Commit Graph

471 Commits

Author SHA1 Message Date
Neal Norwitz 7bcabc60a3 Fix a few more memory leaks
Document more info about the benefits of configuring without
pymalloc when running valgrind
2005-11-20 23:58:38 +00:00
Neal Norwitz 4737b2348b Last batch of ref leaks in new AST code.
Also converted a bunch of assert(0) to SystemError's.

There are still printfs, etc that need to be cleaned up.
2005-11-19 23:58:29 +00:00
Nick Coghlan 944d3eb154 Correctly handle identifiers for anonymous scopes and align genexpr name with symtable.c 2005-11-16 12:46:55 +00:00
Neal Norwitz b6fc9df8fc Fix a lot of memory and ref leaks in error paths.
(Call symtable_exit_block or compiler_exit_scope as appropriate)

Use PyMem_Free on c_future since it was allocated with PyMem_Malloc
2005-11-13 18:50:34 +00:00
Neal Norwitz 2744c6cc35 make internal method static 2005-11-13 01:08:38 +00:00
Neil Schemenauer c396d9edd6 Ensure that compiler_exit_scope() is called as necessary to free memory
allocated by compiler_enter_scope().  Change return type for
compiler_exit_scope() to be void.
2005-10-25 06:30:14 +00:00
Neal Norwitz f1d50684c6 Fix problem handling EXTENDED_ARGs from SF bug # 1333982 2005-10-23 23:00:41 +00:00
Neal Norwitz 7d37f2ff40 cleanup a bit and reuse instrsize (instruction size). working towards fixing problems with EXTENDED_ARG 2005-10-23 22:40:47 +00:00
Neil Schemenauer dad06a159e Remove unnecessary local variable. 2005-10-23 18:52:36 +00:00
Neil Schemenauer 8b528b28f1 Fix private name mangling. The symtable also must do mangles so that
the scope of names can be correctly determined.
2005-10-23 18:37:42 +00:00
Neil Schemenauer 3a44aaa30f Use PyTuple_Pack instead of Py_BuildValue. 2005-10-23 17:21:54 +00:00
Neil Schemenauer d403c45386 Fix arigo's funky LOAD_NAME bug: implicit globals inside classes have
historically been looked up using LOAD_NAME, not LOAD_GLOBAL.
looked up by LOAD_NAME, not
2005-10-23 04:24:49 +00:00
Neil Schemenauer ac699efad8 Don't stop generating code for import statements after the first "import as"
part.  Fixes one bug from #1333982.
2005-10-23 03:45:42 +00:00
Neil Schemenauer ccd19218bc Use <lamba> as the function name for lambdas (matches old compiler). 2005-10-21 18:09:19 +00:00
Armin Rigo 3144130217 ANSI-C-ify the placement of local var declarations. 2005-10-21 12:57:31 +00:00
Jeremy Hylton 3e0055f8c6 Merge ast-branch to head
This change implements a new bytecode compiler, based on a
transformation of the parse tree to an abstract syntax defined in
Parser/Python.asdl.

The compiler implementation is not complete, but it is in stable
enough shape to run the entire test suite excepting two disabled
tests.
2005-10-20 19:59:25 +00:00
Tim Peters ea9dcdc062 com_yield_expr(): Squash new compiler wng about unreferenced local. 2005-08-03 18:33:05 +00:00
Phillip J. Eby 0d6615fd29 PEP 342 implementation. Per Guido's comments, the generator throw()
method still needs to support string exceptions, and allow None for the
third argument.  Documentation updates are needed, too.
2005-08-02 00:46:46 +00:00
Brett Cannon c9371d4a1b Fix signedness of various char variables to stop causing a warning under gcc 4. 2005-06-25 08:23:41 +00:00
Brett Cannon 409d8f2ebd Allow classes to be defined with empty parentheses. This means that
``class C(): pass`` is no longer a syntax error.
2005-03-05 06:47:57 +00:00
Brett Cannon 5dc8ced4a8 Silence a gcc warning about putting in parentheses around && expressions mixed
with || expressions.  Also clarifies intend of 'if' conditional.
2005-03-03 07:01:48 +00:00
Raymond Hettinger e63a078635 Preserve sign of -0.0 when result is run through marshal. 2005-02-23 13:37:55 +00:00
Raymond Hettinger 07359a7fed Document how the pattern recognizer keeps all of its references in bounds.
Add a test in case the underlying assumptions ever change (i.e. the
compiler starts generating code blocks that are not punctuated by
RETURN_VALUE).
2005-02-21 20:03:14 +00:00
Raymond Hettinger 80121491e0 Teach the peepholer to fold unary operations on constants.
Afterwards, -0.5 loads in a single step and no longer requires a runtime
UNARY_NEGATIVE operation.
2005-02-20 12:41:32 +00:00
Raymond Hettinger b615bf0681 Remove the set conversion which didn't work with: [] in (0,) 2005-02-10 01:42:32 +00:00
Raymond Hettinger 5e547969f8 Have set conversion replace existing constant if not used elsewhere. 2005-02-09 21:24:51 +00:00
Raymond Hettinger 7fcb7869ba Adopt Skip's idea to optimize lists of constants in the context
of a "in" or "not in" test.
2005-02-07 19:32:38 +00:00
Raymond Hettinger a164574937 Transform "x in (1,2,3)" to "x in frozenset([1,2,3])".
Inspired by Skip's idea to recognize the throw-away nature of sequences
in this context and to transform their type to one with better performance.
2005-02-06 22:05:42 +00:00
Raymond Hettinger 9feb267caf Do not fold a constant if a large sequence will result.
Saves space in the presence of code like: (None,)*10000
2005-01-26 12:50:05 +00:00
Armin Rigo 664b43b3f4 Re-running python with/without the -Qnew flag uses incorrectly optimized
bytecodes from the previously saved .pyc files.  Fixed by disabling the static
optimization of BINARY_DIVIDE between two constants.
2005-01-07 18:10:51 +00:00
Raymond Hettinger c34f8673a1 Teach the peephole optimizer to fold simple constant expressions. 2005-01-02 06:17:33 +00:00
Jeremy Hylton 16b047904c SF patch 1025636: Check for NULL returns in compile.c:com_import_stmt
There is no test for this change, because there is no way to provoke memory errors on demand.  Test suite passes, though.
2004-11-07 14:04:00 +00:00
Raymond Hettinger 5dec096e6a Maintain peepholer's cumlc invariant by updating the running total
everytime a LOAD_CONSTANT is encountered, created, or overwritten.

