Mailman 3 September 2000 - Pythonmac-SIG

Re: [Pythonmac-SIG] Embedding Python w/o SIOUX popping up
by Jack Jansen April 18, 2001

April 18, 2001

> I am embedding Python in a Code Resource that loads as an extension to the > program Igor. My extension loads the PythonCore Shared Library. When I > call my command from the Igor command window and the code in my extension > calls PyMac_Initialize() a SIOX window pops up and another apple menu is > added at the top of the screen. There is porbably some way to use the > funcitonality of PythonCode without loading the Window/menubar, and if > there is, I would love any advice anyone may have on this > situation. Thanks. Your application needs python preferences resources (a "Popt" and a "GU<sigma> I" resource). The easiest way to get these is dropping your application on EditPythonPrefs. An alternative (often handier because you don't have to do it after every build) is to create a resource file that includes the Popt and GUSI resource and include this in your build. -- Jack Jansen | ++++ stop the execution of Mumia Abu-Jamal ++++ Jack.Jansen(a)oratrix.com | ++++ if you agree copy these lines to your sig ++++ www.oratrix.nl/~jack | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm

3 3

Re: [Pythonmac-SIG] Re: Pythonmac-SIG digest, Vol 1 #455 - 5 msgs
by Anthony Ingraldi March 6, 2001

March 6, 2001

> B) is the TK stuff needed to use Gnuplot, or is just what you used to > create your "GUI-driven" solution. > You can drive Mac Gnuplot from python via the "standard" Gnuplot.py module. The module is available at http://monsoon.harvard.edu/~mhagger/Gnuplot/Gnuplot.html. The latest version includes support for the Mac. Also, the Mac gnuplot distribution contains a simple console written in python. You can obtain the Mac port of gnuplot at http://users.ece.gatech.edu/~schooley/gnuplot.html. It looks like it hasn't been updated in over 2 years. -- Tony Ingraldi A.M.INGRALDI(a)LaRC.NASA.GOV Phone : (757) 864-3039 Fax : (757) 864-7892

2 1

State of BBPy? More general Tkinter development question.
by Russell E Owen Oct. 17, 2000

Oct. 17, 2000

I have gotten BBPy to work with Pyton 1.6a2 (but not 1.5.2) by changing "from Types.py import *" to "from types.py import *" (i.e. types must be lower case) in PythonSlave.py. One simple script I tried ran fine. However, when I tried to run a Tkinter script I'm working on Python hung and so did BBEdit. (In general BBEdit seems susceptible to hanging if Python hangs when using BBPy). Is this a known limitation in BBPy, that it fails with Tkinter scripts? Or did I just get unlucky and have something fail, or what? Any suggestions for a good environment in which to build Tkinter applications would be very much appreciated. It doesn't have to be on the Mac. The present business of editing the file, dropping it onto PythonInterpreter, having it choke, analyzing the output and doing it all over again is pretty painful. Also, any suggestions on ways to interact with a Tkinter script while building it would be appreciated. I'd like to start things up and then be able to talk to my widgets and generally exercise my code interactively. Regards, -- Russell

2 1

Re: [Pythonmac-SIG] Line endings and readline()
by Jack Jansen Oct. 5, 2000

Oct. 5, 2000

Recently, Chris Barker <cbarker(a)jps.net> said: > Andrew Watson wrote: > > Is there some easy way of convincing python to change what it expects as th > e > > end-of-line character, to deal with this sort of crodd-platform problem? > > No, and I wish there was. There are some cases where it can be useful, but in general you should leave this to whoever did the transfer from mac to unix or vice-versa. After all Python isn't the only Unix program that'll have trouble with Mac line endings: all unix programs will. FTP clients, archivers, and remote file server packages will usually do the conversion for you if you set them up correctly. And on the Mac this should be a breeze because all files of type TEXT should be translated (unix and windows programs will have to gues whether a file is text based on the extension, so you'll have to do the correct setup there yourself). A quick trick for converting any known convention to \n convention is the following (if you can afford reading your whole file into memory) data = f.read() data = string.replace(data, '\r\n', '\n') data = string.replace(data, '\r', '\n') It'll even work for confused files, as you can often find in webpages and such that have been edited on multiple systems. The only thing it won't handle is files with \n\r (which can happen if you send a file back and forth between all three platforms and don't get the conversion right), but this is unsolvable anyway. -- Jack Jansen | ++++ stop the execution of Mumia Abu-Jamal ++++ Jack.Jansen(a)oratrix.com | ++++ if you agree copy these lines to your sig ++++ www.oratrix.nl/~jack | see http://www.xs4all.nl/~tank/spg-l/sigaction.htm

3 3

W applications
by Tattoo Mabonzo K. Oct. 4, 2000

Oct. 4, 2000

Yo, Is it correct that the fact that W relies on globals such as _application and _signature makes it (essentially) impossible to have several W-based application running at the same time? TIA for any input! = tmk =

