jinx/python-uncompyle6
1
Fork 0
mirror of https://github.com/rocky/python-uncompyle6.git synced 2025-08-04 09:22:40 +08:00
Code Issues Packages Projects Releases Wiki Activity
793 Commits 10 Branches 160 Tags
50c2e1bda9acff29a19de6c3c06516c706f8a6ad
Commit Graph

75 Commits

Author SHA1 Message Date
rocky
50c2e1bda9 Revert augassign change but.. 
Make note of what's going on and add grammar test for bad
situations we have in Python 2.6 (and perhaps others)
 2016-11-11 09:08:02 -05:00
rocky
0f536b18fa Bug in detecting 3.3 default value in lambda 2016-11-10 23:59:51 -05:00
rocky
6fb879d0d8 Detect some erroneous decompilations 
Until we can actually prevent these in grammar rules, we will warn of
improper decompilations.

Also, we now stop when we hit a decompile error. Previously we were not.
 2016-11-10 22:29:39 -05:00
rocky
2328ca7a55 Break out make_function() into its own file. 
It is already too complex and will get worse in Python 3.6.

Note: make_function in fragments.py is still inside and
probably needs fixup.
 2016-10-29 07:22:58 -04:00
rocky
fe072d8b57 DRY fragments.py preorder code 
pysource.py: doc typo
 2016-10-11 22:24:00 -04:00
rocky
c7788e4545 disassemble -> ingest where appropriate 
As part of tokenization for (de)parsing, we need to do something like a
disassembly, but is is really a little different.

Disassembly, strictly speaking, is done by the xdis module now.
What "ingestion" does is massage the instruction tokens to a form that is
more amenable for parsing.

In sum, ingestion is different than disassembly, although disassembly is
generally the first part of ingestion.
 2016-09-04 11:43:02 -04:00
rocky
3ef0325cb8 Small changes 2016-07-29 19:54:16 -04:00
rocky
f33f425692 Add is_pypy parameter to places that need it 2016-07-27 16:08:40 -04:00
rocky
c3f6fa32db Customize tables better for specific Python versions 2016-07-27 06:56:02 -04:00
rocky
74f440bd0b Usuability fixes 
* try using format for __str__
* Explicitly nuke self.attr and self.pattr when no arg
* Sync pysource and format wrt make_function
 2016-07-26 20:49:41 -04:00
rocky
7c4316d4fb Start handling pypy 2.7 
Need to understand whether we care compiling pypy.
Pypy 2.7 list comprehensions are different and use
its own opcode.
 2016-07-21 02:58:50 -04:00
rocky
4e6e38358d 3.x list comprehenions list_if_not, comp_ifnot bug 
Saw only list_if_not bug, but might also be applicable
to comp_ifnot
 2016-07-14 11:30:42 -04:00
rocky
bc86b73cf0 Document/correct hide_internal and store_locals 2016-07-12 12:07:48 -04:00
rocky
694e1edd00 More offsets captrued Add %b specifer 
%b - associate text before specifier
pysource.py: small doc correction
 2016-07-03 21:57:46 -04:00
rocky
fa7d8f955a WIP Python-2.6 but don't remove opcodes 
The scheme for turning 2.6 bytecode into 2.7 psuedo bytecode
I think is a lose.

I won't work for fragment handling.

Instead, change the grammar and syntax rules

This also has the benefits:

* We see how code generation changed over releases
  by looking at grammar and semantic rules rather
  than arbitrary code
* We can better assocate with what's running
  (in a sense this is a restatement of broken fragment
   handling)
* With the right structure in place we are in a better position to
  handle 2.5, 2.4, etc. That is, after a while, the incremental changes
  to get say from python 2.3 bytecode to python 2.7 are great.

