jinx/python-uncompyle6
1
Fork 0
mirror of https://github.com/rocky/python-uncompyle6.git synced 2025-08-03 00:45:53 +08:00
Code Issues Packages Projects Releases Wiki Activity

Partial sync with decompyle3

Browse Source
This commit is contained in:
rocky
2024-02-05 16:57:59 -05:00
parent 33bc80bb24
commit f605f859ae
1 changed files with 70 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files

107
uncompyle6/semantics/pysource.py
View File

@@ -131,6 +131,7 @@ Python.
import sys import sys
from io import StringIO from io import StringIO
from typing import Optional
from spark_parser import GenericASTTraversal from spark_parser import GenericASTTraversal
from xdis import COMPILER_FLAG_BIT, iscode from xdis import COMPILER_FLAG_BIT, iscode
@@ -159,7 +160,11 @@ from uncompyle6.semantics.consts import (
) )
from uncompyle6.semantics.customize import customize_for_version from uncompyle6.semantics.customize import customize_for_version
from uncompyle6.semantics.gencomp import ComprehensionMixin from uncompyle6.semantics.gencomp import ComprehensionMixin
from uncompyle6.semantics.helper import find_globals_and_nonlocals, print_docstring from uncompyle6.semantics.helper import (
find_globals_and_nonlocals,
is_lambda_mode,
print_docstring,
)
from uncompyle6.semantics.make_function1 import make_function1 from uncompyle6.semantics.make_function1 import make_function1
from uncompyle6.semantics.make_function2 import make_function2 from uncompyle6.semantics.make_function2 import make_function2
from uncompyle6.semantics.make_function3 import make_function3 from uncompyle6.semantics.make_function3 import make_function3
@@ -213,7 +218,7 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
def __init__( def __init__(
self, self,
version, version: tuple,
out, out,
scanner, scanner,
showast=TREE_DEFAULT_DEBUG, showast=TREE_DEFAULT_DEBUG,
@@ -223,7 +228,7 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
linestarts={}, linestarts={},
tolerate_errors=False, tolerate_errors=False,
): ):
"""`version' is the Python version (a float) of the Python dialect """`version' is the Python version of the Python dialect
of both the syntax tree and language we should produce. of both the syntax tree and language we should produce.
`out' is IO-like file pointer to where the output should go. It `out' is IO-like file pointer to where the output should go. It
@@ -235,9 +240,12 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
If `showast' is True, we print the syntax tree. If `showast' is True, we print the syntax tree.
`compile_mode' is is either 'exec' or 'single'. It is the compile `compile_mode` is is either `exec`, `single` or `lambda`.
mode that was used to create the Syntax Tree and specifies a
grammar variant within a Python version to use. For `lambda`, the grammar that can be used in lambda
expressions is used. Otherwise, it is the compile mode that
was used to create the Syntax Tree and specifies a grammar
variant within a Python version to use.
`is_pypy` should be True if the Syntax Tree was generated for PyPy. `is_pypy` should be True if the Syntax Tree was generated for PyPy.
@@ -262,10 +270,8 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
self.currentclass = None self.currentclass = None
self.classes = [] self.classes = []
self.debug_parser = dict(debug_parser) self.debug_parser = dict(debug_parser)
# Initialize p_lambda on demand
self.line_number = 1 self.line_number = 1
self.linemap = {} self.linemap = {}
self.p_lambda = None
self.params = params self.params = params
self.param_stack = [] self.param_stack = []
self.ERROR = None self.ERROR = None
@@ -276,11 +282,15 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
self.pending_newlines = 0 self.pending_newlines = 0
self.linestarts = linestarts self.linestarts = linestarts
self.treeTransform = TreeTransform(version=self.version, show_ast=showast) self.treeTransform = TreeTransform(version=self.version, show_ast=showast)
# FIXME: have p.insts update in a better way # FIXME: have p.insts update in a better way
# modularity is broken here # modularity is broken here
self.insts = scanner.insts self.insts = scanner.insts
self.offset2inst_index = scanner.offset2inst_index self.offset2inst_index = scanner.offset2inst_index
# Initialize p_lambda on demand
self.p_lambda = None
# This is in Python 2.6 on. It changes the way # This is in Python 2.6 on. It changes the way
# strings get interpreted. See n_LOAD_CONST # strings get interpreted. See n_LOAD_CONST
self.FUTURE_UNICODE_LITERALS = False self.FUTURE_UNICODE_LITERALS = False
@@ -507,19 +517,19 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
def pp_tuple(self, tup): def pp_tuple(self, tup):
"""Pretty print a tuple""" """Pretty print a tuple"""
last_line = self.f.getvalue().split("\n")[-1] last_line = self.f.getvalue().split("\n")[-1]
l = len(last_line) + 1 ll = len(last_line) + 1
indent = " " * l indent = " " * ll
self.write("(") self.write("(")
sep = "" sep = ""
for item in tup: for item in tup:
self.write(sep) self.write(sep)
l += len(sep) ll += len(sep)
s = better_repr(item, self.version) s = better_repr(item, self.version)
l += len(s) ll += len(s)
self.write(s) self.write(s)
sep = "," sep = ","
if l > LINE_LENGTH: if ll > LINE_LENGTH:
l = 0 ll = 0
sep += "\n" + indent sep += "\n" + indent
else: else:
sep += " " sep += " "
@@ -699,9 +709,10 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
""" """
# print("-----") # print("-----")
# print(startnode) # print(startnode.kind)
# print(entry[0]) # print(entry[0])
# print('======') # print('======')
fmt = entry[0] fmt = entry[0]
arg = 1 arg = 1
i = 0 i = 0
@@ -794,13 +805,9 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
node[index].kind, node[index].kind,
) )
else: else:
assert ( assert node[tup[0]] in tup[1], (
node[tup[0]] in tup[1] f"at {node.kind}[{tup[0]}], expected to be in '{tup[1]}' "
), "at %s[%d], expected to be in '%s' node; got '%s'" % ( f"node; got '{node[tup[0]].kind}'"
node.kind,
arg,
index[1],
node[index[0]].kind,
) )
else: else:
@@ -869,7 +876,7 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
d = node.__dict__ d = node.__dict__
try: try:
self.write(eval(expr, d, d)) self.write(eval(expr, d, d))
except: except Exception:
raise raise
m = escape.search(fmt, i) m = escape.search(fmt, i)
self.write(fmt[i:]) self.write(fmt[i:])
@@ -1190,7 +1197,7 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
is_lambda=False, is_lambda=False,
noneInNames=False, noneInNames=False,
is_top_level_module=False, is_top_level_module=False,
): ) -> GenericASTTraversal:
# FIXME: DRY with fragments.py # FIXME: DRY with fragments.py
# assert isinstance(tokens[0], Token) # assert isinstance(tokens[0], Token)
@@ -1242,7 +1249,7 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
# Build a parse tree from a tokenized and massaged disassembly. # Build a parse tree from a tokenized and massaged disassembly.
try: try:
# FIXME: have p.insts update in a better way # FIXME: have p.insts update in a better way
# modularity is broken here # Modularity is broken here.
p_insts = self.p.insts p_insts = self.p.insts
self.p.insts = self.scanner.insts self.p.insts = self.scanner.insts
self.p.offset2inst_index = self.scanner.offset2inst_index self.p.offset2inst_index = self.scanner.offset2inst_index
@@ -1255,6 +1262,7 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
checker(ast, False, self.ast_errors) checker(ast, False, self.ast_errors)
self.customize(customize) self.customize(customize)
transform_tree = self.treeTransform.transform(ast, code) transform_tree = self.treeTransform.transform(ast, code)
self.maybe_show_tree(ast, phase="before") self.maybe_show_tree(ast, phase="before")
@@ -1270,13 +1278,15 @@ class SourceWalker(GenericASTTraversal, NonterminalActions, ComprehensionMixin):
def code_deparse( def code_deparse(
co, co,
out=sys.stdout, out=sys.stdout,
version=None, version: Optional[tuple] = None,
debug_opts=DEFAULT_DEBUG_OPTS, debug_opts=DEFAULT_DEBUG_OPTS,
code_objects={}, code_objects={},
compile_mode="exec", compile_mode="exec",
is_pypy=IS_PYPY, is_pypy=IS_PYPY,
walker=SourceWalker, walker=SourceWalker,
): start_offset: int = 0,
stop_offset: int = -1,
) -> Optional[SourceWalker]:
""" """
ingests and deparses a given code block 'co'. If version is None, ingests and deparses a given code block 'co'. If version is None,
we will use the current Python interpreter version. we will use the current Python interpreter version.
