2
## Copyright (C) 2004, 2005 Red Hat, Inc.
3
## Written by Gary Benson <gbenson@redhat.com>
5
## This program is free software; you can redistribute it and/or modify
6
## it under the terms of the GNU General Public License as published by
7
## the Free Software Foundation; either version 2 of the License, or
8
## (at your option) any later version.
10
## This program is distributed in the hope that it will be useful,
11
## but WITHOUT ANY WARRANTY; without even the implied warranty of
12
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
## GNU General Public License for more details.
15
"""Read Java(TM) class files."""
17
import cStringIO as StringIO
21
def __init__(self, arg):
22
if hasattr(arg, "read"):
24
elif type(arg) == type(""):
25
if arg.startswith("\xca\xfe\xba\xbe"):
26
self.fp = StringIO.StringIO(arg)
28
self.fp = open(arg, "r")
30
raise TypeError, type(arg)
32
magic = self._read_int()
33
assert magic == 0xcafebabeL
34
minor, major = self._read(">HH")
35
self.version = (major, minor)
37
self.pool_integrity_checks = None
40
except AssertionError:
41
self.pool_integrity_checks = []
43
self._read_constants_pool()
45
self.access_flags = self._read_short()
46
self.name = self._read_reference_Class()
47
self.super = self._read_reference_Class()
49
self.interfaces = self._read_interfaces()
50
self.fields = self._read_fieldsormethods()
51
self.methods = self._read_fieldsormethods()
52
self.attributes = self._read_attributes()
54
if self.pool_integrity_checks is not None:
55
for index, tag in self.pool_integrity_checks:
56
assert self.constants[index][0] == tag
58
del self.fp, self.pool_integrity_checks
62
attrs = [attr for attr in dir(self)
63
if not attr.startswith("_") and attr != "Member"]
66
result.append("%-13s %s" % (
67
attr + ":", attr == "constants" and
68
"<ELIDED>" or repr(getattr(self, attr))))
69
return "\n".join(result)
71
def _read_constants_pool(self):
74
for i in xrange(1, self._read_short()):
79
1: "Utf8", 3: "Integer", 4: "Float", 5: "Long",
80
6: "Double", 7: "Class", 8: "String", 9: "Fieldref",
81
10: "Methodref", 11: "InterfaceMethodref",
82
12: "NameAndType"}[self._read_byte()]
83
skip = tag in ("Long", "Double") # crack crack crack!
84
self.constants[i] = (tag, getattr(self, "_read_constant_" + tag)())
86
def _read_interfaces(self):
88
for i in xrange(self._read_short()):
89
result.append(self._read_reference_Class())
92
def _read_fieldsormethods(self):
94
for i in xrange(self._read_short()):
95
result.append(self.Member(self))
99
def __init__(self, source):
100
self.access_flags = source._read_short()
101
self.name = source._read_reference_Utf8()
102
self.descriptor = source._read_reference_Utf8()
103
self.attributes = source._read_attributes()
107
attrs = [attr for attr in dir(self) if not attr.startswith("_")]
110
value = getattr(self, attr)
111
if attr == "attributes" and value.has_key("Code"):
113
value.update({"Code": "<ELIDED>"})
114
result.append("%-13s %s" % (
115
attr + ":", repr(value).replace(
116
"'Code': '<ELIDED>'", "'Code': <ELIDED>")))
117
return ("\n%s" % (15 * " ")).join(result)
119
def _read_attributes(self):
121
for i in xrange(self._read_short()):
122
name = self._read_reference_Utf8()
123
data = self.fp.read(self._read_int())
124
assert not result.has_key(name)
128
# Constants pool reference reader convenience functions
130
def _read_reference_Utf8(self):
131
return self._read_references("Utf8")[0]
133
def _read_reference_Class(self):
134
return self._read_references("Class")[0]
136
def _read_reference_Class_NameAndType(self):
137
return self._read_references("Class", "NameAndType")
139
def _read_references(self, *args):
142
index = self._read_short()
143
if self.pool_integrity_checks is not None:
144
self.pool_integrity_checks.append((index, arg))
148
# Constants pool constant reader functions
150
def _read_constant_Utf8(self):
151
constant = self.fp.read(self._read_short())
153
constant = constant.decode("utf-8")
155
constant = _bork_utf8_decode(constant)
157
constant = constant.encode("us-ascii")
162
def _read_constant_Integer(self):
163
return self._read_int()
165
def _read_constant_Float(self):
166
return self._read(">f")[0]
168
def _read_constant_Long(self):
169
return self._read(">q")[0]
171
def _read_constant_Double(self):
172
return self._read(">d")[0]
174
_read_constant_Class = _read_reference_Utf8
175
_read_constant_String = _read_reference_Utf8
176
_read_constant_Fieldref = _read_reference_Class_NameAndType
177
_read_constant_Methodref = _read_reference_Class_NameAndType
178
_read_constant_InterfaceMethodref = _read_reference_Class_NameAndType
180
def _read_constant_NameAndType(self):
181
return self._read_reference_Utf8(), self._read_reference_Utf8()
183
# Generic reader functions
186
# XXX how else to read 32 bits on a 64-bit box?
187
h, l = map(long, self._read(">HH"))
190
def _read_short(self):
191
return self._read(">H")[0]
193
def _read_byte(self):
194
return self._read("B")[0]
196
def _read(self, fmt):
197
return struct.unpack(fmt, self.fp.read(struct.calcsize(fmt)))
199
def _bork_utf8_decode(data):
201
bytes, unicode = map(ord, data), ""
207
assert b2 & 0xC0 == 0x80
211
assert b3 & 0xC0 == 0x80
213
((b1 & 0x0f) << 12) + ((b2 & 0x3f) << 6) + (b3 & 0x3f))
215
unicode += unichr(((b1 & 0x1f) << 6) + (b2 & 0x3f))
217
unicode += unichr(b1)
220
if __name__ == "__main__":
221
print Class("/usr/share/katana/build/ListDependentClasses.class")