jinx/python-uncompyle6
1
Fork 0
mirror of https://github.com/rocky/python-uncompyle6.git synced 2025-08-02 16:44:46 +08:00
Code Issues Packages Projects Releases Wiki Activity

Lint stuff

Browse Source
This commit is contained in:
rocky
2015-12-11 18:49:12 -05:00
parent b3c8cbb83f
commit 39842cef1f
9 changed files with 711 additions and 78 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
Makefile
View File

@@ -60,7 +60,7 @@ sdist:
#: Style check. Set env var LINT to pyflakes, flake, or flake8
lint:
$(LINT) trepan_deparse/deparser.py
$(LINT) uncompyle6
#: Create binary egg distribution
bdist_egg:

7
uncompyle6/Makefile Normal file
View File

@@ -0,0 +1,7 @@
# Whatever it is you want to do, it should be forwarded to the
# to top-level irectories
PHONY=check all
all: check
%:
$(MAKE) -C .. $@

20
uncompyle6/__init__.py
View File

@@ -43,7 +43,7 @@ def _load_file(filename):
load a Python source file and compile it to byte-code
_load_module(filename: string): code_object
filename: name of file containing Python source code
(normally a .py)
(normally a .py)
code_object: code_object compiled from this source code
This function does NOT write any file!
'''
@@ -62,7 +62,7 @@ def _load_module(filename):
load a module without importing it
_load_module(filename: string): code_object
filename: name of file containing Python byte-code object
(normally a .pyc)
(normally a .pyc)
code_object: code_object from this file
'''
@@ -74,7 +74,7 @@ def _load_module(filename):
raise ImportError("Unknown magic number %s in %s" % (ord(magic[0])+256*ord(magic[1]), filename))
if (version > 2.7) or (version < 2.5):
raise ImportError("This is a Python %s file! Only Python 2.5 to 2.7 files are supported." % version)
#print version
# print version
fp.read(4) # timestamp
co = dis.marshalLoad(fp)
fp.close()
@@ -85,7 +85,7 @@ def uncompyle(version, co, out=None, showasm=0, showast=0):
diassembles a given code block 'co'
'''
assert type(co) == types.CodeType
assert isinstance(co, types.CodeType)
# store final output stream for case of error
__real_out = out or sys.stdout
@@ -121,7 +121,7 @@ def uncompyle(version, co, out=None, showasm=0, showast=0):
del ast[0]
if ast[-1] == walker.RETURN_NONE:
ast.pop() # remove last node
#todo: if empty, add 'pass'
# todo: if empty, add 'pass'
except:
pass
walk.mod_globs = walker.find_globals(ast, set())
@@ -137,8 +137,8 @@ def uncompyle_file(filename, outstream=None, showasm=0, showast=0):
"""
version, co = _load_module(filename)
if type(co) == list:
for con in co:
uncompyle(version, con, outstream, showasm, showast)
for con in co:
uncompyle(version, con, outstream, showasm, showast)
else:
uncompyle(version, co, outstream, showasm, showast)
co = None
@@ -182,7 +182,7 @@ def main(in_base, out_base, files, codes, outfile=None,
of = outfile
tot_files = okay_files = failed_files = verify_failed_files = 0
#for code in codes:
# for code in codes:
# version = sys.version[:3] # "2.5"
# with open(code, "r") as f:
# co = compile(f.read(), "", "exec")
@@ -190,7 +190,7 @@ def main(in_base, out_base, files, codes, outfile=None,
for file in files:
infile = os.path.join(in_base, file)
#print >>sys.stderr, infile
# print >>sys.stderr, infile
if of: # outfile was given as parameter
outstream = _get_outstream(outfile)
@@ -199,7 +199,7 @@ def main(in_base, out_base, files, codes, outfile=None,
else:
outfile = os.path.join(out_base, file) + '_dis'
outstream = _get_outstream(outfile)
#print >>sys.stderr, outfile
# print >>sys.stderr, outfile
# try to decomyple the input file
try:

32
uncompyle6/disas.py
View File

@@ -37,9 +37,8 @@ def dis(x=None):
elif isinstance(x, str):
disassemble_string(x)
else:
raise TypeError, \
"don't know how to disassemble %s objects" % \
type(x).__name__
raise (TypeError,
"don't know how to disassemble %s objects" % type(x).__name__)
def distb(tb=None):
"""Disassemble a traceback (default: last traceback)."""
@@ -47,7 +46,7 @@ def distb(tb=None):
try:
tb = sys.last_traceback
except AttributeError:
raise RuntimeError, "no last traceback to disassemble"
raise(RuntimeError, "no last traceback to disassemble")
while tb.tb_next: tb = tb.tb_next
disassemble(tb.tb_frame.f_code, tb.tb_lasti)
@@ -123,7 +122,7 @@ def disassemble_string(code, lasti=-1, varnames=None, names=None,
if constants:
print '(' + repr(constants[oparg]) + ')',
else:
print '(%d)'%oparg,
print '(%d)' % oparg,
elif op in hasname:
if names is not None:
print '(' + names[oparg] + ')',
@@ -140,7 +139,7 @@ def disassemble_string(code, lasti=-1, varnames=None, names=None,
print '(' + cmp_op[oparg] + ')',
print
disco = disassemble
disco = disassemble
# XXX For backwards compatibility
def findlabels(code):
@@ -191,17 +190,17 @@ def marshalLoad(fp):
global internStrings
internStrings = []
return load(fp)
def load(fp):
"""
Load marshal
Load marshal
"""
global internStrings
marshalType = fp.read(1)
if marshalType == 'c':
Code = types.CodeType
co_argcount = unpack('i', fp.read(4))[0]
co_nlocals = unpack('i', fp.read(4))[0]
co_stacksize = unpack('i', fp.read(4))[0]
@@ -216,12 +215,13 @@ def load(fp):
co_name = load(fp)
co_firstlineno = unpack('i', fp.read(4))[0]
co_lnotab = load(fp)
return Code(co_argcount, co_nlocals, co_stacksize, co_flags, co_code, co_consts, co_names,\
co_varnames, co_filename, co_name, co_firstlineno, co_lnotab, co_freevars, co_cellvars)
return Code(co_argcount, co_nlocals, co_stacksize, co_flags, co_code,
co_consts, co_names, co_varnames, co_filename, co_name,
co_firstlineno, co_lnotab, co_freevars, co_cellvars)
# const type
elif marshalType == '.':
return Ellipsis
return Ellipsis
elif marshalType == '0':
raise KeyError, marshalType
return None
@@ -253,7 +253,7 @@ def load(fp):
n = unpack('i', fp.read(4))[0]
if n == 0:
return long(0)
size = abs(n);
size = abs(n);
d = long(0)
for j in range(0, size):
md = int(unpack('h', fp.read(2))[0])
@@ -296,7 +296,7 @@ def load(fp):
return None
else:
sys.stderr.write("Unkown type %i (hex %x)\n" % (ord(marshalType), ord(marshalType)))
def _test():
"""Simple test program to disassemble a file."""
if sys.argv[1:]:

43
uncompyle6/parser.py
View File

@@ -27,6 +27,7 @@ class AST(UserList):
UserList.__init__(self, kids)
def __getslice__(self, low, high): return self.data[low:high]
def __eq__(self, o):
if isinstance(o, AST):
return self.type == o.type \
@@ -147,7 +148,6 @@ class Parser(GenericASTBuilder):
genexpr_func ::= LOAD_FAST FOR_ITER designator comp_iter JUMP_BACK
'''
def p_dictcomp(self, args):
'''
expr ::= dictcomp
@@ -159,7 +159,6 @@ class Parser(GenericASTBuilder):
'''
def p_augmented_assign(self, args):
'''
stmt ::= augassign1
@@ -707,7 +706,7 @@ class Parser(GenericASTBuilder):
rv = args[0]
rv.append(args[1])
else:
rv = GenericASTBuilder.nonterminal(self, nt, args)
rv = GenericASTBuilder.nonterminal(self, nt, args)
return rv
def __ambiguity(self, children):
@@ -720,7 +719,7 @@ class Parser(GenericASTBuilder):
return 'funcdef'
if 'grammar' in list and 'expr' in list:
return 'expr'
#print >> sys.stderr, 'resolve', str(list)
# print >> sys.stderr, 'resolve', str(list)
return GenericASTBuilder.resolve(self, list)
nop = lambda self, args: None
@@ -728,22 +727,22 @@ nop = lambda self, args: None
p = Parser()
def parse(tokens, customize):
#
# Special handling for opcodes that take a variable number
# of arguments -- we add a new rule for each:
#
# expr ::= {expr}^n BUILD_LIST_n
# expr ::= {expr}^n BUILD_TUPLE_n
# unpack_list ::= UNPACK_LIST {expr}^n
# unpack ::= UNPACK_TUPLE {expr}^n
# unpack ::= UNPACK_SEQEUENE {expr}^n
# mkfunc ::= {expr}^n LOAD_CONST MAKE_FUNCTION_n
# mkfunc ::= {expr}^n load_closure LOAD_CONST MAKE_FUNCTION_n
# expr ::= expr {expr}^n CALL_FUNCTION_n
# expr ::= expr {expr}^n CALL_FUNCTION_VAR_n POP_TOP
# expr ::= expr {expr}^n CALL_FUNCTION_VAR_KW_n POP_TOP
# expr ::= expr {expr}^n CALL_FUNCTION_KW_n POP_TOP
#
'''
Special handling for opcodes that take a variable number
of arguments -- we add a new rule for each:
expr ::= {expr}^n BUILD_LIST_n
expr ::= {expr}^n BUILD_TUPLE_n
unpack_list ::= UNPACK_LIST {expr}^n
unpack ::= UNPACK_TUPLE {expr}^n
unpack ::= UNPACK_SEQEUENE {expr}^n
mkfunc ::= {expr}^n LOAD_CONST MAKE_FUNCTION_n
mkfunc ::= {expr}^n load_closure LOAD_CONST MAKE_FUNCTION_n
expr ::= expr {expr}^n CALL_FUNCTION_n
expr ::= expr {expr}^n CALL_FUNCTION_VAR_n POP_TOP
expr ::= expr {expr}^n CALL_FUNCTION_VAR_KW_n POP_TOP
expr ::= expr {expr}^n CALL_FUNCTION_KW_n POP_TOP
'''
global p
for k, v in list(customize.items()):
# avoid adding the same rule twice to this parser
@@ -751,7 +750,7 @@ def parse(tokens, customize):
continue
p.customized[k] = None
#nop = lambda self, args: None
# nop = lambda self, args: None
op = k[:string.rfind(k, '_')]
if op in ('BUILD_LIST', 'BUILD_TUPLE', 'BUILD_SET'):
rule = 'build_list ::= ' + 'expr '*v + k
@@ -762,7 +761,7 @@ def parse(tokens, customize):
elif op in ('DUP_TOPX', 'RAISE_VARARGS'):
# no need to add a rule
continue
#rule = 'dup_topx ::= ' + 'expr '*v + k
# rule = 'dup_topx ::= ' + 'expr '*v + k
elif op == 'MAKE_FUNCTION':
p.addRule('mklambda ::= %s LOAD_LAMBDA %s' %
('expr '*v, k), nop)

