Skip to content

Commit daf54d2

Browse files
rockyrocky
committed
2.6 scanner show -A headers now
1 parent bf59e3c commit daf54d2

File tree

1 file changed

+76
-45
lines changed

1 file changed

+76
-45
lines changed

‎uncompyle6/scanners/scanner26.py‎

Lines changed: 76 additions & 45 deletions
Original file line numberDiff line numberDiff line change
@@ -23,27 +23,31 @@
2323
"""
2424

2525
import sys
26-
import uncompyle6.scanners.scanner2 as scan
2726

2827
# bytecode verification, verify(), uses JUMP_OPs from here
2928
from xdis import iscode
30-
from xdis.opcodes import opcode_26
3129
from xdis.bytecode import _get_const_info
30+
from xdis.opcodes import opcode_26
3231

32+
import uncompyle6.scanners.scanner2 as scan
3333
from uncompyle6.scanner import Token
3434

3535
intern = sys.intern
3636

3737
JUMP_OPS = opcode_26.JUMP_OPS
3838

39+
3940
class Scanner26(scan.Scanner2):
4041
def __init__(self, show_asm=False):
4142
super(Scanner26, self).__init__((2, 6), show_asm)
4243

4344
# "setup" opcodes
44-
self.setup_ops = frozenset([
45-
self.opc.SETUP_EXCEPT, self.opc.SETUP_FINALLY,
46-
])
45+
self.setup_ops = frozenset(
46+
[
47+
self.opc.SETUP_EXCEPT,
48+
self.opc.SETUP_FINALLY,
49+
]
50+
)
4751

4852
return
4953

@@ -76,8 +80,9 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
7680

7781
# show_asm = 'after'
7882
if show_asm in ("both", "before"):
83+
print("\n# ---- before tokenization:")
7984
for instr in bytecode.get_instructions(co):
80-
print(instr.disassemble())
85+
print(instr.disassemble(self.opc))
8186

8287
# Container for tokens
8388
tokens = []
@@ -96,17 +101,18 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
96101
# 'LOAD_ASSERT' is used in assert statements.
97102
self.load_asserts = set()
98103
for i in self.op_range(0, codelen):
99-
100104
# We need to detect the difference between:
101105
# raise AssertionError
102106
# and
103107
# assert ...
104-
if (self.code[i] == self.opc.JUMP_IF_TRUE and
105-
i + 4 < codelen and
106-
self.code[i+3] == self.opc.POP_TOP and
107-
self.code[i+4] == self.opc.LOAD_GLOBAL):
108-
if names[self.get_argument(i+4)] == 'AssertionError':
109-
self.load_asserts.add(i+4)
108+
if (
109+
self.code[i] == self.opc.JUMP_IF_TRUE
110+
and i + 4 < codelen
111+
and self.code[i + 3] == self.opc.POP_TOP
112+
and self.code[i + 4] == self.opc.LOAD_GLOBAL
113+
):
114+
if names[self.get_argument(i + 4)] == "AssertionError":
115+
self.load_asserts.add(i + 4)
110116

111117
jump_targets = self.find_jump_targets(show_asm)
112118
# contains (code, [addrRefToCode])
@@ -131,7 +137,8 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
131137
i += 1
132138
op = self.code[offset]
133139
op_name = self.opname[op]
134-
oparg = None; pattr = None
140+
oparg = None
141+
pattr = None
135142

136143
if offset in jump_targets:
137144
jump_idx = 0
@@ -142,28 +149,37 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
142149
# properly. For example, a "loop" with an "if" nested in it should have the
143150
# "loop" tag last so the grammar rule matches that properly.
144151
last_jump_offset = -1
145-
for jump_offset in sorted(jump_targets[offset], reverse=True):
152+
for jump_offset in sorted(jump_targets[offset], reverse=True):
146153
if jump_offset != last_jump_offset:
147-
tokens.append(Token(
148-
'COME_FROM', jump_offset, repr(jump_offset),
149-
offset="%s_%d" % (offset, jump_idx),
150-
has_arg = True))
154+
tokens.append(
155+
Token(
156+
"COME_FROM",
157+
jump_offset,
158+
repr(jump_offset),
159+
offset="%s_%d" % (offset, jump_idx),
160+
has_arg=True,
161+
)
162+
)
151163
jump_idx += 1
152164
last_jump_offset = jump_offset
153165
elif offset in self.thens:
154-
tokens.append(Token(
155-
'THEN', None, self.thens[offset],
156-
offset="%s_0" % offset,
157-
has_arg = True))
166+
tokens.append(
167+
Token(
168+
"THEN",
169+
None,
170+
self.thens[offset],
171+
offset="%s_0" % offset,
172+
has_arg=True,
173+
)
174+
)
158175