Added two tests to cover cases affected by the patch.
2004-11-02 04:20:10 +00:00
Raymond Hettinger 099ecfbec9 Simplify and future proof NOP counting in the peepholer.
No longer assumes that the input is NOP free.
2004-11-01 15:19:11 +00:00
Armin Rigo d7bcf4deb1 Fixed a comment and added another one. 2004-10-30 21:08:59 +00:00
Raymond Hettinger effb3931ea Adopt some peepholer suggestions from Armin Rigo:
* Use simpler, faster two pass algorithm for markblocks().
* Free the blocks variable if not NULL and exiting without change.
* Verify that the rest of the compiler has not set an exception.
* Make the test for tuple of constants less restrictive.
* Embellish the comment for chained conditional jumps.
2004-10-30 08:55:08 +00:00
Raymond Hettinger 23109ef11e SF bug #1053819: Segfault in tuple_of_constants
Peepholer could be fooled into misidentifying a tuple_of_constants.
Added code to count consecutive occurrences of LOAD_CONST.
Use the count to weed out the misidentified cases.
Added a unittest.
2004-10-26 08:59:14 +00:00
Raymond Hettinger 9047c8f73d SF bug #1048870: call arg of lambda not updating 2004-10-24 00:10:06 +00:00
Raymond Hettinger fec0c466ad Improve error message for augmented assignments to genexps or listcomps.
Rather than introduce new logic, took the approach of making the message
itself more general.
2004-09-29 23:54:08 +00:00
Raymond Hettinger 8ffc1410e2 Fix two erroneous error messages. 2004-09-29 21:47:10 +00:00
Raymond Hettinger 06cc973d06 Replaced a test with an assertion.
(Suggested by Michael Hudson.)
2004-09-28 17:22:12 +00:00
Raymond Hettinger 2c31a058eb SF patch #1031667: Fold tuples of constants into a single constant
Example:
>>> import dis
>>> dis.dis(compile('1,2,3', '', 'eval'))
  0           0 LOAD_CONST               3 ((1, 2, 3))
              3 RETURN_VALUE
2004-09-22 18:44:21 +00:00
Anthony Baxter 1a4ddaecc7 SF patch #1007189, multi-line imports, for instance:
"from blah import (foo, bar
baz, bongo)"
2004-08-31 10:07:13 +00:00
Raymond Hettinger 1792bfbf90 Bypass peepholing of code with lineno tables having intervals >= 255.
Allows the lineno fixup code to remain simple and not have to deal with
multibyte codings.

* Add an assertion to that effect.
* Remove the XXX comment on the subject.
2004-08-25 17:19:38 +00:00
Raymond Hettinger 65d3c0537a Fix typo in comment and add clarification. 2004-08-25 15:15:56 +00:00
Raymond Hettinger ef0a82b682 Simplify chains of conditional jumps.
(Suggested by Neal Norwitz.)
2004-08-25 03:18:29 +00:00
Raymond Hettinger a12fa148d7 Incorporate review comments courtesy of Neal Norwitz:
* Perform the code length check earlier.
* Eliminate the extra PyMem_Free() upon hitting an EXTENDED_ARG.
* Assert that the NOP count used in jump retargeting matches the NOPs
  eliminated in the final step.
* Add an XXX note to indicate that more work is being to done to
  handle linenotab with intervals > 255.
2004-08-24 04:34:16 +00:00
Raymond Hettinger fd2d1f7870 SF Patch #1013667: Cleanup Peepholer Output
* Make a pass to eliminate NOPs.  Produce code that is more readable,
  more compact, and a tiny bit faster.  Makes the peepholer more flexible
  in the scope of allowable transformations.

* With Guido's okay, bumped up the magic number so that this patch gets
  widely exercised before the alpha goes out.
2004-08-23 23:37:48 +00:00
Raymond Hettinger 1a7892924f Move the bytecode optimizer upstream so that its results are saved in pyc
files and not re-optimized upon import.  Saves a bit of startup time while
still remaining decoupled from the rest of the compiler.