9
uncompyle6/scanner26.py
View File

@@ -42,6 +42,7 @@ class Scanner26(scan.Scanner):
# class and names
if classname:
classname = '_' + classname.lstrip('_') + '__'
def unmangle(name):
if name.startswith(classname) and name[-2:] != '__':
return name[len(classname) - 2:]
@@ -150,7 +151,7 @@ class Scanner26(scan.Scanner):
# verify uses 'pattr' for comparism, since 'attr'
# now holds Code(const) and thus can not be used
# for comparism (todo: think about changing this)
#pattr = 'code_object @ 0x%x %s->%s' %\
# pattr = 'code_object @ 0x%x %s->%s' %\
# (id(const), const.co_filename, const.co_name)
pattr = '<code_object ' + const.co_name + '>'
else:
@@ -248,7 +249,7 @@ class Scanner26(scan.Scanner):
if destFor == i+opsize+4:
setupLoop = self.last_instr(0, jmpabs1target, SETUP_LOOP)
standarFor = self.last_instr(setupLoop, jmpabs1target, GET_ITER)
if standarFor == None:
if standarFor is None:
self.restructJump(jmpabs1target, destFor+self.op_size(POP_BLOCK))
toDel += [setupLoop, i+opsize+1, i+opsize+4]
@@ -280,7 +281,7 @@ class Scanner26(scan.Scanner):
# log JA/POP_TOP to del and update PJIF
while start < end:
start = self.first_instr(start, end, (PJIF,PJIT))
if start == None: break
if start is None: break
target = self.get_target(start)
if self.code[target] == POP_TOP and self.code[target-3] == JA:
toDel += [target, target-3]
@@ -433,7 +434,7 @@ class Scanner26(scan.Scanner):
listDel = []
for i in self.op_range(0, len(self.code)):
ret = self.getOpcodeToDel(i)
if ret != None:
if ret is not None:
listDel += ret
# change code structure after deleting byte

26
uncompyle6/scanner27.py
View File

@@ -139,8 +139,8 @@ class Scanner27(scan.Scanner):
# verify uses 'pattr' for comparism, since 'attr'
# now holds Code(const) and thus can not be used
# for comparism (todo: think about changing this)
#pattr = 'code_object @ 0x%x %s->%s' %\
# (id(const), const.co_filename, const.co_name)
# pattr = 'code_object @ 0x%x %s->%s' %\
# (id(const), const.co_filename, const.co_name)
pattr = '<code_object ' + const.co_name + '>'
else:
pattr = const
@@ -209,7 +209,7 @@ class Scanner27(scan.Scanner):
def build_stmt_indices(self):
code = self.code
start = 0;
start = 0
end = len(code)
stmt_opcodes = {
@@ -269,7 +269,7 @@ class Scanner27(scan.Scanner):
j = self.prev[s]
while code[j] == JA:
j = self.prev[j]
if code[j] == LIST_APPEND: #list comprehension
if code[j] == LIST_APPEND: # list comprehension
stmts.remove(s)
continue
elif code[s] == POP_TOP and code[self.prev[s]] == ROT_TWO:
@@ -336,7 +336,7 @@ class Scanner27(scan.Scanner):
if op is None:
op = code[pos]
## Detect parent structure
# Detect parent structure
parent = self.structs[0]
start = parent['start']
end = parent['end']
@@ -347,7 +347,7 @@ class Scanner27(scan.Scanner):
start = _start
end = _end
parent = s
## We need to know how many new structures were added in this run
# We need to know how many new structures were added in this run
origStructCount = len(self.structs)
if op == SETUP_LOOP:
@@ -416,15 +416,15 @@ class Scanner27(scan.Scanner):
end = self.restrict_to_parent(target, parent)
if target != end:
self.fixed_jumps[pos] = end
#print target, end, parent
## Add the try block
# print target, end, parent
# Add the try block
self.structs.append({'type': 'try',
'start': start,
'end': end-4})
## Now isolate the except and else blocks
# Now isolate the except and else blocks
end_else = start_else = self.get_target(self.prev[end])
## Add the except blocks
# Add the except blocks
i = end
while self.code[i] != END_FINALLY:
jmp = self.next_except_jump(i)
@@ -593,8 +593,8 @@ class Scanner27(scan.Scanner):
self.structs = [{'type': 'root',
'start': 0,
'end': n-1}]
self.loops = [] ## All loop entry points
self.fixed_jumps = {} ## Map fixed jumps to their real destination
self.loops = [] # All loop entry points
self.fixed_jumps = {} # Map fixed jumps to their real destination
self.ignore_if = set()
self.build_stmt_indices()
self.not_continue = set()
@@ -604,7 +604,7 @@ class Scanner27(scan.Scanner):
for i in self.op_range(0, n):
op = code[i]
## Determine structures and fix jumps for 2.3+
# Determine structures and fix jumps for 2.3+
self.detect_structure(i, op)
if op >= HAVE_ARGUMENT:

628
uncompyle6/scanner34.py Executable file
View File

@@ -0,0 +1,628 @@
from __future__ import print_function
'''
Copyright (c) 1999 John Aycock
Copyright (c) 2000-2002 by hartmut Goebel <h.goebel@crazy-compilers.com>
Copyright (c) 2005 by Dan Pascu <dan@windowmaker.org>
See main module for license.
'''
import types
from collections import namedtuple
from array import array
from operator import itemgetter
from uncompyle6.opcodes.opcode_27 import *
import disas as dis
import scanner as scan
class Scanner27(scan.Scanner):
def __init__(self):
self.Token = scan.Scanner.__init__(self, 2.7) # check
def disassemble(self, co, classname=None):
'''
Disassemble a code object, returning a list of 'Token'.
The main part of this procedure is modelled after
dis.disassemble().
'''
rv = []
customize = {}
Token = self.Token # shortcut
self.code = array('B', co.co_code)
for i in self.op_range(0, len(self.code)):
if self.code[i] in (RETURN_VALUE, END_FINALLY):
n = i + 1
self.code = array('B', co.co_code[:n])
self.prev = [0]
# mapping adresses of instru & arg
for i in self.op_range(0, n):
op = self.code[i]
self.prev.append(i)
if op >= HAVE_ARGUMENT:
self.prev.append(i)
self.prev.append(i)
self.lines = []
linetuple = namedtuple('linetuple', ['l_no', 'next'])
j = 0
# linestarts contains bloc code adresse (addr,block)
linestarts = list(dis.findlinestarts(co))
linestartoffsets = {a for (a, _) in linestarts}
(prev_start_byte, prev_line_no) = linestarts[0]
for (start_byte, line_no) in linestarts[1:]:
while j < start_byte:
self.lines.append(linetuple(prev_line_no, start_byte))
j += 1
last_op = self.code[self.prev[start_byte]]
(prev_start_byte, prev_line_no) = (start_byte, line_no)
while j < n:
self.lines.append(linetuple(prev_line_no, n))
j+=1
# self.lines contains (block,addrLastInstr)
if classname:
classname = '_' + classname.lstrip('_') + '__'
def unmangle(name):
if name.startswith(classname) and name[-2:] != '__':
return name[len(classname) - 2:]
return name
free = [ unmangle(name) for name in (co.co_cellvars + co.co_freevars) ]
names = [ unmangle(name) for name in co.co_names ]
varnames = [ unmangle(name) for name in co.co_varnames ]
else:
free = co.co_cellvars + co.co_freevars
names = co.co_names
varnames = co.co_varnames
self.load_asserts = set()
for i in self.op_range(0, n):
if self.code[i] == PJIT and self.code[i+3] == LOAD_GLOBAL:
if names[self.get_argument(i+3)] == 'AssertionError':
self.load_asserts.add(i+3)
cf = self.find_jump_targets(self.code)
# contains (code, [addrRefToCode])
last_stmt = self.next_stmt[0]
i = self.next_stmt[last_stmt]
replace = {}
while i < n-1:
if self.lines[last_stmt].next > i:
if self.code[last_stmt] == PRINT_ITEM:
if self.code[i] == PRINT_ITEM:
replace[i] = 'PRINT_ITEM_CONT'
elif self.code[i] == PRINT_NEWLINE:
replace[i] = 'PRINT_NEWLINE_CONT'
last_stmt = i
i = self.next_stmt[i]
imports = self.all_instr(0, n, (IMPORT_NAME, IMPORT_FROM, IMPORT_STAR))
if len(imports) > 1:
last_import = imports[0]
for i in imports[1:]:
if self.lines[last_import].next > i:
if self.code[last_import] == IMPORT_NAME == self.code[i]:
replace[i] = 'IMPORT_NAME_CONT'
last_import = i
extended_arg = 0
for offset in self.op_range(0, n):
if offset in cf:
k = 0
for j in cf[offset]:
rv.append(Token('COME_FROM', None, repr(j),
offset="%s_%d" % (offset, k)))
k += 1
op = self.code[offset]
op_name = opname[op]
oparg = None; pattr = None
if op >= HAVE_ARGUMENT:
oparg = self.get_argument(offset) + extended_arg
extended_arg = 0
if op == EXTENDED_ARG:
extended_arg = oparg * 65536
continue
if op in hasconst:
const = co.co_consts[oparg]
if type(const) == types.CodeType:
oparg = const
if const.co_name == '<lambda>':
assert op_name == 'LOAD_CONST'
op_name = 'LOAD_LAMBDA'
elif const.co_name == '<genexpr>':
op_name = 'LOAD_GENEXPR'
elif const.co_name == '<dictcomp>':
op_name = 'LOAD_DICTCOMP'
elif const.co_name == '<setcomp>':
op_name = 'LOAD_SETCOMP'
# verify uses 'pattr' for comparism, since 'attr'
# now holds Code(const) and thus can not be used
# for comparism (todo: think about changing this)
#pattr = 'code_object @ 0x%x %s->%s' %\
# (id(const), const.co_filename, const.co_name)
pattr = '<code_object ' + const.co_name + '>'
else:
pattr = const
elif op in hasname:
pattr = names[oparg]
elif op in hasjrel:
pattr = repr(offset + 3 + oparg)
elif op in hasjabs:
pattr = repr(oparg)
elif op in haslocal:
pattr = varnames[oparg]
elif op in hascompare:
pattr = cmp_op[oparg]
elif op in hasfree:
pattr = free[oparg]
if op in (BUILD_LIST, BUILD_TUPLE, BUILD_SET, BUILD_SLICE,
UNPACK_SEQUENCE,
MAKE_FUNCTION, CALL_FUNCTION, MAKE_CLOSURE,
CALL_FUNCTION_VAR, CALL_FUNCTION_KW,
CALL_FUNCTION_VAR_KW, DUP_TOPX, RAISE_VARARGS
):
# CE - Hack for >= 2.5
# Now all values loaded via LOAD_CLOSURE are packed into
# a tuple before calling MAKE_CLOSURE.
if op == BUILD_TUPLE and \
self.code[self.prev[offset]] == LOAD_CLOSURE:
continue
else:
op_name = '%s_%d' % (op_name, oparg)
if op != BUILD_SLICE:
customize[op_name] = oparg
elif op == JA:
target = self.get_target(offset)
if target < offset:
if offset in self.stmts and self.code[offset+3] not in (END_FINALLY, POP_BLOCK) \
and offset not in self.not_continue:
op_name = 'CONTINUE'
else:
op_name = 'JUMP_BACK'
elif op == LOAD_GLOBAL:
if offset in self.load_asserts:
op_name = 'LOAD_ASSERT'
elif op == RETURN_VALUE:
if offset in self.return_end_ifs:
op_name = 'RETURN_END_IF'
if offset not in replace:
rv.append(Token(op_name, oparg, pattr, offset, linestart = offset in linestartoffsets))
else:
rv.append(Token(replace[offset], oparg, pattr, offset, linestart = offset in linestartoffsets))
if self.showasm:
out = self.out # shortcut
for t in rv:
print(t, file=out)
print(file=out)
return rv, customize
def op_size(self, op):
if op < self.opc.HAVE_ARGUMENT:
return 1
else:
return 3
def build_stmt_indices(self):
code = self.code
start = 0;
end = len(code)
stmt_opcodes = {
SETUP_LOOP, BREAK_LOOP, CONTINUE_LOOP,
SETUP_FINALLY, END_FINALLY, SETUP_EXCEPT, SETUP_WITH,
POP_BLOCK, STORE_FAST, DELETE_FAST, STORE_DEREF,
STORE_GLOBAL, DELETE_GLOBAL, STORE_NAME, DELETE_NAME,
STORE_ATTR, DELETE_ATTR, STORE_SUBSCR, DELETE_SUBSCR,
RETURN_VALUE, RAISE_VARARGS, POP_TOP,
PRINT_EXPR, PRINT_ITEM, PRINT_NEWLINE, PRINT_ITEM_TO, PRINT_NEWLINE_TO,
STORE_SLICE_0, STORE_SLICE_1, STORE_SLICE_2, STORE_SLICE_3,
DELETE_SLICE_0, DELETE_SLICE_1, DELETE_SLICE_2, DELETE_SLICE_3,
JUMP_ABSOLUTE, EXEC_STMT,
}
stmt_opcode_seqs = [(PJIF, JF), (PJIF, JA), (PJIT, JF), (PJIT, JA)]
designator_ops = {
STORE_FAST, STORE_NAME, STORE_GLOBAL, STORE_DEREF, STORE_ATTR,
STORE_SLICE_0, STORE_SLICE_1, STORE_SLICE_2, STORE_SLICE_3,
STORE_SUBSCR, UNPACK_SEQUENCE, JA
}
prelim = self.all_instr(start, end, stmt_opcodes)
stmts = self.stmts = set(prelim)
pass_stmts = set()
for seq in stmt_opcode_seqs:
for i in self.op_range(start, end-(len(seq)+1)):
match = True
for elem in seq:
if elem != code[i]:
match = False
break
i += self.op_size(code[i])
if match:
i = self.prev[i]
stmts.add(i)
pass_stmts.add(i)
if pass_stmts:
stmt_list = list(stmts)
stmt_list.sort()
else:
stmt_list = prelim
last_stmt = -1
self.next_stmt = []
slist = self.next_stmt = []
i = 0
for s in stmt_list:
if code[s] == JA and s not in pass_stmts:
target = self.get_target(s)
if target > s or self.lines[last_stmt].l_no == self.lines[s].l_no:
stmts.remove(s)
continue
j = self.prev[s]
while code[j] == JA:
j = self.prev[j]
if code[j] == LIST_APPEND: #list comprehension
stmts.remove(s)
continue
elif code[s] == POP_TOP and code[self.prev[s]] == ROT_TWO:
stmts.remove(s)
continue
elif code[s] in designator_ops:
j = self.prev[s]
while code[j] in designator_ops:
j = self.prev[j]
if code[j] == FOR_ITER:
stmts.remove(s)
continue
last_stmt = s
slist += [s] * (s-i)
i = s
slist += [end] * (end-len(slist))
def remove_mid_line_ifs(self, ifs):
filtered = []
for i in ifs:
if self.lines[i].l_no == self.lines[i+3].l_no:
if self.code[self.prev[self.lines[i].next]] in (PJIT, PJIF):
continue
filtered.append(i)
return filtered
def next_except_jump(self, start):
'''
Return the next jump that was generated by an except SomeException:
construct in a try...except...else clause or None if not found.
'''
if self.code[start] == DUP_TOP:
except_match = self.first_instr(start, len(self.code), POP_JUMP_IF_FALSE)
if except_match:
jmp = self.prev[self.get_target(except_match)]
self.ignore_if.add(except_match)
self.not_continue.add(jmp)
return jmp
count_END_FINALLY = 0
count_SETUP_ = 0
for i in self.op_range(start, len(self.code)):
op = self.code[i]
if op == END_FINALLY:
if count_END_FINALLY == count_SETUP_:
assert self.code[self.prev[i]] in (JA, JF, RETURN_VALUE)
self.not_continue.add(self.prev[i])