Conflicts:
	uncompyle6/parsers/astnode.py
 2016-06-24 18:15:59 -04:00
rocky
493ffa62fe Python 3.5 and setcomp_func's 
A comprehension walk in Python 3.5 needs to include
setcomp_func's.
 2016-06-22 22:21:54 -04:00
rocky
480e6a125f Add Python 3.5 yield from and ... 
* fragments.py: Handle pass stmt sometimes
* scanners: regularize Python 2 scanners some
* test/test_pyenvlib.py: add python 3.5.1 option
 2016-06-22 13:08:11 -04:00
rocky
faa630902d Python 3 comprehension fixes 
Sync up fragments.py
 2016-06-22 10:50:11 -04:00
rocky
75592795b3 Python 2.6 tolerance 2016-06-22 00:05:19 -04:00
rocky
649da8fbc0 Btter Python 3 fragment set comprehensions ... 
fragment handling for "break" and "continue"
 2016-06-21 23:40:26 -04:00
rocky
cdb02fa591 Handle fragment key/value offsets better 2016-06-21 04:08:36 -04:00
rocky
24d4cfb150 Python 3 set comprehension bug 2016-06-20 18:47:03 -04:00
rocky
8c374904f5 3.x make closure kw args handling bug 2016-06-20 10:01:18 -04:00
rocky
078f15013e Python 3.x class decorator bug 2016-06-20 00:32:00 -04:00
rocky
efb4012087 Correct kw+pos args semantics on 3.3 
Sync fragment make_function code
 2016-06-19 21:41:40 -04:00
rocky
5c268ee2a6 2.7 and 3.x bug in dict comprehensions 2016-06-19 13:45:07 -04:00
rocky
ff014a8393 Cover more offsets 2016-06-19 04:22:55 -04:00
rocky
d7b79c2b59 3.4 dictionary comprehension bug 
Sync up fragment code with recent changesa
 2016-06-19 00:49:22 -04:00
rocky
fe56ca96c2 Add fragment offsets for more instructions 
Save and restore class name across switches
 2016-06-07 23:21:01 -04:00
rocky
e9c8c11071 Set comprehension code is not in 2.6 
So we need more care in test programs.
 2016-06-06 18:06:44 -04:00
rocky
bdfe14069c Add class offset of Python3 2016-06-06 17:26:08 -04:00
rocky
70d4841a6a Fix python 3 set comprehension and ... 
Add a few set/list comprehension offsets for Python 3
 2016-06-06 17:04:59 -04:00
rocky
4f8714ff4c include offset for starting listcomp 2016-06-06 10:10:19 -04:00
Daniel Bradburn
b3182e804d showasm and showast now accept file like objects which are used for writing the the asm or ast to. 2016-06-01 10:02:04 +02:00
rocky
4398b5b2e0 Use xdis for code, magics, and marshal 2016-05-27 19:18:06 -04:00
Daniel Bradburn
ebee84228b Fixed bug in pysource / fragments where the use of deparse_code from fragments could break subsequent calls to deparse_code from pysource due to amending to the global TABLE_DIRECT. 2016-05-25 20:48:35 +02:00
rocky
79593d6ef1 Small change 
0 shouldn't be used as False.
 2016-05-24 12:38:44 -04:00
rocky
68ff878b3e Small pyflakes stuff 2016-05-20 21:37:46 -04:00
rocky
f834b46b84 Fragment fixes 
fragments.py:
 * Use "%x" specifier if for iterators
 * Add '%D' interpretation

pysource.py:
  TABLE_DIRECT can get messed up from running fragments
  duplicate "%x" specifier to igore fragment stuff
 2016-05-20 06:18:41 -04:00
rocky
9462e33f48 Fix Python 3.x bugs 
* class definitions made via closures
* Add "make check-short" to top-level
* parse3.py: Python 3.3 uses STORE_LOGALS
 2016-05-17 04:00:54 -04:00
rocky
09a01dc783 Correct bugs in Python 3.2 source generation 2016-05-16 15:50:45 -04:00
rocky
6be6632e96 Bug in Python 3.x genexpr 2016-05-15 19:35:45 -04:00
rocky
2fdcdcb154 Fix bug in kvlist handling on Python 2.7 
Created from previous DRY code in Python 3
 2016-05-15 03:47:30 -04:00
rocky
b9692c9b1f DRY scanner34 and scanner35 
handle 3.0..3.4 build maps as key/value pairs
 2016-05-15 03:29:08 -04:00
rocky
2db380a77f More fragment bugs 2016-05-13 23:55:27 -04:00
rocky
de0ec195b7 More packaging crap. 
Did I tell you how much I hate python packaging?

fragments.py: track recent change in class decorators.
 2016-05-13 23:35:31 -04:00
rocky
6f6f1db576 Misc fixups/cleanups 
* parse3.py Had botched if-for-else test by grammar addition
* semantics/*.py: Show errorstack in failed parse when -g (requires sparck 1.2.0)
* some optimization in scanner3
 2016-05-12 09:27:25 -04:00
rocky
631d7be921 Redo make_function for *, arg 
main(*, file='foo') and things like that now work
 2016-05-11 20:34:20 -04:00
rocky
41f9e9e53e Track recent lc changes in fragment semantics 2016-05-09 06:57:13 -04:00
rocky
debb46b0fe Semantic routine nonterminal typo 2016-05-08 13:49:15 -04:00