As a side benefit, handcoded bytecode is not run through the optimizer
when new code objects are created.  Hopefully, a handcoder has already
created exactly what they want to have run.

(Idea suggested by Armin Rigo and Michael Hudson.  Initially avoided
 because of worries about compiler coupling; however, only the nexus
 point needed to be moved so there won't be a conflict when the AST
 branch is loaded.)
2004-08-18 05:22:06 +00:00
Michael W. Hudson 0ccff074cd This is Mark Russell's patch:
[ 1009560 ] Fix @decorator evaluation order

From the description:

Changes in this patch:

- Change Grammar/Grammar to require
newlines between adjacent decorators.

- Fix order of evaluation of decorators
in the C (compile.c) and python
(Lib/compiler/pycodegen.py) compilers

- Add better order of evaluation check
to test_decorators.py (test_eval_order)

- Update the decorator documentation in
the reference manual (improve description
of evaluation order and update syntax
description)

and the comment:

Used Brett's evaluation order (see
http://mail.python.org/pipermail/python-dev/2004-August/047835.html)

(I'm checking this in for Anthony who was having problems getting SF to
talk to him)
2004-08-17 17:29:16 +00:00
Brett Cannon add33601c2 Correct the order of application for decorators. Meant to be bottom-up and not
top-down.  Now matches the PEP.
2004-08-15 07:21:25 +00:00
Brett Cannon 31f8350f43 Fix incorrect comment for (struct compiling)->c_cellvars 2004-08-15 01:15:01 +00:00
Tim Peters d459f536c5 code_new(): Wouldn't compile on Windows, because of gcc'ism. 2004-08-12 18:16:43 +00:00
Michael W. Hudson 6093462739 Fix bug
[ 1005248 ] new.code() not cleanly checking its arguments

using the result of new.code() can still destroy the sun, but merely
calling the function shouldn't any more.

I also rewrote the existing tests of new.code() to use vastly less
un-bogus arguments, and added tests for the previous insane behaviours.
2004-08-12 17:56:29 +00:00
Raymond Hettinger 98bd1814e2 SF bug #1004088: big code objects (>64K) may be optimized incorrectly
Will backport.
2004-08-06 19:46:34 +00:00
Michael W. Hudson e51c4f9a10 Revert 2.312; turns out interning the file name did do some good (reducing
.pyc size) after all.
2004-08-04 10:26:08 +00:00
Michael W. Hudson 782d8ffb42 Don't intern the filename of a file being compiled.
Hard to believe it ever helped anything, and it hurts finding reference
leaks.
2004-08-03 10:24:03 +00:00
Anthony Baxter 4e7785aa4f fix for @decorators under a debug build. 2004-08-02 11:08:15 +00:00
Anthony Baxter c2a5a63654 PEP-0318, @decorator-style. In Guido's words:
"@ seems the syntax that everybody can hate equally"
Implementation by Mark Russell, from SF #979728.
2004-08-02 06:10:11 +00:00
Raymond Hettinger 11a70c79b9 Upgrade None assignment SyntaxWarning to a SyntaxError. 2004-07-17 21:46:25 +00:00
Tim Peters db5860b7c7 optimize_code(): Repaired gross error in new special-casing for None.
The preceding case statement was missing a terminating "break" stmt,
so fell into the new code by mistake.  This caused uncaught out-of-bounds
accesses to the "names" tuple, leading to a variety of insane behaviors.
2004-07-17 05:00:52 +00:00
Raymond Hettinger 76d962d700 Treat None as a constant. 2004-07-16 12:16:48 +00:00
Neal Norwitz 93468eac72 Remove unused macros in .c files 2004-07-08 01:49:00 +00:00
Raymond Hettinger 43ea47ff06 Move NOP to end of code transformation. 2004-06-24 09:25:39 +00:00
Raymond Hettinger 9c18e81fb2 Install two code generation optimizations that depend on NOP.
Reduces the cost of "not" to almost zero.
2004-06-21 16:31:15 +00:00
Martin v. Löwis 737ea82a5a Patch #774665: Make Python LC_NUMERIC agnostic. 2004-06-08 18:52:54 +00:00
Raymond Hettinger 354433a59d SF patch #872326: Generator expression implementation
(Code contributed by Jiwon Seo.)

The documentation portion of the patch is being re-worked and will be
checked-in soon.  Likewise, PEP 289 will be updated to reflect Guido's
rationale for the design decisions on binding behavior (as described in
in his patch comments and in discussions on python-dev).

