Track recent xdis changes

This commit is contained in:
rocky 2023-08-26 14:39:42 -04:00
parent 20c58e2e2a
commit 803678e9b4
3 changed files with 19 additions and 12 deletions

View File

@ -48,8 +48,7 @@ class Scanner26(scan.Scanner2):
return
def ingest(self, co, classname=None, code_objects={}, show_asm=None):
"""
Create "tokens" the bytecode of an Python code object. Largely these
"""Create "tokens" the bytecode of an Python code object. Largely these
are the opcode name, but in some cases that has been modified to make parsing
easier.
returning a list of uncompyle6 Token's.
@ -57,14 +56,17 @@ class Scanner26(scan.Scanner2):
Some transformations are made to assist the deparsing grammar:
- various types of LOAD_CONST's are categorized in terms of what they load
- COME_FROM instructions are added to assist parsing control structures
- operands with stack argument counts or flag masks are appended to the opcode name, e.g.:
- operands with stack argument counts or flag masks are appended to the
opcode name, e.g.:
* BUILD_LIST, BUILD_SET
* MAKE_FUNCTION and FUNCTION_CALLS append the number of positional arguments
* MAKE_FUNCTION and FUNCTION_CALLS append the number of positional
arguments
- EXTENDED_ARGS instructions are removed
Also, when we encounter certain tokens, we add them to a set which will cause custom
grammar rules. Specifically, variable arg tokens like MAKE_FUNCTION or BUILD_LIST
cause specific rules for the specific number of arguments they take.
Also, when we encounter certain tokens, we add them to a set
which will cause custom grammar rules. Specifically, variable
arg tokens like MAKE_FUNCTION or BUILD_LIST cause specific
rules for the specific number of arguments they take.
"""
if not show_asm:

View File

@ -221,7 +221,7 @@ class Scanner37Base(Scanner):
if show_asm in ("both", "before"):
print("\n# ---- before tokenization:")
bytecode.disassemble_bytes(
self.insts = bytecode.disassemble_bytes(
co.co_code,
varnames=co.co_varnames,
names=co.co_names,
@ -229,6 +229,9 @@ class Scanner37Base(Scanner):
cells=bytecode._cell_names,
linestarts=bytecode._linestarts,
asm_format="extended",
filename=co.co_filename,
show_source=True,
first_line_number=co.co_firstlineno,
)
# "customize" is in the process of going away here
@ -302,6 +305,8 @@ class Scanner37Base(Scanner):
inst.starts_line,
inst.is_jump_target,
inst.has_extended_arg,
None,
None,
)
# Get jump targets
@ -348,9 +353,9 @@ class Scanner37Base(Scanner):
j = tokens_append(
j,
Token(
come_from_name,
jump_offset,
repr(jump_offset),
opname=come_from_name,
attr=jump_offset,
pattr=repr(jump_offset),
offset="%s_%s" % (inst.offset, jump_idx),
has_arg=True,
opc=self.opc,

View File

@ -3,7 +3,7 @@
# More could be done here though.
from math import copysign
from xdis.codetype import UnicodeForPython3
from xdis.cross_types import UnicodeForPython3
from xdis.version_info import PYTHON_VERSION_TRIPLE
def get_code_name(code) -> str: