Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER
Unsupported opcode: GET_YEILD_FROM_ITER

And yield from is still valid in 3.5 as long as you have the @asyncio.coroutine decorator on top of the function.

it outputs this:

    def onPressKeyDict[app.DIK_1]():
        return self._GameWindow__PressNumKey(1)

but schould be:

onPressKeyDict[app.DIK_1] = lambda : self._GameWindow__PressNumKey(1)

In Python 3.5 the opcode for WITH_CLEANUP is replaced by the pair of opcodes WITH_CLEANUP_START and WITH_CLEANUP_FINISH. This is to support async "with". These opcodes are not handled by pycdc, so the "with" statement compiled to bytecode with python 3.5 cannot be decompiled.

A quick and dirty solution is to add the following after the Pyc::WITH_CLEANUP block in ASTree.cpp. As the comment below says, this does not support async "with". In the case of not using async "with", the two opcodes will follow eachother immediately, and so should do the same as the old WITH_CLEANUP.

        case Pyc::WITH_CLEANUP_START:
            {
                // This is just a copy of the WITH_CLEANUP block
                // Stack top should be a None. Ignore it.
                PycRef<ASTNode> none = stack.top();
                stack.pop();

                if (none != Node_NULL) {
                    fprintf(stderr, "Something TERRIBLE happened!\n");
                    break;
                }

                if (curblock->blktype() == ASTBlock::BLK_WITH
                        && curblock->end() == curpos) {
                    PycRef<ASTBlock> with = curblock;
                    blocks.pop();
                    curblock = blocks.top();
                    curblock->append(with.cast<ASTNode>());
                }
                else {
                    fprintf(stderr, "Something TERRIBLE happened! No matching with block found for WITH_CLEANUP at %d\n", curpos);
                }
            }
            break;
        case Pyc::WITH_CLEANUP_FINISH:
            {
                // Do nothing. (This does not support async with.)
            }
            break;

when I run pycdc on some pyc, pycdc crashes. So is there any guide to find which cause this problem, and then we can fix it. I know we can debug the C source file, but it is some difficult for C newbie. So if there some way can aid debug, it will be good for this project and the user. Thanks

On python 3.2.6, this

# From python 3.2 configparser.py
def __init__(self, defaults=None, dict_type=_default_dict,
             allow_no_value=False, *, delimiters=('=', ':'),
             comment_prefixes=('#', ';'), inline_comment_prefixes=None,
             strict=True, empty_lines_in_values=True,
             default_section=DEFAULTSECT,
             interpolation=_UNSET):
    return 

gives:

def __init__(self = None, defaults = None, dict_type = None, allow_no_value = None):
    pass

At one point I went though all of the test case from that repo and there were/are a number of other failures, including some SEGV's. If it is of interest I can open a ticket for each one.

BTW great work. I wish uncompyle6 were as simple and tight and handled its control structures as well as pycdc.

After I've compiled complete package (using Microsoft Visual Studio 2012), whenever I try to decompile or disassemble ANY python .pyc file (for example - any file located in "tests" directory) I'm getting as a result

Bad MAGIC!
Could not load file test_with.pyc

Is there any specific reason why it doesn't work?

thanks

I use this Tool to decompyle .pyc Files and want write this to a File.
Now i use the freopen Method like this: "freopen( "output.py", "w", stdout );"

My problem is that every time I decompyle a file will be named output.py. But I want that when I e.g. decompyle test.pyc this is then named test.py and when I load test-2.pyc this is then named test-2.py.

Can you help me?

a = lambda x: [i**2 for i in x]
def f4(x):
    return a(x)

compile and use pycdc to decompile ,generate:

a = lambda x: continue[ i ** 2 for i in x ]

def f4(x):
    return a(x)

In particular，when lambda return a generator like this:

a = lambda x: (i**2 for i in x)
def f4(x):
    return a(x)

use pycdc to decompile ,generate:

a = lambda x: (lambda .0: continue)(x)
def f4(x):
    return a(x)

look like something wrong....

when use pycdc
there is an error
PycData::get16()
{
/* Ensure endianness */
int result = getByte() & 0xFF;
result |= (getByte() & 0xFF) << 8;
return (result | -(result & 0x8000));
}

Tuple unpacking in for loops is not handled correctly:

def map_with_index(func, lst):
    res = []
    for i, x in enumerate(lst):
        res.append(func(i, x)) 
    return res

is reconstructed as

def map_with_index(func, lst):
    res = []
    for None in enumerate(lst):
        (i, x) = None
    return res

If I unpack the tuple inside the loop, there is no problem, aka

def map_with_index_2(func, lst):
    res = []
    for t in enumerate(lst):
        i, x = t
        res.append(func(i, x))
    return res

is fine.

Python version: 2.7.3

This came from decompyle's suite of Python 2.2 tests, so I'm presuming that it's actually a valid PYC file. pycdc and pycdas both hang while trying to parse it.

I'm not familiar with the bytecode parsing code, so my attempts to debug it have just led me in circles. It doesn't crash at any point, it just seems to either keep parsing the same thing or sit there waiting for something.

when do the file program abends,file like this on the web : https://www.dropbox.com/s/tn2krl17h4v7134/Root.pyc

Your Python Decompiler is awesome! At least works for me on python2.2 generated ancient .pyc files.

I am currently struggling to port the CASIO IDES waveconverter (synth patch compressor) on Linux. Fortunately, it is a py2exe PE containing all the .py/.pyc. With your handy tool, the interface (wxWidgets) got working on my LinuxMint 17.2 x64 (yet there are .pyd libraries).

However, I had to $git clone $cmake $make just to run the $pycdc for the first time. Why don't you create some Releases with binary files on GitHub? That would be easier for some people, if, say, they wanted to just decompyle a bunch of .pyc. (I even found out I had no CMake in the system)

Hi,

After downloading CMake as suggested and building the project files in Visual Studio 2013 I took the precaution of throwing the exe (pycdc.exe) at Virus total. AegisLab reports Troj.W32.Gen. Has anyone come across this before?

I hope it is anomalous..

Kithara

Input:

try:
    x = 1
except:
    x = 2
else:
    x = 3

Output:

try:
    x = 1
except:
    x = 2
x = 3

Works correctly for a try..except..else..finally.

Today I found this decompiler don't support following codes:

def useVoxCheck(arg):
  global useVoxCtrl
  useVoxCtrl = True if arg == "true" else False
  cmds.floatFieldGrp('voxSizeText', edit=True, enable=1 if arg == "false" else 0)  # crash at this line

this will crash pycdc, after several unit test, I'm sure it is cause by the arg in line 3: 'edit=True, enable=1 if arg == "false" else 0', but pydas is fine on this. I tried debuging it but failed.

Would anyone fix it ? Thanks!

Hey nice Project i modified it to use it in one of my project as dll im actually running it for every *.PYC file in the directory.

Its a for loop my only problem is i get the following error:
Warning: block stack is not empty!
Warning: Stack history is not empty!
Warning: Stack history is not empty!
Warning: Stack history is not empty!

Why is this happening im calling

        if((stream = freopen(p3, "w", stdout)) == NULL)
          exit(-1);
        mod.loadFromFile(p1);
        if (!mod.isValid()) {
            fprintf(stderr, "Could not load file %s\n", p2);
            return 1;
        }
        decompyle(mod.code(), &mod);
        stream = freopen("CON", "w", stdout); // Redirect the output back to console
        return 1;

Kind Regards

What python versions does pycdc support?

def myfunc((a, b) = (0, 1)):
    print a
    print b

gets decompyled into

def myfunc(.0 = (0, 1))
    (a, b) = .0
    print a
    print b

This is especially problematic for lambdas, and requires some way of replacing function argument names with nodes as the code is parsed.

Any function that changes the value of a global variable should declare that global variable in scope with the global keyword.

This would require each function to keep a list/set of global variables that are used within that function, and print out that list with the global keyword as the first line of disassembled output.

This is mostly just a reminder that this needs to be done at some point.