The test file, test_genexps.py, is written in doctest format and is
meant to exercise all aspects of the the patch.  Further additions are
welcome from everyone.  Please stress test this new feature as much as
possible before the alpha release.
2004-05-19 08:20:33 +00:00
Armin Rigo 80d937e986 Fix for line events in the case:
def f(a):
    if a:
      print 5
    else:
      pass
2004-03-22 17:52:53 +00:00
Raymond Hettinger ff5bc50bb0 Improve byte coding for multiple assignments.
Gives 30% speedup on "a,b=1,2" and 25% on "a,b,c=1,2,3".
2004-03-21 15:12:00 +00:00
Raymond Hettinger dd80f76265 SF patch #910929: Optimize list comprehensions
Add a new opcode, LIST_APPEND, and apply it to the code generation for
list comprehensions.  Reduces the per-loop overhead by about a third.
2004-03-07 07:31:06 +00:00
Guido van Rossum 6c9e130524 - Removed FutureWarnings related to hex/oct literals and conversions
and left shifts.  (Thanks to Kalle Svensson for SF patch 849227.)
  This addresses most of the remaining semantic changes promised by
  PEP 237, except for repr() of a long, which still shows the trailing
  'L'.  The PEP appears to promise warnings for operations that
  changed semantics compared to Python 2.3, but this is not
  implemented; we've suffered through enough warnings related to
  hex/oct literals and I think it's best to be silent now.
2003-11-29 23:52:13 +00:00
Martin v. Löwis dd7eb146a8 Patch #792869: Clarify error message for parameters declared global,
rename LOCAL_GLOBAL to PARAM_GLOBAL.
2003-10-18 22:05:25 +00:00
Raymond Hettinger 8ae4689657 Simplify and speedup uses of Py_BuildValue():
* Py_BuildValue("(OOO)",a,b,c)  -->  PyTuple_Pack(3,a,b,c)
* Py_BuildValue("()",a)         -->  PyTuple_New(0)
* Py_BuildValue("O", a)         -->  Py_INCREF(a)
2003-10-12 19:09:37 +00:00
Jeremy Hylton 9832613beb Fix SF bug [ 808594 ] leak on lambda with duplicate arguments error.
Refactor code so that one helper routine sets error location and
increments st_errors.

Bug fix candidate.
2003-09-22 04:26:44 +00:00
Raymond Hettinger cc1798e0c0 Improve the leak fix so that PyTuple_New is only called when needed. 2003-09-16 04:27:52 +00:00
Raymond Hettinger 37a724d718 Fix leak discovered in test_new by Michael Hudson.
Will backport to 2.3.1
2003-09-15 21:43:16 +00:00
Jeremy Hylton 376e63d5cd Fix for SF bug [ 784075 ] Fatal Python error: unknown scope
Make sure the inner function is not compiled when there is a syntax
error in the default arguments.
2003-08-28 14:42:14 +00:00
Jeremy Hylton 1955fcf67a SF patch 763201: handling of SyntaxErrors in symbol table build
Fixes for three related bugs, including errors that caused a script to
be ignored without printing an error message.  The key problem was a bad
interaction between syntax warnings and syntax errors.  If an
exception was already set when a warning was issued, the warning could
clobber the exception.

The PyErr_Occurred() check in issue_warning() isn't entirely
satisfying (the caller should know whether there was already an
error), but a better solution isn't immediately obvious.

Bug fix candidate.
2003-07-15 20:23:26 +00:00
Raymond Hettinger b9572c3456 Removed bytecode transformation for sequence packing/unpacking.
It depended on the previously removed basic block checker to
prevent a jump into the middle of the transformed block.

Clears SF 757818: tuple assignment -- SystemError: unknown opcode
2003-06-20 16:13:17 +00:00
Neal Norwitz a1d654e13a SF bug #644345, Poor error message for augmented assign
Update error message.  Hopefully this is clearer to some people.
2003-05-22 22:00:04 +00:00
Jeremy Hylton f9415e6245 Use macro to get length of list.
Remove comment about how code used to work.
2003-05-22 16:22:33 +00:00
Jeremy Hylton 521482d84f Remove comment with very long lines that explained what the code used
to do.

XXX Please be careful when checking in patches to avoid checking in
junk that explains what the patched code used to do.
2003-05-22 15:47:02 +00:00
Jeremy Hylton 4d508adae3 Fix for SF [ 734869 ] Lambda functions in list comprehensions
The compiler was reseting the list comprehension tmpname counter for each function, but the symtable was using the same counter for the entire module.  Repair by move tmpname into the symtable entry.

Bugfix candidate.
2003-05-21 17:34:50 +00:00
Tim Peters 12d55a7caa cmp_type(): The grammar stopped allowing '=' as a comparison operator
about a decade ago.  Put the code still allowing for it in cmp_type()
out of its lonely misery.
2003-05-12 19:16:52 +00:00
Martin v. Löwis a94568a753 Patch #734231: Update RiscOS support. In particular, correct
riscospath.extsep, and use os.extsep throughout.
2003-05-10 07:36:56 +00:00
Martin v. Löwis cd12bfc142 Patch #708604: Check more function results. Will backport to 2.2. 2003-05-03 10:53:08 +00:00
Michael W. Hudson 26848a34d1 Use Tim's suggestion to fix
[ 708901 ] Lineno calculation sometimes broken

A one line patch to compile.c and a rather-more-than-one-line patch
to test_dis.  Hey ho.

Possibly a backport candidate -- tho' lnotab is less used in 2.2...
2003-04-29 17:07:36 +00:00
Raymond Hettinger f4cf76dd5e Revert the previous enhancement to the bytecode optimizer.
The additional code complexity and new NOP opcode were not worth it.
2003-04-24 05:45:23 +00:00
Raymond Hettinger 060641d511 Improved the bytecode optimizer.
* Can now test for basic blocks.
* Optimize inverted comparisions.
* Optimize unary_not followed by a conditional jump.
* Added a new opcode, NOP, to keep code size constant.
* Applied NOP to previous transformations where appropriate.

