]> granicus.if.org Git - python/log
python
28 years agoBasic README file
Guido van Rossum [Thu, 30 Jan 1997 03:24:00 +0000 (03:24 +0000)]
Basic README file

28 years agoAdded robots.txt support, using Skip Montanaro's parser.
Guido van Rossum [Thu, 30 Jan 1997 03:19:41 +0000 (03:19 +0000)]
Added robots.txt support, using Skip Montanaro's parser.
Fixed occasional inclusion of unpicklable objects (Message in errors).
Changed indent of a few messages.

28 years agoSkip Montanaro's robots.txt parser.
Guido van Rossum [Thu, 30 Jan 1997 03:18:23 +0000 (03:18 +0000)]
Skip Montanaro's robots.txt parser.

28 years agoweb tree checker
Guido van Rossum [Thu, 30 Jan 1997 02:44:48 +0000 (02:44 +0000)]
web tree checker

28 years agomime types guesser
Guido van Rossum [Thu, 30 Jan 1997 02:44:20 +0000 (02:44 +0000)]
mime types guesser

28 years agoDid nobody ever notice that "make test" didn't print the summary line
Guido van Rossum [Wed, 29 Jan 1997 16:03:45 +0000 (16:03 +0000)]
Did nobody ever notice that "make test" didn't print the summary line
any more?  This was because the *parent* process in test_socket.py did
an os._exit(0)!  Getting rid of that now...

28 years agoSmall lookmapping nits:
Guido van Rossum [Wed, 29 Jan 1997 15:53:56 +0000 (15:53 +0000)]
Small lookmapping nits:
- remove bogus initialization using uninitialized i
- derive initial incr from hash
- copy mp->ma_table into a local variable

28 years agoposix_execve(): Accept any mapping protocol object for the env
Barry Warsaw [Wed, 29 Jan 1997 15:08:24 +0000 (15:08 +0000)]
posix_execve(): Accept any mapping protocol object for the env
argument, not hardwired to a dictionary.

28 years agoFix bug reported by Per Lindqvist: "%#06x" % 1 stuck the 0 padding
Guido van Rossum [Wed, 29 Jan 1997 06:00:24 +0000 (06:00 +0000)]
Fix bug reported by Per Lindqvist: "%#06x" % 1 stuck the 0 padding
in front of the 0x, like such: "0000x1".

28 years agoFinal three poly table entries corrected by Tim Peters.
Guido van Rossum [Wed, 29 Jan 1997 04:45:16 +0000 (04:45 +0000)]
Final three poly table entries corrected by Tim Peters.
Reindented the whole table.

28 years agoChange bsddbmodule.o -> bsddbmodule.c.
Guido van Rossum [Tue, 28 Jan 1997 02:21:56 +0000 (02:21 +0000)]
Change bsddbmodule.o -> bsddbmodule.c.

28 years agoAdded LDLAST variable, substituted by configure script, for the final
Guido van Rossum [Tue, 28 Jan 1997 02:21:24 +0000 (02:21 +0000)]
Added LDLAST variable, substituted by configure script, for the final
argument to the linker (required for DEC Alpha threads).

28 years agoChanged the lookup algorithm again, based on Reimer Behrends's post.
Guido van Rossum [Tue, 28 Jan 1997 00:00:11 +0000 (00:00 +0000)]
Changed the lookup algorithm again, based on Reimer Behrends's post.
The table size is now constrained to be a power of two, and we use a
variable increment based on GF(2^n)-{0} (not that I have the faintest
idea what that is :-) which helps avoid the expensive '%' operation.

Some of the entries in the table of polynomials have been modified
according to a post by Tim Peters.

28 years agoTwo small changes:
Guido van Rossum [Mon, 27 Jan 1997 23:42:36 +0000 (23:42 +0000)]
Two small changes:

- Use co->... instead of f->f_code->...; save an extra lookup of what
we already have in a local variable).

- Remove test for nlocals > 0 before setting fastlocals to
f->f_localsplus; 0 is a rare case and the assignment is safe even
then.

