauses are parsed from left to right. Extra separators are taken to be
part of the final clause. For instance, the last parameter (header-value or
replace) below is "foo/bar/":
/one/two/foo/bar/
"""
sep, rem = option[0], option[1:]
parts = rem.split(sep, 2)
if len(parts) == 2:
flow_filter_pattern = ".*"
subject, replacement = parts
elif len(parts) == 3:
flow_filter_pattern, subject, replacement = parts
else:
raise ValueError("Invalid number of parameters (2 or 3 are expected)")
flow_filter = flowfilter.parse(flow_filter_pattern)
if not flow_filter:
raise ValueError(f"Invalid filter pattern: {flow_filter_pattern}")
subject = strutils.escaped_str_to_bytes(subject)
replacement = strutils.escaped_str_to_bytes(replacement)
if replacement.startswith(b"@") and not os.path.isfile(
os.path.expanduser(replacement[1:])):
raise ValueError(f"Invalid file path: {replacement[1:]}")
return ModifySpec(flow_filter_pattern, flow_filter, subject, replacement)
def __init__(self, expr):
self.expr = expr
if self.is_binary:
expr = strutils.escaped_str_to_bytes(expr)
try:
self.re = re.compile(expr, self.flags)
except:
raise ValueError("Cannot compile expression.")
def __init__(self, expr):
self.expr = expr
if self.is_binary:
expr = strutils.escaped_str_to_bytes(expr)
try:
self.re = re.compile(expr, self.flags)
except:
raise ValueError("Cannot compile expression.")
def test_escaped_str_to_bytes():
assert strutils.escaped_str_to_bytes("foo") == b"foo"
assert strutils.escaped_str_to_bytes("\x08") == b"\b"
assert strutils.escaped_str_to_bytes("&!?=\\\\)") == br"&!?=\)"
assert strutils.escaped_str_to_bytes(u"\\x08") == b"\b"
assert strutils.escaped_str_to_bytes(u"&!?=\\\\)") == br"&!?=\)"
assert strutils.escaped_str_to_bytes(u"\u00fc") == b'\xc3\xbc'
with tutils.raises(ValueError):
strutils.escaped_str_to_bytes(b"very byte")
def test_escaped_str_to_bytes():
assert strutils.escaped_str_to_bytes("foo") == b"foo"
assert strutils.escaped_str_to_bytes("\x08") == b"\b"
assert strutils.escaped_str_to_bytes("&!?=\\\\)") == br"&!?=\)"
assert strutils.escaped_str_to_bytes(u"\\x08") == b"\b"
assert strutils.escaped_str_to_bytes(u"&!?=\\\\)") == br"&!?=\)"
assert strutils.escaped_str_to_bytes(u"\u00fc") == b'\xc3\xbc'
with pytest.raises(ValueError):
strutils.escaped_str_to_bytes(b"very byte")
def get_data(self) -> bytes:
txt = self._w.get_text()[0].strip()
try:
return strutils.escaped_str_to_bytes(txt)
except ValueError:
signals.status_message.send(self,
message="Invalid data.",
expire=1000)
raise
def replace(self, pattern, repl, flags=0, count=0):
"""
Replaces a regular expression pattern with repl in both the headers
and the body of the message. Encoded body will be decoded
before replacement, and re-encoded afterwards.
Returns:
The number of replacements made.
"""
if isinstance(pattern, str):
pattern = strutils.escaped_str_to_bytes(pattern)
if isinstance(repl, str):
repl = strutils.escaped_str_to_bytes(repl)
replacements = 0
if self.content:
self.content, replacements = re.subn(pattern, repl, self.content, flags=flags, count=count)
replacements += self.headers.replace(pattern, repl, flags=flags, count=count)
return replacements
def inject_tcp(self, flow: Flow, to_client: bool, message: str):
if not isinstance(flow, tcp.TCPFlow):
ctx.log.warn("Cannot inject TCP messages into non-TCP flows.")
message_bytes = strutils.escaped_str_to_bytes(message)
event = TcpMessageInjected(flow, tcp.TCPMessage(not to_client, message_bytes))
try:
self.inject_event(event)
except ValueError as e:
ctx.log.warn(str(e))
def get_data(self) -> bytes:
txt = self._w.get_text()[0].strip()
try:
return strutils.escaped_str_to_bytes(txt)
except ValueError:
signals.status_message.send(
self,
message="Invalid Python-style string encoding.",
expire=1000
)
raise
def replace(self, pattern, repl, flags=0, count=0):
"""
Replaces a regular expression pattern with repl in the headers, the
request path and the body of the request. Encoded content will be
decoded before replacement, and re-encoded afterwards.
Returns:
The number of replacements made.
"""
if isinstance(pattern, str):
pattern = strutils.escaped_str_to_bytes(pattern)
if isinstance(repl, str):
repl = strutils.escaped_str_to_bytes(repl)
c = super().replace(pattern, repl, flags, count)
self.path, pc = re.subn(
pattern, repl, self.data.path, flags=flags, count=count
)
c += pc
return c
def replace(self, pattern, repl, flags=0, count=0):
"""
Replaces a regular expression pattern with repl in the headers, the
request path and the body of the request. Encoded content will be
decoded before replacement, and re-encoded afterwards.
Returns:
The number of replacements made.