return self.prev[i]
count_END_FINALLY += 1
elif op in (SETUP_EXCEPT, SETUP_WITH, SETUP_FINALLY):
count_SETUP_ += 1
def detect_structure(self, pos, op=None):
'''
Detect type of block structures and their boundaries to fix optimizied jumps
in python2.3+
'''
# TODO: check the struct boundaries more precisely -Dan
code = self.code
# Ev remove this test and make op a mandatory argument -Dan
if op is None:
op = code[pos]
## Detect parent structure
parent = self.structs[0]
start = parent['start']
end = parent['end']
for s in self.structs:
_start = s['start']
_end = s['end']
if (_start <= pos < _end) and (_start >= start and _end <= end):
start = _start
end = _end
parent = s
## We need to know how many new structures were added in this run
origStructCount = len(self.structs)
if op == SETUP_LOOP:
start = pos+3
target = self.get_target(pos, op)
end = self.restrict_to_parent(target, parent)
if target != end:
self.fixed_jumps[pos] = end
(line_no, next_line_byte) = self.lines[pos]
jump_back = self.last_instr(start, end, JA,
next_line_byte, False)
if jump_back and jump_back != self.prev[end] and code[jump_back+3] in (JA, JF):
if code[self.prev[end]] == RETURN_VALUE or \
(code[self.prev[end]] == POP_BLOCK and code[self.prev[self.prev[end]]] == RETURN_VALUE):
jump_back = None
if not jump_back: # loop suite ends in return. wtf right?
jump_back = self.last_instr(start, end, RETURN_VALUE) + 1
if not jump_back:
return
if code[self.prev[next_line_byte]] not in (PJIF, PJIT):
loop_type = 'for'
else:
loop_type = 'while'
self.ignore_if.add(self.prev[next_line_byte])
target = next_line_byte
end = jump_back + 3
else:
if self.get_target(jump_back) >= next_line_byte:
jump_back = self.last_instr(start, end, JA, start, False)
if end > jump_back+4 and code[end] in (JF, JA):
if code[jump_back+4] in (JA, JF):
if self.get_target(jump_back+4) == self.get_target(end):
self.fixed_jumps[pos] = jump_back+4
end = jump_back+4
elif target < pos:
self.fixed_jumps[pos] = jump_back+4
end = jump_back+4
target = self.get_target(jump_back, JA)
if code[target] in (FOR_ITER, GET_ITER):
loop_type = 'for'
else:
loop_type = 'while'
test = self.prev[next_line_byte]
if test == pos:
loop_type = 'while 1'
elif self.code[test] in hasjabs+hasjrel:
self.ignore_if.add(test)
test_target = self.get_target(test)
if test_target > (jump_back+3):
jump_back = test_target
self.not_continue.add(jump_back)
self.loops.append(target)
self.structs.append({'type': loop_type + '-loop',
'start': target,
'end': jump_back})
if jump_back+3 != end:
self.structs.append({'type': loop_type + '-else',
'start': jump_back+3,
'end': end})
elif op == SETUP_EXCEPT:
start = pos+3
target = self.get_target(pos, op)
end = self.restrict_to_parent(target, parent)
if target != end:
self.fixed_jumps[pos] = end
#print target, end, parent
## Add the try block
self.structs.append({'type': 'try',
'start': start,
'end': end-4})
## Now isolate the except and else blocks
end_else = start_else = self.get_target(self.prev[end])
## Add the except blocks
i = end
while self.code[i] != END_FINALLY:
jmp = self.next_except_jump(i)
if self.code[jmp] == RETURN_VALUE:
self.structs.append({'type': 'except',
'start': i,
'end': jmp+1})
i = jmp + 1
else:
if self.get_target(jmp) != start_else:
end_else = self.get_target(jmp)
if self.code[jmp] == JF:
self.fixed_jumps[jmp] = -1
self.structs.append({'type': 'except',
'start': i,
'end': jmp})
i = jmp + 3
## Add the try-else block
if end_else != start_else:
r_end_else = self.restrict_to_parent(end_else, parent)
self.structs.append({'type': 'try-else',
'start': i+1,
'end': r_end_else})
self.fixed_jumps[i] = r_end_else
else:
self.fixed_jumps[i] = i+1
elif op in (PJIF, PJIT):
start = pos+3
target = self.get_target(pos, op)
rtarget = self.restrict_to_parent(target, parent)
pre = self.prev
if target != rtarget and parent['type'] == 'and/or':
self.fixed_jumps[pos] = rtarget
return
#does this jump to right after another cond jump?
# if so, it's part of a larger conditional
if (code[pre[target]] in (JUMP_IF_FALSE_OR_POP, JUMP_IF_TRUE_OR_POP,
PJIF, PJIT)) and (target > pos):
self.fixed_jumps[pos] = pre[target]
self.structs.append({'type': 'and/or',
'start': start,
'end': pre[target]})
return
# is this an if and
if op == PJIF:
match = self.rem_or(start, self.next_stmt[pos], PJIF, target)