28 years agonew.code() has grown another foot, eh, another two arguments.
Guido van Rossum [Mon, 27 Jan 1997 23:25:37 +0000 (23:25 +0000)]
new.code() has grown another foot, eh, another two arguments.

28 years agoPlug a leak with calling something other than a function or method is
Guido van Rossum [Mon, 27 Jan 1997 21:30:09 +0000 (21:30 +0000)]
Plug a leak with calling something other than a function or method is
called with keyword arguments -- the keyword and value were leaked.
This affected for instance with a __call__() method.

Bug reported and fix supplied by Jim Fulton.

28 years agoPatches for (two forms of) optional dynamic execution profiling --
Guido van Rossum [Fri, 24 Jan 1997 13:49:28 +0000 (13:49 +0000)]
Patches for (two forms of) optional dynamic execution profiling --
i.e., counting opcode frequencies, or (with DXPAIRS defined) opcode
pair frequencies.  Define DYNAMIC_EXECUTION_PROFILE on the command
line (for this file and for sysmodule.c) to enable.

28 years agoChange the control flow for error handling in the function prelude to
Guido van Rossum [Fri, 24 Jan 1997 04:19:24 +0000 (04:19 +0000)]
Change the control flow for error handling in the function prelude to
jump to the "Kill locals" section at the end.  Add #ifdef macintosh
bandaid to make sure we call sigcheck() on the Mac.

28 years agoAdded optional interface for dynamic execution profile (to be gathered
Guido van Rossum [Fri, 24 Jan 1997 04:07:45 +0000 (04:07 +0000)]
Added optional interface for dynamic execution profile (to be gathered
in ceval.c).

28 years agoGet the line number from PyCode_Addr2Line instead of believing
Guido van Rossum [Fri, 24 Jan 1997 04:02:55 +0000 (04:02 +0000)]
Get the line number from PyCode_Addr2Line instead of believing
tb_lineno.  Store it in tb_lineno for the user.

28 years agoWhen a recycled frame has more local+stack slots than needed,
Guido van Rossum [Fri, 24 Jan 1997 04:00:21 +0000 (04:00 +0000)]
When a recycled frame has more local+stack slots than needed,
give the extra slots to the stack rather than than forgetting about
them (this reduces the number of reallocs done).

28 years agoTwo more arguments to newcodeobject -- first lineno and lineno table.
Guido van Rossum [Fri, 24 Jan 1997 03:58:52 +0000 (03:58 +0000)]
Two more arguments to newcodeobject -- first lineno and lineno table.

28 years agoNew magin number (because of linenumber table).
Guido van Rossum [Fri, 24 Jan 1997 03:44:53 +0000 (03:44 +0000)]
New magin number (because of linenumber table).

28 years agoMarshal the line number table of code objects.
Guido van Rossum [Fri, 24 Jan 1997 03:44:17 +0000 (03:44 +0000)]
Marshal the line number table of code objects.

28 years agoInstead of emitting SET_LINENO instructions, generate a line number
Guido van Rossum [Fri, 24 Jan 1997 03:43:35 +0000 (03:43 +0000)]
Instead of emitting SET_LINENO instructions, generate a line number
table which is incorporated in the code object.  This way, the runtime
overhead to keep track of line numbers is only incurred when an
exception has to be reported.

28 years agoAdd co_firstlineno and co_lnotab, for the line number table generated
Guido van Rossum [Fri, 24 Jan 1997 03:41:09 +0000 (03:41 +0000)]
Add co_firstlineno and co_lnotab, for the line number table generated
by the compiler (in lieu of SET_LINENO instructions).

28 years agoInitialize the ob_type field of PyAST_Type dynamically (in
Guido van Rossum [Thu, 23 Jan 1997 23:29:44 +0000 (23:29 +0000)]
Initialize the ob_type field of PyAST_Type dynamically (in
initparser()) instead of statically (in the initializer).  The static
initialization, using the address of an object in a different DLL, is
too much for the Microsoft VC++ compiler, and we want to be able to
build this module as a separate DLL (it's nice to have but we don't
want to increase the core DLL's size by 25K).  This same trick has
been applied to a number of modules, e.g. NumPy and _tkinter.

28 years agoForget keeping track of whether a dictionary contains all interned
Guido van Rossum [Thu, 23 Jan 1997 19:39:29 +0000 (19:39 +0000)]
Forget keeping track of whether a dictionary contains all interned
string keys.  Just doing a pointer compare before the string compare
(in fact before the hash compare!) is just as fast.

28 years agoVarious notes added:
Guido van Rossum [Wed, 22 Jan 1997 21:00:32 +0000 (21:00 +0000)]
Various notes added:

- -mt option for Solaris threads with Solaris compiler
- make clean when switching static link status for Linux
- DEC alpha --with-dec-threads option
- SunOS removed unnecessary warnings, mention -Xa for SunPro
- Setup.local file
- warn to try make clean after changing readline option
- mention --with-threads as alias for --with-thread

28 years agoThe usual
Guido van Rossum [Wed, 22 Jan 1997 20:54:01 +0000 (20:54 +0000)]
The usual

28 years agoA number of new symbols have appeared (courtesy autoconf 2.12?)
Guido van Rossum [Wed, 22 Jan 1997 20:53:30 +0000 (20:53 +0000)]
A number of new symbols have appeared (courtesy autoconf 2.12?)

28 years agoAdd --with-dec-threads, for DEL Alpha, which requires -threads as the
Guido van Rossum [Wed, 22 Jan 1997 20:51:58 +0000 (20:51 +0000)]
Add --with-dec-threads, for DEL Alpha, which requires -threads as the
*final* linker argument.  This implies --with-thread.

Reimplemented --with-threads as an alias for --with-thread.

28 years agoAdded PyCObject_Import.
Guido van Rossum [Wed, 22 Jan 1997 20:48:48 +0000 (20:48 +0000)]
Added PyCObject_Import.

28 years ago(libparser.tex): Fix typo found by Mark Hammond.
Fred Drake [Wed, 22 Jan 1997 14:25:21 +0000 (14:25 +0000)]
(libparser.tex):  Fix typo found by Mark Hammond.

28 years agoBump version to 1.5a1.
Guido van Rossum [Wed, 22 Jan 1997 03:46:13 +0000 (03:46 +0000)]
Bump version to 1.5a1.

28 years agoBump install version to 1.5.
Guido van Rossum [Wed, 22 Jan 1997 03:45:38 +0000 (03:45 +0000)]
Bump install version to 1.5.

28 years agoRemove GetOwner, add GetBuildInfo...
Guido van Rossum [Wed, 22 Jan 1997 01:41:30 +0000 (01:41 +0000)]
Remove GetOwner, add GetBuildInfo...

28 years agoAdded getbuildinfo.c to the project.
Guido van Rossum [Wed, 22 Jan 1997 01:40:35 +0000 (01:40 +0000)]
Added getbuildinfo.c to the project.

28 years agoFix splitext() to go up to the last dot, not the first.
Guido van Rossum [Wed, 22 Jan 1997 00:17:26 +0000 (00:17 +0000)]
Fix splitext() to go up to the last dot, not the first.

28 years agoAvoid traceback in selection_own_get() when no Tk widget owns the
Guido van Rossum [Tue, 21 Jan 1997 23:22:03 +0000 (23:22 +0000)]
Avoid traceback in selection_own_get() when no Tk widget owns the
selection; return None instead.

28 years agoDarn. Updated version number for configure.in.
Guido van Rossum [Tue, 21 Jan 1997 22:03:53 +0000 (22:03 +0000)]
Darn.  Updated version number for configure.in.

28 years agoLet --with-threads be an alias for --with-thread.
Guido van Rossum [Tue, 21 Jan 1997 22:02:36 +0000 (22:02 +0000)]
Let --with-threads be an alias for --with-thread.

28 years agoKill all local variables on function return. This closes a gigantic
Guido van Rossum [Tue, 21 Jan 1997 21:18:36 +0000 (21:18 +0000)]
Kill all local variables on function return.  This closes a gigantic
leak of memory and file descriptors (thanks for Roj for reporting
that!).  Alas, the speed goes down by 5%. :-(

28 years agoDon't use static buffers internally for formatstring().
Guido van Rossum [Tue, 21 Jan 1997 16:12:09 +0000 (16:12 +0000)]
Don't use static buffers internally for formatstring().

28 years agoOnly call sigcheck() at the ticker code if we don't have true signals.
Guido van Rossum [Tue, 21 Jan 1997 06:15:24 +0000 (06:15 +0000)]
Only call sigcheck() at the ticker code if we don't have true signals.
This is safe now that both intrcheck() and signalmodule.c schedule a
sigcheck() call via Py_AddPendingCall().

This gives another 7% speedup (never run such a test twice ;-).

28 years agoArrange for PyErr_CheckSignals() to be called via Py_AddPendingCall().
Guido van Rossum [Tue, 21 Jan 1997 06:13:09 +0000 (06:13 +0000)]
Arrange for PyErr_CheckSignals() to be called via Py_AddPendingCall().
This avoids having to call sigcheck() (the same routine by its old
name :-) in the ticker code in ceval.c's main interpreter loop.

28 years agoIn the Unix version, arrange for sigcheck() to be called via
Guido van Rossum [Tue, 21 Jan 1997 06:00:33 +0000 (06:00 +0000)]
In the Unix version, arrange for sigcheck() to be called via
Py_AddPendingCall().  This avoids having to call sigcheck() in the
ticker code in ceval.c's main interpreter loop.

28 years agoCleanup:
Guido van Rossum [Tue, 21 Jan 1997 05:34:20 +0000 (05:34 +0000)]
Cleanup:
- fix bug in Py_MakePendingCalls() with threading
- fix return type of do_raise
- remove build_slice (same as PySlice_New)
- remove code inside #if 0
- remove code inside #ifdef CHECK_STACK
- remove code inside #ifdef SUPPORT_OBSOLETE_ACCESS
- comment about newimp.py should refer to ni.py

28 years agoget build info from elsewhere
Guido van Rossum [Mon, 20 Jan 1997 18:34:55 +0000 (18:34 +0000)]
get build info from elsewhere

28 years agowork build# into version string
Guido van Rossum [Mon, 20 Jan 1997 18:34:26 +0000 (18:34 +0000)]
work build# into version string

28 years agoBugfix: remove ref to fast before it's defined.
Guido van Rossum [Mon, 20 Jan 1997 04:29:16 +0000 (04:29 +0000)]
Bugfix: remove ref to fast before it's defined.

28 years agoChanges for frame object speedup:
Guido van Rossum [Mon, 20 Jan 1997 04:26:20 +0000 (04:26 +0000)]
Changes for frame object speedup:

- get fastlocals differently
- call newframeobject() with fewer arguments
- toss getowner(), which was unused anyway

28 years agoChanges that appear to give another 12% speedup.
Guido van Rossum [Mon, 20 Jan 1997 04:20:52 +0000 (04:20 +0000)]
Changes that appear to give another 12% speedup.

Rather than allocating a list object for the fast locals and another
(extensible one) for the value stack and allocating the block stack
dynamically, allocate the block stack with a fixed size (CO_MAXBLOCKS
from compile.h), and stick the locals and value stack at the end of
the object (this is now possible since the stack size is known
beforehand).  Get rid of the owner field and the nvalues argument --
it is available in the code object, like nlocals.

This requires small changes in ceval.c only.

28 years agoNew lay-out of frame object, for fewer mallocs.
Guido van Rossum [Mon, 20 Jan 1997 04:16:40 +0000 (04:16 +0000)]
New lay-out of frame object, for fewer mallocs.
See frameobject.c checkin message.

28 years agoRemove unused variable.
Guido van Rossum [Sat, 18 Jan 1997 20:04:05 +0000 (20:04 +0000)]
Remove unused variable.

28 years agoIntern the string "__complex__".
Guido van Rossum [Sat, 18 Jan 1997 08:04:16 +0000 (08:04 +0000)]
Intern the string "__complex__".

28 years agoIntern all names and varnames in newcodeobject(), plus those string
Guido van Rossum [Sat, 18 Jan 1997 08:02:57 +0000 (08:02 +0000)]
Intern all names and varnames in newcodeobject(), plus those string
literals that look like identifiers.  Also intern all strings used as
names during the compilation.

28 years agoIntern the various string objects created to speed up lookups.
Guido van Rossum [Sat, 18 Jan 1997 07:59:12 +0000 (07:59 +0000)]
Intern the various string objects created to speed up lookups.

28 years agoIntern the strings "__builtins__".
Guido van Rossum [Sat, 18 Jan 1997 07:58:41 +0000 (07:58 +0000)]
Intern the strings "__builtins__".

28 years agoIntern the strings created in getattr() and setattr().
Guido van Rossum [Sat, 18 Jan 1997 07:57:16 +0000 (07:57 +0000)]
Intern the strings created in getattr() and setattr().

28 years agoString interning.
Guido van Rossum [Sat, 18 Jan 1997 07:55:05 +0000 (07:55 +0000)]
String interning.

28 years agoBump API version due to the various speedup hacks.
Guido van Rossum [Sat, 18 Jan 1997 07:54:03 +0000 (07:54 +0000)]
Bump API version due to the various speedup hacks.

28 years agoTwo speedup hacks. Caching the hash saves recalculation of a string's
Guido van Rossum [Sat, 18 Jan 1997 07:53:23 +0000 (07:53 +0000)]
Two speedup hacks.  Caching the hash saves recalculation of a string's
hash value.  Interning strings (which requires hash caching) tries to
ensure that only one string object with a given value exists, so
equality tests are one pointer comparison.  Together, these can speed
the interpreter up by as much as 20%.  Each costs the size of a long
or pointer per string object.  In addition, interned strings live
until the end of times.  If you are concerned about memory footprint,
simply comment the #define out here (and rebuild everything!).

28 years agoAdd "if (x != NULL) continue;" (or similar for err==0) before the
Guido van Rossum [Sat, 18 Jan 1997 02:46:13 +0000 (02:46 +0000)]
Add "if (x != NULL) continue;" (or similar for err==0) before the
break to most cases, as suggested by Tim Peters.  This gives another
8-10% speedup.

28 years agoVersion 1.1. Fix memory leak and expensive comparison with None.
Guido van Rossum [Sat, 18 Jan 1997 02:20:37 +0000 (02:20 +0000)]
Version 1.1.  Fix memory leak and expensive comparison with None.

28 years agoFix the _setmode() patch for MS_WINDOWS: include <fcntl.h> and use
Guido van Rossum [Fri, 17 Jan 1997 22:05:38 +0000 (22:05 +0000)]
Fix the _setmode() patch for MS_WINDOWS: include <fcntl.h> and use
fileno(std*).

28 years agoRemove obsolete PyFrame_ExtendStack.
Guido van Rossum [Fri, 17 Jan 1997 22:03:32 +0000 (22:03 +0000)]
Remove obsolete PyFrame_ExtendStack.

28 years agoRemoved accessobject.c from file list.
Guido van Rossum [Fri, 17 Jan 1997 21:56:08 +0000 (21:56 +0000)]
Removed accessobject.c from file list.

28 years agoAdd stacksize argument to new.code().
Guido van Rossum [Fri, 17 Jan 1997 21:12:06 +0000 (21:12 +0000)]
Add stacksize argument to new.code().

28 years agoComment out extend_stack() -- it is no longer needed.
Guido van Rossum [Fri, 17 Jan 1997 21:07:57 +0000 (21:07 +0000)]
Comment out extend_stack() -- it is no longer needed.

28 years agoMarshal the new stacksize item in code objects.
Guido van Rossum [Fri, 17 Jan 1997 21:07:08 +0000 (21:07 +0000)]
Marshal the new stacksize item in code objects.