Note, the NOP would not be necessary if other functions were
added to re-target jump addresses and update the co_lnotab mapping.
That would yield slightly faster and cleaner bytecode at the
expense of optimizer simplicity and of keeping it decoupled
from the line-numbering structure.
2003-04-22 06:49:11 +00:00
Raymond Hettinger 255a3d08a1 Extend SF patch #707257: Improve code generation
to cover the case for:  "x,y,z=1,2,3".  Gives a 30% speed-up.

Also, added FOR_ITER to the list of opcodes that can jump.
2003-04-15 10:35:07 +00:00
Raymond Hettinger 5b75c38cc9 Factored out test for absolute jumps. 2003-03-28 12:05:00 +00:00
Raymond Hettinger f6f575ae6f SF patch #707257: Improve code generation
Adds a single function to improve generated bytecode.  Has a single line
attachment point, so it is completely de-coupled from both the compiler
and ceval.c.

Makes three simple transforms that do not require a basic block analysis
or re-ordering of code.  Gives improved timings on pystone, pybench,
and any code using either "while 1" or "x,y=y,x".
2003-03-26 01:07:54 +00:00
Tim Peters b39903b0a0 symtable_cellvar_offsets(): This leaked references to little integers
in normal cases, and also in error cases.

Bugfix candidate.
2003-03-24 17:22:24 +00:00
Guido van Rossum 66b1259dbc SF #660455 : patch by NNorwitz.
"Unsigned" (i.e., positive-looking, but really negative) hex/oct
constants with a leading minus sign are once again properly negated.
The micro-optimization for negated numeric constants did the wrong
thing for such hex/oct constants.  The patch avoids the optimization
for all hex/oct constants.

This needs to be backported to Python 2.2!
2003-02-12 16:57:47 +00:00
Just van Rossum 3aaf42c613 patch #683515: "Add unicode support to compile(), eval() and exec"
Incorporated nnorwitz's comment re. Py__USING_UNICODE.
2003-02-10 08:21:10 +00:00
Jeremy Hylton 985eba53f5 Small function call optimization and special build option for call stats.
-DCALL_PROFILE: Count the number of function calls executed.

When this symbol is defined, the ceval mainloop and helper functions
count the number of function calls made.  It keeps detailed statistics
about what kind of object was called and whether the call hit any of
the special fast paths in the code.

Optimization:

When we take the fast_function() path, which seems to be taken for
most function calls, and there is minimal frame setup to do, avoid
call PyEval_EvalCodeEx().  The eval code ex function does a lot of
work to handle keywords args and star args, free variables,
generators, etc.  The inlined version simply allocates the frame and
copies the arguments values into the frame.

The optimization gets a little help from compile.c which adds a
CO_NOFREE flag to code objects that don't have free variables or cell
variables.  This change allows fast_function() to get into the fast
path with fewer tests.

I measure a couple of percent speedup in pystone with this change, but
there's surely more that can be done.
2003-02-05 23:13:00 +00:00
Michael W. Hudson 976249be74 A. Lloyd Flanagan pointed out a spelling error on c.l.py. 2003-01-16 15:39:07 +00:00
Jeremy Hylton accb62b28e SF patch [ 597919 ] compiler package and SET_LINENO
A variety of changes from Michael Hudson to get the compiler working
with 2.3.  The primary change is the handling of SET_LINENO:

# The set_lineno() function and the explicit emit() calls for
# SET_LINENO below are only used to generate the line number table.
# As of Python 2.3, the interpreter does not have a SET_LINENO
# instruction.  pyassem treats SET_LINENO opcodes as a special case.

A few other small changes:
 - Remove unused code from pycodegen and pyassem.
 - Fix error handling in parsermodule.  When PyParser_SimplerParseString()
   fails, it sets an exception with detailed info.  The parsermodule
   was clobbering that exception and replacing it was a generic
   "could not parse string" exception.  Keep the original exception.
2002-12-31 18:17:44 +00:00
Guido van Rossum 633d90c7a3 Oops. Roll back that last change. It wasn't ready for release. :-( 2002-12-23 16:51:42 +00:00
Guido van Rossum 9c8a0866c9 Add warning for assignment to None, True and False. This is patch
549213 by Jeremy (checking in for him since he's away and busy).
2002-12-23 16:35:23 +00:00
Neal Norwitz 06982221bb SF # 654960, remove unnecessary static variable
The static variable (implicit) was not necessary.
The c_globals can be None or True now.
2002-12-18 01:18:44 +00:00
Gustavo Niemeyer 78429a6aa6 Fixing bug
[#448679] Left to right

* Python/compile.c
  (com_dictmaker): Reordered evaluation of dictionaries to follow strict
  LTR evaluation.

* Lib/compiler/pycodegen.py
  (CodeGenerator.visitDict): Reordered evaluation of dictionaries to
  follow strict LTR evaluation.

* Doc/ref/ref5.tex
  Documented the general LTR evaluation order idea.

* Misc/NEWS
  Documented change in evaluation order of dictionaries.
2002-12-16 13:54:02 +00:00
Martin v. Löwis 95292d6caa Constify filenames and scripts. Fixes #651362. 2002-12-11 14:04:59 +00:00
Walter Dörwald 4c6c76559d Move three variables that are only used inside an if block into the block,
so the --disable-unicode build doesn't complain about unused variables.
2002-11-21 20:13:40 +00:00
Michael W. Hudson 3ae3315f44 Clamp code objects' tp_compare result to [-1, 1].
Bugfix candidate.
2002-10-03 09:50:47 +00:00
Michael W. Hudson 53d58bb369 Further SET_LINENO reomval fixes. See comments in patch #587933.
Use a slightly different strategy to determine when not to call the line
trace function.  This removes the need for the RETURN_NONE opcode, so
that's gone again.  Update docs and comments to match.

Thanks to Neal and Armin!

Also add a test suite.  This should have come with the original patch...
2002-08-30 13:09:51 +00:00
Guido van Rossum b7164621fa Add warnings for arguments named None. All set. (I could add a
warning for 'global None', but that's either accompanied by an
assignment to None, which will trigger a warning, or not, in which
case it's harmless. :-)
2002-08-16 02:48:11 +00:00
Guido van Rossum 63dd79ac04 Add warning for None used as keyword argument name in function call.
Still to do: function definition arguments (including *None and
**None).
2002-08-16 02:24:56 +00:00
Guido van Rossum 3ac99d4910 Add warnings for assignment or deletion of variables and attributes
named 'None'.  Still to do: function definition parameter lists, and
function call keyword arguments.
2002-08-16 02:13:49 +00:00
Guido van Rossum b081e0c74c Minor cleanup of parsename() and parsestr(): the 'struct compiling *'
argument should be called 'c', like everywhere else.  Renamed a
complex variable 'c' to 'z' and moved it inside the only scope where
it's used.
2002-08-16 01:57:32 +00:00
Michael W. Hudson dd32a91cc0 This is my patch
[ 587993 ] SET_LINENO killer

Remove SET_LINENO.  Tracing is now supported by inspecting co_lnotab.

Many sundry changes to document and adapt to this change.
2002-08-15 14:59:02 +00:00
Barry Warsaw 9f00739551 Added a FutureWarning for constructs that will change semantically in
the future.  Changed PEP 237 hex constant warnings from
DeprecationWarning to FutureWarning.  Updated the documentation.
2002-08-14 15:51:29 +00:00
Martin v. Löwis 8a8da798a5 Patch #505705: Remove eval in pickle and cPickle. 2002-08-14 07:46:28 +00:00
Guido van Rossum 715eca932e Use PyErr_WarnExplicit() to warn about hex/oct constants, so the
correct filename and line number are reported.
2002-08-12 21:54:46 +00:00
Guido van Rossum 3cb8e54da0 Reset errno to zero after calling PyErr_Warn(). It can potentially do
a lot of work, including I/O (e.g. to import warnings.py), which might
affect errno.
2002-08-11 14:06:15 +00:00
Guido van Rossum 078151da90 Implement stage B0 of PEP 237: add warnings for operations that
currently return inconsistent results for ints and longs; in
particular: hex/oct/%u/%o/%x/%X of negative short ints, and x<<n that
either loses bits or changes sign.  (No warnings for repr() of a long,
though that will also change to lose the trailing 'L' eventually.)

This introduces some warnings in the test suite; I'll take care of
those later.
2002-08-11 04:24:12 +00:00
Martin v. Löwis 2863c10a86 Use Py_FatalError instead of abort. 2002-08-07 15:18:57 +00:00
Martin v. Löwis 019934b3cc Fix PEP 263 code --without-unicode. Fixes #591943. 2002-08-07 12:33:18 +00:00
Martin v. Löwis 00f1e3f5a5 Patch #534304: Implement phase 1 of PEP 263. 2002-08-04 17:29:52 +00:00
Andrew MacIntyre 80d4e2acf5 SF patch #578297:
Change the parser and compiler to use PyMalloc.

Only the files implementing processes that will request memory
allocations small enough for PyMalloc to be a win have been
changed, which are:-
 - Python/compile.c
 - Parser/acceler.c
 - Parser/node.c
 - Parser/parsetok.c

This augments the aggressive overallocation strategy implemented by
Tim Peters in PyNode_AddChild() [Parser/node.c], in reducing the
impact of platform malloc()/realloc()/free() corner case behaviour.
Such corner cases are known to be triggered by test_longexp and
test_import.

Jeremy Hylton, in accepting this patch, recommended this as a
bugfix candidate for 2.2.  While the changes to Python/compile.c
and Parser/node.c backport easily (and could go in), the changes
to Parser/acceler.c and Parser/parsetok.c require other not
insignificant changes as a result of the differences in the memory
APIs between 2.3 and 2.2, which I'm not in a position to work
through at the moment.  This is a pity, as the Parser/parsetok.c
changes are the most important after the Parser/node.c changes, due
to the size of the memory requests involved and their frequency.
2002-08-04 06:28:21 +00:00
Skip Montanaro 21ee37c94e remove spurious SET_LINENO from com_list_for and com_list_if. All they do
is slow things down unnecessarily and make tracing much more verbose.
Something like

    def f(n):
	return [i for i in range(n) if i%2]

should have at most two SET_LINENO instructions, not four.  When tracing,
the current line number should be printed once, not 2*n+1 times.
2002-07-25 06:18:42 +00:00
Raymond Hettinger 0ae0c07661 SF 569257 -- Name mangle double underscored variable names in __slots__. 2002-06-20 22:23:15 +00:00
Guido van Rossum bea18ccde6 SF patch 568629 by Oren Tirosh: types made callable.
These built-in functions are replaced by their (now callable) type:

    slice()
    buffer()

and these types can also be called (but have no built-in named
function named after them)

    classobj (type name used to be "class")
    code
    function
    instance
    instancemethod (type name used to be "instance method")

The module "new" has been replaced with a small backward compatibility
placeholder in Python.

A large portion of the patch simply removes the new module from
various platform-specific build recipes.  The following binary Mac
project files still have references to it:

    Mac/Build/PythonCore.mcp
    Mac/Build/PythonStandSmall.mcp
    Mac/Build/PythonStandalone.mcp

[I've tweaked the code layout and the doc strings here and there, and
added a comment to types.py about StringTypes vs. basestring.  --Guido]
2002-06-14 20:41:17 +00:00
Jeremy Hylton 05ab2e693c Fix SF bug [ 561825 ] Confusing error for "del f()"
In the error message, say del for del and assign for everything else.
2002-05-31 14:08:29 +00:00
Guido van Rossum 05459c5e5e Accept u"..." literals even when Unicode is disabled. But these
literals must not contain \u, \U or \N escapes.  (XXX Should they also
not contain non-ASCII characters?)
2002-05-28 18:47:29 +00:00
Guido van Rossum 2d3b986480 Disambiguate the grammar for backtick.
The old syntax suggested that a trailing comma was OK inside backticks,
but in fact (due to ideosyncrasies of pgen) it was not.  Fix the grammar
to avoid the ambiguity.  Fred: you may want to update the refman.
2002-05-24 15:47:06 +00:00
Neil Schemenauer 89e3ee0ccf If Py_OptimizeFlag is false then always evaluate assert conditions, don't
test __debug__ at runtime.  Closes SF patch #548833.
2002-04-26 01:58:53 +00:00
Tim Peters 51e7f5caba Moving pymalloc along.
+ Redirect PyMem_{Del, DEL} to the object allocator's free() when
  pymalloc is enabled.  Needed so old extensions can continue to
  mix PyObject_New with PyMem_DEL.

+ This implies that pgen needs to be able to see the PyObject_XYZ
  declarations too.  pgenheaders.h now includes Python.h.  An
  implication is that I expect obmalloc.o needs to get linked into
  pgen on non-Windows boxes.

+ When PYMALLOC_DEBUG is defined, *all* Py memory API functions
  now funnel through the debug allocator wrapper around pymalloc.
  This is the default in a debug build.

+ That caused compile.c to fail:  it indirectly mixed PyMem_Malloc
  with raw platform free() in one place.  This is verbotten.
2002-04-22 02:33:27 +00:00
Martin v. Löwis 314fc79ce8 Patch #542659: Eliminate duplicate check for NULL of freevars/cellvars. 2002-04-14 09:53:49 +00:00
Martin v. Löwis cfeb3b6ab8 Patch #50002: Display line information for bad \x escapes:
- recognize "SyntaxError"s by the print_file_and_line attribute.
- add the syntaxerror attributes to all exceptions in compile.c.
Fixes #221791
2002-03-03 21:30:27 +00:00
Neal Norwitz 150d09d360 Fix missing space between words. Bugfix candidate. 2002-01-29 00:56:37 +00:00
Neal Norwitz 2a47c0fa23 Fix spelling mistakes. Bugfix candidates. 2002-01-29 00:53:41 +00:00
Guido van Rossum 340cbe74b9 A tentative fix for SF bug #503837 (Roeland Rengelink):
type.__module__ problems (again?)

This simply initializes the __module__ local in a class statement from
the __name__ global.  I'm not 100% sure that this is the correct fix,
although it usually does the right thing.  The problem is that if the
class statement executes in a custom namespace, the __name__ global
may be taken from __builtins__, in which case it would have the value
__builtin__, or it may not exist at all (if the custom namespace also
has a custom __builtins__), in which case the class statement will
fail.

Nevertheless, unless someone finds a better solution, this is a 2.2.1
bugfix too.
2002-01-15 21:06:07 +00:00
Martin v. Löwis 7198a525f3 Patch #494783: Rename cmp_op enumerators. 2002-01-01 19:59:11 +00:00
Jeremy Hylton 733c8935f9 Fix for SF bug [ #492403 ] exec() segfaults on closure's func_code
Based on the patch from Danny Yoo.  The fix is in exec_statement() in
ceval.c.

There are also changes to introduce use of PyCode_GetNumFree() in
several places.
2001-12-13 19:51:56 +00:00
Jeremy Hylton 86424e333f SF bug #488687 reported by Neal Norwitz
The error for assignment to __debug__ used ste->ste_opt_lineno instead
of n->n_lineno.  The latter was at best incorrect; often the slot was
uninitialized.  Two fixes here: Use the correct lineno for the error.
Initialize ste_opt_lineno in PySymtable_New(); while there are no
current cases where it is referenced unless it has already been
assigned to, there is no harm in initializing it.
2001-12-04 02:41:46 +00:00
Barry Warsaw 8f6d868bbb code_repr(), com_addop_varname(), com_list_comprehension(),
com_arglist(), symtable_check_unoptimized(), symtable_params(),
symtable_global(), symtable_list_comprehension():

    Conversion of sprintf() to PyOS_snprintf() for buffer overrun
    avoidance.
2001-11-28 21:10:39 +00:00
Marc-André Lemburg d4c0a9c59b Fixes for possible buffer overflows in sprintf() usages. 2001-11-28 11:47:00 +00:00
Jeremy Hylton 9f64caaf00 Use PyObject_CheckReadBuffer(). 2001-11-09 22:02:48 +00:00
Jeremy Hylton 778e265462 Fix SF buf #480096: Assign to __debug__ still allowed
Easy enough to catch assignment in the compiler.  The perverse user
can still change the value of __debug__, but that may be the least he
can do.
2001-11-09 19:50:08 +00:00
Tim Peters 8c5e41559c Part of SF bug #478003 possible memory leaks in err handling.
PyNode_CompileSymtable:  if symtable_init() fails, free the memory
allocated for the PyFutureFeatures struct.
2001-11-04 19:26:58 +00:00
Jeremy Hylton 961dfe0d85 Fix for SF bug [ #471928 ] global made w/nested list comprehensions
The symbol table pass didn't have an explicit case for the list_iter
node which is used only for a nested list comprehension.  As a result,
the target of the list comprehension was treated as a use instead of
an assignment.  Fix is to add a case to symtable_node() to handle
list_iter.

Also, rework and document a couple of the subtler implementation
issues in the symbol table pass.  The symtable_node() switch statement
depends on falling through the last several cases, in order to handle
some of the more complicated nodes like atom.  Add a comment
explaining the behavior before the first fall through case.  Add a
comment /* fall through */ at the end of case so that it is explicitly
marked as such.

Move the for_stmt case out of the fall through logic, which simplifies
both for_stmt and default.  (The default used the local variable start
to skip the first three nodes of a for_stmt when it fell through.)

Rename the flag argument to symtable_assign() to def_flag and add a
comment explaining its use:

   The third argument to symatble_assign() is a flag to be passed to
   symtable_add_def() if it is eventually called.  The flag is useful
   to specify the particular type of assignment that should be
   recorded, e.g. an assignment caused by import.
2001-10-18 16:15:10 +00:00
Jeremy Hylton 93a569d634 Fix computation of stack depth for classdef and closures.
Also minor tweaks to internal routines.
Use PyCF_MASK instead of explicit list of flags.

For the MAKE_CLOSURE opcode, the number of items popped off the stack
depends on both the oparg and the number of free variables for the
code object.  Fix the code so it accounts for the free variables.

In com_classdef(), record an extra pop to account for the STORE call
after the BUILD_CLASS.

Get rid of some commented out debugging code in com_push() and
com_pop().

Factor string resize logic into helper routine com_check_size().

In com_addbyte(), remove redudant if statement after assert.  (They
test the same condition.)

In several routines, use string macros instead of string functions.
2001-10-17 13:22:22 +00:00
Guido van Rossum 1c917072ca Very subtle syntax change: in a list comprehension, the testlist in
"for <var> in <testlist> may no longer be a single test followed by
a comma.  This solves SF bug #431886.  Note that if the testlist
contains more than one test, a trailing comma is still allowed, for
maximum backward compatibility; but this example is not:

    [(x, y) for x in range(10), for y in range(10)]
                              ^

The fix involved creating a new nonterminal 'testlist_safe' whose
definition doesn't allow the trailing comma if there's only one test:

    testlist_safe: test [(',' test)+ [',']]
2001-10-15 15:44:05 +00:00
Guido van Rossum 6f7993765a Add optional docstrings to member descriptors. For backwards
compatibility, this required all places where an array of "struct
memberlist" structures was declared that is referenced from a type's
tp_members slot to change the type of the structure to PyMemberDef;
"struct memberlist" is now only used by old code that still calls
PyMember_Get/Set.  The code in PyObject_GenericGetAttr/SetAttr now
calls the new APIs PyMember_GetOne/SetOne, which take a PyMemberDef
argument.

As examples, I added actual docstrings to the attributes of a few
types: file, complex, instance method, super, and xxsubtype.spamlist.

Also converted the symtable to new style getattr.
2001-09-20 20:46:19 +00:00
Jeremy Hylton c785f4841c Supply code objects a new-style tp_members slot and tp_getattr impl.
The chief effects are to make dir() do something useful and supply
them with an __class__.
2001-09-14 20:08:07 +00:00
Tim Peters 51e2651b29 SF bug [#458941] Looks like a unary minus bug.
com_factor():  when a unary minus is attached to a float or imaginary zero,
don't optimize the UNARY_MINUS opcode away:  the const dict can't
distinguish between +0.0 and -0.0, so ended up treating both like the
first one added to it.  Optimizing UNARY_PLUS away isn't a problem.

(BTW, I already uploaded the 2.2a3 Windows installer, and this isn't
important enough to delay the release.)
2001-09-07 08:45:55 +00:00
Fred Drake 14ef244dfe When re-writing a factor containing a unary negation of a literal, only
affect nodes without another operator.  This was causing negated
exponentiations to drop the exponentiation.  This closes SF bug #456756.
2001-08-30 18:53:25 +00:00
Sjoerd Mullender a2c2ae62df Removed unreachable goto statement to silence SGI compiler. 2001-08-30 14:06:45 +00:00
Jeremy Hylton 71b6af91d3 If an integer constant can't be generated from an integer literal
because of overflow, generate a long instead.
2001-08-27 19:45:25 +00:00