match = self.remove_mid_line_ifs(match)
if match:
if code[pre[rtarget]] in (JF, JA) \
and pre[rtarget] not in self.stmts \
and self.restrict_to_parent(self.get_target(pre[rtarget]), parent) == rtarget:
if code[pre[pre[rtarget]]] == JA \
and self.remove_mid_line_ifs([pos]) \
and target == self.get_target(pre[pre[rtarget]]) \
and (pre[pre[rtarget]] not in self.stmts or self.get_target(pre[pre[rtarget]]) > pre[pre[rtarget]])\
and 1 == len(self.remove_mid_line_ifs(self.rem_or(start, pre[pre[rtarget]], (PJIF, PJIT), target))):
pass
elif code[pre[pre[rtarget]]] == RETURN_VALUE \
and self.remove_mid_line_ifs([pos]) \
and 1 == (len(set(self.remove_mid_line_ifs(self.rem_or(start, pre[pre[rtarget]], \
(PJIF, PJIT), target))) \
| set(self.remove_mid_line_ifs(self.rem_or(start, pre[pre[rtarget]], \
(PJIF, PJIT, JA), pre[rtarget], True))))):
pass
else:
fix = None
jump_ifs = self.all_instr(start, self.next_stmt[pos], PJIF)
last_jump_good = True
for j in jump_ifs:
if target == self.get_target(j):
if self.lines[j].next == j+3 and last_jump_good:
fix = j
break
else:
last_jump_good = False
self.fixed_jumps[pos] = fix or match[-1]
return
else:
self.fixed_jumps[pos] = match[-1]
return
else: # op == PJIT
if (pos+3) in self.load_asserts:
if code[pre[rtarget]] == RAISE_VARARGS:
return
self.load_asserts.remove(pos+3)
next = self.next_stmt[pos]
if pre[next] == pos:
pass
elif code[next] in (JF, JA) and target == self.get_target(next):
if code[pre[next]] == PJIF:
if code[next] == JF or target != rtarget or code[pre[pre[rtarget]]] not in (JA, RETURN_VALUE):
self.fixed_jumps[pos] = pre[next]
return
elif code[next] == JA and code[target] in (JA, JF):
next_target = self.get_target(next)
if self.get_target(target) == next_target:
self.fixed_jumps[pos] = pre[next]
return
elif code[next_target] in (JA, JF) and self.get_target(next_target) == self.get_target(target):
self.fixed_jumps[pos] = pre[next]
return
# don't add a struct for a while test, it's already taken care of
if pos in self.ignore_if:
return
if code[pre[rtarget]] == JA and pre[rtarget] in self.stmts \
and pre[rtarget] != pos and pre[pre[rtarget]] != pos:
if code[rtarget] == JA and code[rtarget+3] == POP_BLOCK:
if code[pre[pre[rtarget]]] != JA:
pass
elif self.get_target(pre[pre[rtarget]]) != target:
pass
else:
rtarget = pre[rtarget]
else:
rtarget = pre[rtarget]
# does the if jump just beyond a jump op, then this is probably an if statement
if code[pre[rtarget]] in (JA, JF):
if_end = self.get_target(pre[rtarget])
# is this a loop not an if?
if (if_end < pre[rtarget]) and (code[pre[if_end]] == SETUP_LOOP):
if(if_end > start):
return
end = self.restrict_to_parent(if_end, parent)
self.structs.append({'type': 'if-then',
'start': start,
'end': pre[rtarget]})
self.not_continue.add(pre[rtarget])
if rtarget < end:
self.structs.append({'type': 'if-else',
'start': rtarget,
'end': end})
elif code[pre[rtarget]] == RETURN_VALUE:
self.structs.append({'type': 'if-then',
'start': start,
'end': rtarget})
self.return_end_ifs.add(pre[rtarget])
elif op in (JUMP_IF_FALSE_OR_POP, JUMP_IF_TRUE_OR_POP):
target = self.get_target(pos, op)
self.fixed_jumps[pos] = self.restrict_to_parent(target, parent)
def find_jump_targets(self, code):
'''
Detect all offsets in a byte code which are jump targets.
Return the list of offsets.
This procedure is modelled after dis.findlables(), but here
for each target the number of jumps are counted.
'''
n = len(code)
self.structs = [{'type': 'root',
'start': 0,
'end': n-1}]
self.loops = [] # All loop entry points
self.fixed_jumps = {} # Map fixed jumps to their real destination
self.ignore_if = set()
self.build_stmt_indices()
self.not_continue = set()
self.return_end_ifs = set()
targets = {}
for i in self.op_range(0, n):
op = code[i]
# Determine structures and fix jumps for 2.3+
self.detect_structure(i, op)
if op >= HAVE_ARGUMENT:
label = self.fixed_jumps.get(i)
oparg = code[i+1] + code[i+2] * 256
if label is None:
if op in hasjrel and op != FOR_ITER:
label = i + 3 + oparg
elif op in hasjabs:
if op in (JUMP_IF_FALSE_OR_POP, JUMP_IF_TRUE_OR_POP):
if (oparg > i):
label = oparg
if label is not None and label != -1:
targets[label] = targets.get(label, []) + [i]
elif op == END_FINALLY and i in self.fixed_jumps:
label = self.fixed_jumps[i]
targets[label] = targets.get(label, []) + [i]
return targets