159-
has_arg = (op >= self.opc.HAVE_ARGUMENT)
176+
has_arg = op >= self.opc.HAVE_ARGUMENT
160177
if has_arg:
161178
oparg = self.get_argument(offset) + extended_arg
162179
extended_arg = 0
163180
if op == self.opc.EXTENDED_ARG:
164-
extended_arg += self.extended_arg_val(oparg)
165-
continue
166-
181+
extended_arg += self.extended_arg_val(oparg)
182+
continue
167183

168184
# Note: name used to match on rather than op since
169185
# BUILD_SET isn't in earlier Pythons.
@@ -172,7 +188,14 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
172188
"BUILD_SET",
173189
):
174190
t = Token(
175-
op_name, oparg, pattr, offset, self.linestarts.get(offset, None), op, has_arg, self.opc
191+
op_name,
192+
oparg,
193+
pattr,
194+
offset,
195+
self.linestarts.get(offset, None),
196+
op,
197+
has_arg,
198+
self.opc,
176199
)
177200

178201
collection_type = op_name.split("_")[1]
@@ -221,8 +244,8 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
221244
# FIXME: this is a hack to catch stuff like:
222245
# if x: continue
223246
# the "continue" is not on a new line.
224-
if len(tokens) and tokens[-1].kind == 'JUMP_BACK':
225-
tokens[-1].kind = intern('CONTINUE')
247+
if len(tokens) and tokens[-1].kind == "JUMP_BACK":
248+
tokens[-1].kind = intern("CONTINUE")
226249

227250
elif op in self.opc.JABS_OPS:
228251
pattr = repr(oparg)
@@ -240,17 +263,23 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
240263
# CE - Hack for >= 2.5
241264
# Now all values loaded via LOAD_CLOSURE are packed into
242265
# a tuple before calling MAKE_CLOSURE.
243-
if (self.version >= (2, 5) and op == self.opc.BUILD_TUPLE and
244-
self.code[self.prev[offset]] == self.opc.LOAD_CLOSURE):
266+
if (
267+
self.version >= (2, 5)
268+
and op == self.opc.BUILD_TUPLE
269+
and self.code[self.prev[offset]] == self.opc.LOAD_CLOSURE
270+
):
245271
continue
246272
else:
247-
op_name = '%s_%d' % (op_name, oparg)
273+
op_name = "%s_%d" % (op_name, oparg)
248274
customize[op_name] = oparg
249275
elif self.version > (2, 0) and op == self.opc.CONTINUE_LOOP:
250276
customize[op_name] = 0
251-
elif op_name in """
277+
elif (
278+
op_name
279+
in """
252280
CONTINUE_LOOP EXEC_STMT LOAD_LISTCOMP LOAD_SETCOMP
253-
""".split():
281+
""".split()
282+
):
254283
customize[op_name] = 0
255284
elif op == self.opc.JUMP_ABSOLUTE:
256285
# Further classify JUMP_ABSOLUTE into backward jumps
@@ -266,23 +295,24 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
266295
# rule for that.
267296
target = self.get_target(offset)
268297
if target <= offset:
269-
op_name = 'JUMP_BACK'
270-
if (offset in self.stmts
271-
and self.code[offset+3] not in (self.opc.END_FINALLY,
272-
self.opc.POP_BLOCK)):
273-
if ((offset in self.linestarts and
274-
tokens[-1].kind == 'JUMP_BACK')
275-
or offset not in self.not_continue):
276-
op_name = 'CONTINUE'
298+
op_name = "JUMP_BACK"
299+
if offset in self.stmts and self.code[offset + 3] not in (
300+
self.opc.END_FINALLY,
301+
self.opc.POP_BLOCK,
302+
):
303+
if (
304+
offset in self.linestarts and tokens[-1].kind == "JUMP_BACK"
305+
) or offset not in self.not_continue:
306+
op_name = "CONTINUE"
277307
else:
278308
# FIXME: this is a hack to catch stuff like:
279309
# if x: continue
280310
# the "continue" is not on a new line.
281-
if tokens[-1].kind == 'JUMP_BACK':
311+
if tokens[-1].kind == "JUMP_BACK":
282312
# We need 'intern' since we have
283313
# already have processed the previous
284314
# token.
285-
tokens[-1].kind = intern('CONTINUE')
315+
tokens[-1].kind = intern("CONTINUE")
286316

287317
elif op == self.opc.LOAD_GLOBAL:
288318
if offset in self.load_asserts:
@@ -316,6 +346,7 @@ def ingest(self, co, classname=None, code_objects={}, show_asm=None):
316346
pass
317347

318348
if show_asm in ("both", "after"):
349+
print("\n# ---- after tokenization:")
319350
for t in tokens:
320351
print(t.format(line_prefix=""))
321352
print()

0 commit comments

Comments
 (0)