I used pycdc to decompile a pyc file

./pycdc x.pyc
it can print the part of source code
but after running for a moment , it get this error
it cannot print all the source code

I can give the pyc file if needed

Currently the usage of the older C I/O stuff usage and association with "FILE *" (pyc_output) is scattered throughout the code, I have already done an initial re-write of this to use std::ostream, which come out with a bit cleaner code. (Personal opinion)

This is a reasonable size change (non-isolated) so this is more of a discussion 'issue' of the pro's/con's of the different approaches.

Pros of keeping current / Cons of change:

• Giant diff (Counter: No frequent commits, conflicts rare)
• Ain't broke don't fix it (Counter: Limited room for customize-ability)
• printf easier to read (Counter: Personal preference, helpers can be made easier with iostream - less mystical format strings, and 'printf' validation, type safety)

Cons of current / Pros of change:

• (All counter of other)
• Use std::stringstream to store intermediate formats (I found very useful debugging)
• Easier to define different output types (Not something FILE_) (Counter: Not often non-FILE_; Count-counter: Debug buffer)
• Debug output more closely associated with object (Start using operator<<) not switch(x)
• More common common functionality (Similar to previous; could be done with stdio)
• C++ not C (could be straw-man)

input:

pairs = [('a',2)]
def filter_n_sort_pairs():
    z = [int(x) for (x, y) in pairs if (x not in ['c',4] and not isinstance(x, int))]
    return z.sort()

decompilation result: (+warning blocks not empty)

def filter_n_sort_pairs():
    for (x, y) in pairs:
        if isinstance(x, int) and x not in ['c',4]:
            continue
            z = [][int(var_1)]
            z.sort()
            return z

I think checking current block boundaries (when known) could solve this issue, I may take a look at it myself. (v2.7.10)

thank you

Simply use "make" command not found. Where is the Makefile ? Please make a more precise documentation...

I got this msg in visual studio and I can't do anything with that. Any idea how to fix it?

I compiled pycdc with python34.
This file is not decompiled:

REM this is batch file *.bat
pycdc.exe pyc\bughurt.pyc > py\bughurt.py
pycdc.exe pyc\checkers_g.pyc > py\checkers_g.py
pycdc.exe pyc\1_Lisa.pyc > py\1_Lisa.py
pycdc.exe pyc\memgen.pyc > py\memgen.py
pycdc.exe pyc\mirror.pyc > py\mirror.py
pycdc.exe pyc\pixel.pyc > py\pixel.py
pycdc.exe pyc\weather.pyc > py\weather.py
pycdc.exe pyc\imsearch.pyc > py\imsearch.py
pycdc.exe pyc\like.pyc > py\like.py
pycdc.exe pyc\func.pyc > py\func.py
pause

all pyc, py and screenshots files in archive https://drive.google.com/open?id=0B69FzC4j62F9ZmlILUt3eWlvSVE

Please check and help me!

FILE: ASTree.cpp,
FUNCTION: PycRef BuildFromCode(PycRef code, PycModule* mod)
LINE: 1219
REASON: empty stack_hist cause stack_hist.top() core.

I have some code like this:

if x < 2:
    foo()
else:
    bar()

Decompiling the .pyc file from that gives me

if x < 2:
    foo()
continue
bar()

which is obviously incorrect. This is using Python 3.5.

Hello, can I help look after PYC file protection to deal with it? Can e-mail if you can contact me root # linglin.org, please replace the # @

gmake
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c data.cpp -o out/data.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c bytecode.cpp -o out/bytecode.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c pyc_module.cpp -o out/pyc_module.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c pyc_object.cpp -o out/pyc_object.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c pyc_numeric.cpp -o out/pyc_numeric.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c pyc_code.cpp -o out/pyc_code.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c pyc_sequence.cpp -o out/pyc_sequence.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c pyc_string.cpp -o out/pyc_string.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c ASTree.cpp -o out/ASTree.o
g++ -g -Wall -fprofile-arcs -ftest-coverage -DBLOCK_DEBUG -DSTACK_DEBUG -c ASTNode.cpp -o out/ASTNode.o
( cd bytes ; ./comp_map.py )
File "./comp_map.py", line 26
with infile and outfile:
^
SyntaxError: invalid syntax
gmake: *** [bytes/python_10.cpp] Error 1

When I try to build using MinGW's make, I get the following command line output - it's not possible to build:

g++  -g -Wall -Wextra -Werror -c data.cpp -o out/data.o
g++  -g -Wall -Wextra -Werror -c bytecode.cpp -o out/bytecode.o
g++  -g -Wall -Wextra -Werror -c pyc_module.cpp -o out/pyc_module.o
g++  -g -Wall -Wextra -Werror -c pyc_object.cpp -o out/pyc_object.o
g++  -g -Wall -Wextra -Werror -c pyc_numeric.cpp -o out/pyc_numeric.o
g++  -g -Wall -Wextra -Werror -c pyc_code.cpp -o out/pyc_code.o
g++  -g -Wall -Wextra -Werror -c pyc_sequence.cpp -o out/pyc_sequence.o
g++  -g -Wall -Wextra -Werror -c pyc_string.cpp -o out/pyc_string.o
g++  -g -Wall -Wextra -Werror -c ASTree.cpp -o out/ASTree.o
g++  -g -Wall -Wextra -Werror -c ASTNode.cpp -o out/ASTNode.o
( cd bytes ; ./comp_map.py )
Traceback (most recent call last):
  File "./comp_map.py", line 33, in <module>
    outfile.write('/* This file was auto-generated with comp_map.py.  DO NOT EDIT! */\n\n')
TypeError: 'str' does not support the buffer interface
make: *** [bytes/python_10.cpp] Error 1

`int main(int argc, char* argv[])
{
if (argc < 2) {
fprintf(stderr, "No input file specified\n");
return 1;
}

PycModule mod;
mod.loadFromFile(argv[1]);
if (!mod.isValid()) {
    fprintf(stderr, "Could not load file %s\n", argv[1]);
    return 1;
}
const char* dispname = strrchr(argv[1], PATHSEP);
dispname = (dispname == NULL) ? argv[1] : dispname + 1;
fprintf(pyc_output, "# Source Generated with Decompyle++\n");
fprintf(pyc_output, "# File: %s (Python %d.%d%s)\n\n", dispname, mod.majorVer(), mod.minorVer(),
        (mod.majorVer() < 3 && mod.isUnicode()) ? " Unicode" : "");
decompyle(mod.code(), &mod);

return 0;

}
`

Input:

c = x and y

output:

if x:
    pass
c = y

VC++ 2008 Express Edition

1>Compiling...
1>python_10.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_10.cpp': No such file or directory
1>python_11.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_11.cpp': No such file or directory
1>python_13.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_13.cpp': No such file or directory
1>python_14.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_14.cpp': No such file or directory
1>python_15.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_15.cpp': No such file or directory
1>python_16.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_16.cpp': No such file or directory
1>python_20.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_20.cpp': No such file or directory
1>python_21.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_21.cpp': No such file or directory
1>python_22.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_22.cpp': No such file or directory
1>python_23.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_23.cpp': No such file or directory
1>python_24.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_24.cpp': No such file or directory
1>python_25.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_25.cpp': No such file or directory
1>python_26.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_26.cpp': No such file or directory
1>python_27.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_27.cpp': No such file or directory
1>python_30.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_30.cpp': No such file or directory
1>python_31.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_31.cpp': No such file or directory
1>python_32.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_32.cpp': No such file or directory
1>python_33.cpp
1>c1xx : fatal error C1083: Cannot open source file: '.\bytes\python_33.cpp': No such file or directory
1>ASTNode.cpp
1>ASTree.cpp

is Project file outdated?