22
uncompyle6/walker.py
View File

@@ -1282,7 +1282,7 @@ class Walker(GenericASTTraversal, object):
assert type(code) == CodeType
code = Code(code, self.scanner, self.currentclass)
#assert isinstance(code, Code)
# assert isinstance(code, Code)
# add defaults values to parameter names
argc = code.co_argcount
@@ -1303,12 +1303,12 @@ class Walker(GenericASTTraversal, object):
# build parameters
##This would be a nicer piece of code, but I can't get this to work
## now, have to find a usable lambda constuct hG/2000-09-05
##params = map(lambda name, default: build_param(ast, name, default),
## paramnames, defparams)
# This would be a nicer piece of code, but I can't get this to work
# now, have to find a usable lambda constuct hG/2000-09-05
# params = map(lambda name, default: build_param(ast, name, default),
# paramnames, defparams)
params = []
for name, default in map(lambda a,b: (a,b), paramnames, defparams):
for name, default in map(lambda a, b: (a, b), paramnames, defparams):
params.append( build_param(ast, name, default) )
params.reverse() # back to correct order
@@ -1332,7 +1332,6 @@ class Walker(GenericASTTraversal, object):
# docstring exists, dump it
self.print_docstring(indent, code.co_consts[0])
code._tokens = None # save memory
assert ast == 'stmts'
#if isLambda:
@@ -1373,26 +1372,25 @@ class Walker(GenericASTTraversal, object):
del ast[0]
# if docstring exists, dump it
if code.co_consts and code.co_consts[0] != None and ast[0][0] == ASSIGN_DOC_STRING(code.co_consts[0]):
if (code.co_consts and code.co_consts[0] is not None
and ast[0][0] == ASSIGN_DOC_STRING(code.co_consts[0])):
self.print_docstring(indent, code.co_consts[0])
self.print_()
del ast[0]
# the function defining a class normally returns locals(); we
# don't want this to show up in the source, thus remove the node
if ast[-1][0] == RETURN_LOCALS:
del ast[-1] # remove last node
#else:
# else:
# print ast[-1][-1]
for g in find_globals(ast, set()):
self.print_(indent, 'global ', g)
self.print_(indent, 'global ', g)
self.gen_source(ast, code._customize)
code._tokens = None; code._customize = None # save memory
def gen_source(self, ast, customize, isLambda=0, returnNone=False):
"""convert AST to source code"""