"""
if isinstance(pattern, str):
pattern = strutils.escaped_str_to_bytes(pattern)
if isinstance(repl, str):
repl = strutils.escaped_str_to_bytes(repl)
c = super().replace(pattern, repl, flags, count)
self.path, pc = re.subn(
pattern, repl, self.data.path, flags=flags, count=count
)
c += pc
return c
def parse_modify_spec(option, subject_is_regex: bool) -> ModifySpec:
"""
The form for the modify_* options is as follows:
* modify_headers: [/flow-filter]/header-name/[@]header-value
* modify_body: [/flow-filter]/search-regex/[@]replace
The @ allows to provide a file path that is used to read the respective option.
Both ModifyHeaders and ModifyBody use ModifySpec to represent a single rule.
The first character specifies the separator. Example:
:~q:foo:bar
If only two clauses are specified, the flow filter is set to
match universally (i.e. ".*"). Example:
/foo/bar
Clauses are parsed from left to right. Extra separators are taken to be
part of the final clause. For instance, the last parameter (header-value or
replace) below is "foo/bar/":
/one/two/foo/bar/
"""
sep, rem = option[0], option[1:]
parts = rem.split(sep, 2)
if len(parts) == 2:
flow_filter = _match_all
subject, replacement = parts
elif len(parts) == 3:
flow_filter_pattern, subject, replacement = parts
flow_filter = flowfilter.parse(flow_filter_pattern) # type: ignore
if not flow_filter:
raise ValueError(f"Invalid filter pattern: {flow_filter_pattern}")
else:
raise ValueError("Invalid number of parameters (2 or 3 are expected)")
subject = strutils.escaped_str_to_bytes(subject)
if subject_is_regex:
try:
re.compile(subject)
except re.error as e:
raise ValueError(f"Invalid regular expression {subject!r} ({e})")
spec = ModifySpec(flow_filter, subject, replacement)
try:
spec.read_replacement()
except IOError as e:
raise ValueError(f"Invalid file path: {replacement[1:]} ({e})")
return spec
def read_file(filename: str, escaped: bool) -> typing.AnyStr:
filename = os.path.expanduser(filename)
try:
with open(filename, "r" if escaped else "rb") as f:
d = f.read()
except IOError as v:
raise exceptions.CommandError(v)
if escaped:
try:
d = strutils.escaped_str_to_bytes(d)
except ValueError:
raise exceptions.CommandError("Invalid Python-style string encoding.")
return d
def get_data(self) -> strbytes:
txt = self._w.get_text()[0].strip()
try:
if self.type == bytes:
return strutils.escaped_str_to_bytes(txt)
else:
return txt.decode()
except ValueError:
signals.status_message.send(
self,
message="Invalid Python-style string encoding.",
expire=1000)
raise
def read_replacement(self) -> bytes:
"""
Process the replacement str. This usually just involves converting it to bytes.
However, if it starts with `@`, we interpret the rest as a file path to read from.
Raises:
- IOError if the file cannot be read.
"""
if self.replacement_str.startswith("@"):
return Path(self.replacement_str[1:]).expanduser().read_bytes()
else:
# We could cache this at some point, but unlikely to be a problem.
return strutils.escaped_str_to_bytes(self.replacement_str)
def inject_websocket(self, flow: Flow, to_client: bool, message: str, is_text: bool = True):
if not isinstance(flow, http.HTTPFlow) or not flow.websocket:
ctx.log.warn("Cannot inject WebSocket messages into non-WebSocket flows.")
message_bytes = strutils.escaped_str_to_bytes(message)
msg = websocket.WebSocketMessage(
Opcode.TEXT if is_text else Opcode.BINARY,
not to_client,
message_bytes
)
event = WebSocketMessageInjected(flow, msg)
try:
self.inject_event(event)
except ValueError as e:
ctx.log.warn(str(e))
def parse_modify_spec(option: str, subject_is_regex: bool) -> ModifySpec:
flow_filter, subject_str, replacement = parse_spec(option)
subject = strutils.escaped_str_to_bytes(subject_str)
if subject_is_regex:
try:
re.compile(subject)
except re.error as e:
raise ValueError(f"Invalid regular expression {subject!r} ({e})")
spec = ModifySpec(flow_filter, subject, replacement)
try:
spec.read_replacement()
except OSError as e:
raise ValueError(f"Invalid file path: {replacement[1:]} ({e})")
return spec
def read_file(filename: str, callback: Callable[..., None], escaped: bool) -> Optional[str]:
if not filename:
return
filename = os.path.expanduser(filename)
try:
with open(filename, "r" if escaped else "rb") as f:
d = f.read()
except IOError as v:
return str(v)
if escaped:
try:
d = strutils.escaped_str_to_bytes(d)
except ValueError:
return "Invalid Python-style string encoding."
# TODO: Refactor the status_prompt_path signal so that we
# can raise exceptions here and return the content instead.
callback(d)
def read_file(filename: str, callback: Callable[..., None],
escaped: bool) -> Optional[str]:
if not filename:
return
filename = os.path.expanduser(filename)
try:
with open(filename, "r" if escaped else "rb") as f:
d = f.read()
except IOError as v:
return str(v)
if escaped:
try:
d = strutils.escaped_str_to_bytes(d)
except ValueError:
return "Invalid Python-style string encoding."
# TODO: Refactor the status_prompt_path signal so that we
# can raise exceptions here and return the content instead.
callback(d)
def parse(self, manager: "CommandManager", t: type, s: str) -> bytes:
try:
return strutils.escaped_str_to_bytes(s)
except ValueError as e:
raise exceptions.TypeError(str(e))
def __init__(self, val):
self.val = strutils.escaped_str_to_bytes(val)
def __init__(self, val):
self.val = strutils.escaped_str_to_bytes(val)