In my Original Code the Following Classes I have:

class BaseErrors(Exception):
    pass


class FilePathNotProvided(BaseErrors):
    pass


class FileNameNotProvided(BaseErrors):
    pass


class FileNotFound(BaseErrors):
    pass

Becomes this in pycdc:

class BaseErrors(Exception):
    __qualname__ = 'BaseErrors'


class FilePathNotProvided(BaseErrors):
    __qualname__ = 'FilePathNotProvided'


class FileNameNotProvided(BaseErrors):
    __qualname__ = 'FileNameNotProvided'


class FileNotFound(BaseErrors):
    __qualname__ = 'FileNotFound'

But in a Function in my Code:

def this_is_not_a_function_to_keep_this_coment_in_byte_codes():
    """
    I would like for the data returned from this to be importable without  having to cache (only if 
    sys.dont_write_bytecode is True) and without having to generate the normal py file as well.
    Why do this? what is the point?
    The Point for this is to make a interface to make python scripts smaller.
    With a range of 1/3 ~ 1/2 (results vary from file to file) the original sizes this is a dam good api for that.
    the cool thing is it uses python modules that comes with it. No External libs. (yet)
    Sadly this means somehow either recodign the original import statement in probably the C code to python itself.
    or finding how it reads from the pyc files (hopefully from within a python script).
    __init__.py:
        Normal Size: 598 bytes
        Compressed Size: 488 bytes
    api.py (this file):
        Normal Size: 6753 bytes
        Compressed Size: 3071 bytes
    compresspy.py:
        Normal Size: 1152 bytes
        Compressed Size: 746 bytes
    decompresspy.py:
        Normal Size: 1014 bytes
        Compressed Size: 688 bytes
    """
    pass

Becomes this in pycdc:

def this_is_not_a_function_to_keep_this_coment_in_byte_codes():
    '''
    I would like for the data returned from this to be importable without  having to cache (only if 
    sys.dont_write_bytecode is True) and without having to generate the normal py file as well.
    Why do this? what is the point?
    The Point for this is to make a interface to make python scripts smaller.
    With a range of 1/3 ~ 1/2 (results vary from file to file) the original sizes this is a dam good api for that.
    the cool thing is it uses python modules that comes with it. No External libs. (yet)
    Sadly this means somehow either recodign the original import statement in probably the C code to python itself.
    or finding how it reads from the pyc files (hopefully from within a python script).
    __init__.py:
        Normal Size: 598 bytes
        Compressed Size: 488 bytes
    api.py (this file):
        Normal Size: 6753 bytes
        Compressed Size: 3071 bytes
    compresspy.py:
        Normal Size: 1152 bytes
        Compressed Size: 746 bytes
    decompresspy.py:
        Normal Size: 1014 bytes
        Compressed Size: 688 bytes
    '''
    pass

My question is composed by 2 aspects

1. incompleted pyc is valid in python
   for example. I have tested the following a.py and b.py. After running a.py, we can find that b.pyc is created. But obviously, b.py is not a completed code since there is no defination of function "fib". So the bytecode of I-called incompleted pyc is valid too
2. we can use pycdc to disassemble. But is there a way to find just the code of function (for example "fun") and convert it back into pyc? thanks

[code a.py]
import b
[/code]

[code b.py]
n=10

def fun(x, y):
return x+fib(y)
[/code]

Nice job on the Decompiler just a few bugs where the else: statements for certain if blocks get removed for some reason. btw you guys also know of a way to append some interface to normal import statements for any modules imported in the interpreter? basically a overload to the normal one that would allow loading files with a custom compression without havign to use byecodes. funny thing is with my special format it could replace bytecode files to and is 100% easy to decompile (hehe) it is cool tbh.

py2pycx.zip

pycdc.exe!BuildFromCode(PycRef code={...}, PycModule * mod=0x0040fd9c)
case Pyc::END_FINALLY:
curblock->blktype() == ASTBlock::BLK_EXCEPT
blocks.pop(); blocks.size() == 0
behind
blocks.top() assert error
file：https://www.dropbox.com/s/b76j1bsvs6ph3j5/loglevels.pyc
Software may have logic problems,src file: https://www.dropbox.com/s/cmbtl8huyck3ynb/loglevels.py
Try this code：
def exception_to_unicode(e):
if e:
return "\n".join(a for a in e.args)
try:
return unicode(e)
except Exception:
return u"Unknown message"

I'm finding compiled source with defined functions is crashing the decompiler for Python 3.4.x.