@@ -1284,6 +1294,9 @@ def code_deparse(
assert iscode(co) assert iscode(co)
if out is None:
out = sys.stdout
if version is None: if version is None:
version = PYTHON_VERSION_TRIPLE version = PYTHON_VERSION_TRIPLE
@@ -1294,6 +1307,21 @@ def code_deparse(
co, code_objects=code_objects, show_asm=debug_opts["asm"] co, code_objects=code_objects, show_asm=debug_opts["asm"]
) )
if start_offset > 0:
for i, t in enumerate(tokens):
# If t.offset is a string, we want to skip this.
if isinstance(t.offset, int) and t.offset >= start_offset:
tokens = tokens[i:]
break
if stop_offset > -1:
for i, t in enumerate(tokens):
# In contrast to the test for start_offset If t.offset is
# a string, we want to extract the integer offset value.
if t.off2int() >= stop_offset:
tokens = tokens[:i]
break
debug_parser = debug_opts.get("grammar", dict(PARSER_DEFAULT_DEBUG)) debug_parser = debug_opts.get("grammar", dict(PARSER_DEFAULT_DEBUG))
# Build Syntax Tree from disassembly. # Build Syntax Tree from disassembly.
@@ -1317,7 +1345,7 @@ def code_deparse(
tokens, tokens,
customize, customize,
co, co,
is_lambda=(compile_mode == "lambda"), is_lambda=is_lambda_mode(compile_mode),
is_top_level_module=is_top_level_module, is_top_level_module=is_top_level_module,
) )
@@ -1326,7 +1354,7 @@ def code_deparse(
return None return None
# FIXME use a lookup table here. # FIXME use a lookup table here.
if compile_mode == "lambda": if is_lambda_mode(compile_mode):
expected_start = "lambda_start" expected_start = "lambda_start"
elif compile_mode == "eval": elif compile_mode == "eval":
expected_start = "expr_start" expected_start = "expr_start"
@@ -1339,10 +1367,12 @@ def code_deparse(
expected_start = None expected_start = None
else: else:
expected_start = None expected_start = None
if expected_start: if expected_start:
assert ( assert deparsed.ast == expected_start, (
deparsed.ast == expected_start f"Should have parsed grammar start to '{expected_start}'; "
), f"Should have parsed grammar start to '{expected_start}'; got: {deparsed.ast.kind}" f"got: {deparsed.ast.kind}"
)
# save memory # save memory
del tokens del tokens
@@ -1382,7 +1412,7 @@ def code_deparse(
deparsed.ast, deparsed.ast,
name=co.co_name, name=co.co_name,
customize=customize, customize=customize,
is_lambda=compile_mode == "lambda", is_lambda=is_lambda_mode(compile_mode),
debug_opts=debug_opts, debug_opts=debug_opts,
) )
@@ -1410,9 +1440,12 @@ def deparse_code2str(
compile_mode="exec", compile_mode="exec",
is_pypy=IS_PYPY, is_pypy=IS_PYPY,
walker=SourceWalker, walker=SourceWalker,
): start_offset: int = 0,
"""Return the deparsed text for a Python code object. `out` is where any intermediate stop_offset: int = -1,
output for assembly or tree output will be sent. ) -> str:
"""
Return the deparsed text for a Python code object. `out` is where
any intermediate output for assembly or tree output will be sent.
""" """
return code_deparse( return code_deparse(
code, code,
@@ -1427,6 +1460,7 @@ def deparse_code2str(
if __name__ == "__main__": if __name__ == "__main__":
def deparse_test(co): def deparse_test(co):
"""This is a docstring""" """This is a docstring"""
s = deparse_code2str(co) s = deparse_code2str(co)
@@ -1434,5 +1468,4 @@ if __name__ == "__main__":
print(s) print(s)
return return
deparse_test(deparse_test.__code__) deparse_test(deparse_test.__code__)