28 years agoWorking semaphore implementation by Sjoerd.
Guido van Rossum [Fri, 17 Jan 1997 21:06:41 +0000 (21:06 +0000)]
Working semaphore implementation by Sjoerd.

28 years agoNew MAGIC number (code objects have one more item when marshalled).
Guido van Rossum [Fri, 17 Jan 1997 21:06:11 +0000 (21:06 +0000)]
New MAGIC number (code objects have one more item when marshalled).

28 years agoUse the stack size from the code object and the CO_MAXBLOCKS constant
Guido van Rossum [Fri, 17 Jan 1997 21:05:28 +0000 (21:05 +0000)]
Use the stack size from the code object and the CO_MAXBLOCKS constant
from compile.h.  Remove all eval stack overflow checks.

28 years agoAdd co_stacksize field to codeobject structure, and stacksize argument
Guido van Rossum [Fri, 17 Jan 1997 21:04:03 +0000 (21:04 +0000)]
Add co_stacksize field to codeobject structure, and stacksize argument
to PyCode_New() argument list.  Move MAXBLOCKS constant to conpile.h.

Added accurate calculation of the actual stack size needed by the
generated code.

Also commented out all fprintf statements (except for a new one to
diagnose stack underflow, and one in #ifdef'ed out code), and added
some new TO DO suggestions (now that the stacksize is taken of the TO
DO list).

28 years agoAdd co_stacksize field to codeobject structure, and stacksize argument
Guido van Rossum [Fri, 17 Jan 1997 20:59:26 +0000 (20:59 +0000)]
Add co_stacksize field to codeobject structure, and stacksize argument
to PyCode_New() argument list.  Also add CO_MAXBLOCKS constant
indicating the maximum static nesting supported by the compiler.

28 years agoOf course, when the type of the argument to dis() is unsupported, it
Guido van Rossum [Fri, 17 Jan 1997 20:08:18 +0000 (20:08 +0000)]
Of course, when the type of the argument to dis() is unsupported, it
should raise TypeError, not ValueError...

28 years agoMore user friedly interface:
Guido van Rossum [Fri, 17 Jan 1997 20:05:04 +0000 (20:05 +0000)]
More user friedly interface:

dis() still disassembles the last frame of the lats stack trace.

dis(x) disassembles x, which may be a code object, function, or method.

disassemble(co, [lasti]) disassembles a code object; the lasti
argument is now optional.

disco(...) is an alias for disassemble(...), for backward compatibility.

28 years agoAdded Sjoerd's submitted 'ratecv' method. Made corrections for new names.
Roger E. Masse [Fri, 17 Jan 1997 18:12:04 +0000 (18:12 +0000)]
Added Sjoerd's submitted 'ratecv' method.  Made corrections for new names.
Reindented.

28 years agoMake as sure as we can that the forked child does not return and
Guido van Rossum [Fri, 17 Jan 1997 16:13:15 +0000 (16:13 +0000)]
Make as sure as we can that the forked child does not return and
continue the tests.

28 years agoRenamed, but not tested. Guido will you try your test script on this?
Roger E. Masse [Fri, 17 Jan 1997 16:08:55 +0000 (16:08 +0000)]
Renamed, but not tested.  Guido will you try your test script on this?

28 years agoRenamed.
Roger E. Masse [Fri, 17 Jan 1997 16:00:02 +0000 (16:00 +0000)]
Renamed.

28 years agoTest output for the SGI cl C module test script.
Roger E. Masse [Fri, 17 Jan 1997 15:59:16 +0000 (15:59 +0000)]
Test output for the SGI cl C module test script.

28 years agoLame test script for the SGI cl C module.
Roger E. Masse [Fri, 17 Jan 1997 15:58:35 +0000 (15:58 +0000)]
Lame test script for the SGI cl C module.

28 years ago(puremodule.c): New module which exports the Purify and Quantify C API
Barry Warsaw [Fri, 17 Jan 1997 00:01:33 +0000 (00:01 +0000)]
(puremodule.c): New module which exports the Purify and Quantify C API
to Python.  Minimal documentation is included in comments at the top
of the file, and in the Misc/PURIFY.README file.  Note that this
module must be statically linked since Pure doesn't provide shared
stubs libraries.