Example Code:

def func_name():
print ("yelloo")
return False

print ('Hello, World!')
func_name()

Note that the decompiler won't even work unless you implement mancoast's fix for the Magic Number here: #49

the line in ASTree.cpp fails:
if (strcmp(code_src->name()->value(), "") == 0) {
because code_src->name() is empty. in fact the whole structure is filled with nulls. Something isn't parsing right.

Hello All!

I have been working on updates for 3.4.0.
Issue is simple to fix and easy to replicate:
1.) make a simple Hello.py script with contents:
print ('Hello, World!')

2.) make compile script Compile.py with contents:
import py_compile
py_compile.compile('Hello.py')

3.) execute "python Compile.py" to produce the output file Hello.cpython-34.pyc

4.) now execute "pycdc Hello.cpython-34.pyc"

Fix:

Results:

I am not sure if the magic number MAGIC_3_4 = 0x0A0D0CE4 is for a different 3.4?

Thanks,
mancoast

Got a segfault when trying to decompile a particular .pyc. A quick inspection by building with debug symbols and firing up gdb shows it's in the function PycRef::operator=.

    Stopped reason: SIGSEGV

0x000000000044db54 in PycRef<ASTBlock>::operator= (this=0x7fffffff5cc0, obj=...)
    at /home/dtouch3d/src/pycdc/pyc_object.h:39
39              if (obj.m_obj)
gdb-peda$ bt
#0  0x000000000044db54 in PycRef<ASTBlock>::operator= (this=0x7fffffff5cc0, obj=...)
    at /home/dtouch3d/src/pycdc/pyc_object.h:39
#1  0x0000000000436b8f in BuildFromCode (code=..., mod=0x7fffffffe600)
    at /home/dtouch3d/src/pycdc/ASTree.cpp:1243
#2  0x00000000004497ff in decompyle (code=..., mod=0x7fffffffe600)
    at /home/dtouch3d/src/pycdc/ASTree.cpp:2798
#3  0x0000000000445861 in print_src (node=..., mod=0x7fffffffe600)
    at /home/dtouch3d/src/pycdc/ASTree.cpp:2452
#4  0x000000000044731d in print_src (node=..., mod=0x7fffffffe600)
    at /home/dtouch3d/src/pycdc/ASTree.cpp:2656
#5  0x0000000000444f42 in print_src (node=..., mod=0x7fffffffe600)
    at /home/dtouch3d/src/pycdc/ASTree.cpp:2386
#6  0x000000000044a2e3 in decompyle (code=..., mod=0x7fffffffe600)
...
gdb-peda$ p obj
$1 = (const PycRef<ASTBlock> &) @0xdd00000200: <error reading variable>
gdb-peda$ 

I am guessing it is has something to do with the FastStack implementation, although I did not have much time to investigate further.

The offending file is snapconnect/SnapXmlRpc.pyc at http://forums.synapse-wireless.com/upload/snapconnect-3.1.0-python2.7.zip

Thank you.

There is the error message:

$ /opt/pycdc/pycdc __pycache__/stupid_test.cpython-27-PYTEST.pyc 
CreateObject: Got unsupported type 0x48
# Source Generated with Decompyle++
# File: stupid_test.cpython-27-PYTEST.pyc (Python 2.7)

Segmentation fault

Any idea on what is causing it ?

Here's the pyc file: encode.pyc

I renamed it to txt so GitHub could accept it. Rename it to pyc.

Decompiling generates an unused import random.

Good afternoon, Sir.

I find out a crash problem when used Pycdc to decompile one of the Pyc files and I want to know what's the matter .

The fellow picture shows the stack information when it crashed.

Download here: http://tool.duowan.com/wotbox/client/other/battleresults.pyc
if you need the file.

Looking forward to your reply.Thank you!

<255>

Cannot decompile for Python 3.3. It fails with the above message.