(Setup.in): Added commented template for pure module

(syslogmodule.c): ins() function wasn't declared static.

28 years agoREADME file for hints on Purify'ing or Quantify'ing the Python
Barry Warsaw [Thu, 16 Jan 1997 23:55:38 +0000 (23:55 +0000)]
README file for hints on Purify'ing or Quantify'ing the Python
interpreter.  It also mentions the soon to be checked in pure module.

28 years agoAlready renamed. Indented.
Roger E. Masse [Thu, 16 Jan 1997 22:05:33 +0000 (22:05 +0000)]
Already renamed.  Indented.

28 years agoOutput from test script for the bsddb C extension module.
Roger E. Masse [Thu, 16 Jan 1997 22:04:35 +0000 (22:04 +0000)]
Output from test script for the bsddb C extension module.

28 years agoTest script for the bsddb C extension module.
Roger E. Masse [Thu, 16 Jan 1997 22:04:10 +0000 (22:04 +0000)]
Test script for the bsddb C extension module.

28 years agoRewrote lookmapping() according to suggestions by Jyrki Alakuijala.
Guido van Rossum [Thu, 16 Jan 1997 21:06:45 +0000 (21:06 +0000)]
Rewrote lookmapping() according to suggestions by Jyrki Alakuijala.

28 years agoMinor output message change
Roger E. Masse [Thu, 16 Jan 1997 18:53:01 +0000 (18:53 +0000)]
Minor output message change

28 years agoMerge several mods:
Guido van Rossum [Thu, 16 Jan 1997 18:52:24 +0000 (18:52 +0000)]
Merge several mods:

- add opcodes BINARY_LSHIFT ... BINARY_OR

- remove RESERVE_FAST

- Skip M's suggestion for displaying which comparison operator is meant

28 years agoRenamed.
Roger E. Masse [Thu, 16 Jan 1997 17:10:22 +0000 (17:10 +0000)]
Renamed.

28 years agoAs per GvR recomendations, remove the restriction that binhex only run on
Roger E. Masse [Thu, 16 Jan 1997 16:51:57 +0000 (16:51 +0000)]
As per GvR recomendations, remove the restriction that binhex only run on
big-endian machines.  This is done by directing the struct module's pack
and unpack methods to treat the data always in bin endian format.
This has been tested on irix (big endian) and solaris x86 (little endian)
but not yet on the mac.

28 years agoset_key(): Use Py_CHARMASK macro to get the character from the array
Barry Warsaw [Thu, 16 Jan 1997 16:49:44 +0000 (16:49 +0000)]
set_key(): Use Py_CHARMASK macro to get the character from the array
unsigned.  This fixes the 8bit-char-in-key platform incompatibility.

I also removed the old backwards compatibility code, and the commented
lisp rotor code.  I retained the lisp docstrings as comments preceding
each function.

28 years agotest output for the binascii C module test script.
Roger E. Masse [Thu, 16 Jan 1997 16:44:40 +0000 (16:44 +0000)]
test output for the binascii C module test script.

28 years agotest script for the binascii C module.
Roger E. Masse [Thu, 16 Jan 1997 16:44:09 +0000 (16:44 +0000)]
test script for the binascii C module.

28 years agoreformatted
Barry Warsaw [Thu, 16 Jan 1997 00:15:11 +0000 (00:15 +0000)]
reformatted

memory leak in Tkapp_(Create|Delete)FileHandler plugged.

standard eyeballing

28 years agotest is no longer needed
Barry Warsaw [Thu, 16 Jan 1997 00:13:11 +0000 (00:13 +0000)]
test is no longer needed

28 years agoOpps. Added tests for the last two methods that were missing from
Roger E. Masse [Wed, 15 Jan 1997 20:58:55 +0000 (20:58 +0000)]
Opps.  Added tests for the last two methods that were missing from
 the test program: 'grey2rgb' and 'rgb2grey'