3 2

Re: Python on Mac OS X w/shared modules
by Martin v. Loewis Sept. 30, 2000

Sept. 30, 2000

> Anybody have any ideas what ___eprintf is for? I'm using the > development tools copied from Darwin; does anybody here have "real" > Mac OS X Developer Tools, and if so does this error happen for them? __eprintf is used inside the assert macro, and defined in libgcc2.c: #include <stdio.h> /* This is used by the `assert' macro. */ void __eprintf (const char *string, const char *expression, unsigned int line, const char *filename) { fprintf (stderr, string, expression, line, filename); fflush (stderr); abort (); } It then appears that you are not linking with libgcc. Anyway, thanks for the information. Regards, Martin

1 0

Python on Mac OS X w/shared modules
by Tony Lownds Sept. 30, 2000

Sept. 30, 2000

Hi, I've *almost* got Python working with shared modules, by using --with-next-framework and hacking the configure script to recognize that Mac OS X needs Next-like settings. This is the command I used, and the patch is below. ./configure --with-dyld --with-next-framework --with-suffix=.x Unfortunately it bombs out with a missing symbol in libtool: libtool -o libpython2.0.dylib -dynamic libpython2.0.a -framework System -U _environ ld: for architecture ppc ld: Undefined symbols: ___eprintf libtool: internal link edit command failed make[1]: *** [libpython2.0.dylib] Error 1 make: *** [python.x] Error 2 If I ask the linker to ignore that symbol, it doesn't work either: [localhost:~/python/dist/src] tony% libtool -o libpython2.0.dylib -dynamic libpython2.0.a -framework System -U _environ -U ___eprintf [localhost:~/python/dist/src] tony% ./python.x dyld: ./python.x Undefined symbols: ___eprintf Anybody have any ideas what ___eprintf is for? I'm using the development tools copied from Darwin; does anybody here have "real" Mac OS X Developer Tools, and if so does this error happen for them? Thanks, -Tony Lownds *** configure Fri Sep 29 17:08:15 2000 --- configure.hacked Fri Sep 29 17:58:41 2000 *************** *** 596,601 **** --- 596,608 ---- sed -e 's/://' -e 's/\./_/'` && \ ac_sys_system=next && ac_sys_release=$4 + if test -z "$ac_sys_release" + then + set X `hostinfo | egrep 'Darwin Kernel.*:' | \ + sed -e 's/://'` && \ + ac_sys_system=darwin && ac_sys_release=$5 + fi + MACHDEP="$ac_sys_system$ac_sys_release" fi fi *************** *** 2583,2588 **** --- 2590,2598 ---- # Minor variations in building a framework between NextStep versions 4 and 5 case $ac_sys_system/$ac_sys_release in + darwin/1.2) + ns_undef_sym='_environ' + LIBTOOL_CRUFT="-U $ns_undef_sym" ;; next/4*) ns_undef_sym='__environ' LIBTOOL_CRUFT="-U $ns_undef_sym" ;; *************** *** 2666,2672 **** hp*|HP*) LDSHARED="ld -b";; OSF*) LDSHARED="ld -shared -expect_unresolved \"*\"";; DYNIX/ptx*) LDSHARED="ld -G";; ! next/*) if test "$ns_dyld" then LDSHARED='$(CC) $(LDFLAGS) -bundle -prebind' else LDSHARED='$(CC) $(CFLAGS) -nostdlib -r'; --- 2676,2682 ---- hp*|HP*) LDSHARED="ld -b";; OSF*) LDSHARED="ld -shared -expect_unresolved \"*\"";; DYNIX/ptx*) LDSHARED="ld -G";; ! darwin/*|next/*) if test "$ns_dyld" then LDSHARED='$(CC) $(LDFLAGS) -bundle -prebind' else LDSHARED='$(CC) $(CFLAGS) -nostdlib -r'; *************** *** 2741,2746 **** --- 2751,2758 ---- # crt1.o) gets erroneously defined as common, which breaks dynamic # loading of any modules which reference it in System.framework next/4*|next/5*) LINKFORSHARED="-u __dummy -framework System" ;; + # darwin doesn't like __dummy + darwin/*) LINKFORSHARED="-framework System" ;; SCO_SV*) LINKFORSHARED="-Bdynamic -dy -Wl,-Bexport";; ReliantUNIX*) LINKFORSHARED="-W1 -Blargedynsym";; FreeBSD*|NetBSD*) *************** *** 3882,3888 **** AIX*) DYNLOADFILE="dynload_aix.o";; BeOS*) DYNLOADFILE="dynload_beos.o";; hp*|HP*) DYNLOADFILE="dynload_hpux.o";; ! next/*) DYNLOADFILE="dynload_next.o";; *) # use dynload_shlib.c and dlopen() if we have it; otherwise stub # out any dynamic loading --- 3894,3900 ---- AIX*) DYNLOADFILE="dynload_aix.o";; BeOS*) DYNLOADFILE="dynload_beos.o";; hp*|HP*) DYNLOADFILE="dynload_hpux.o";; ! darwin/*|next/*) DYNLOADFILE="dynload_next.o";; *) # use dynload_shlib.c and dlopen() if we have it; otherwise stub # out any dynamic loading *************** *** 5461,5467 **** # check for --with-libm=... case $ac_sys_system in ! next) ;; BeOS) ;; *) LIBM=-lm esac --- 5473,5479 ---- # check for --with-libm=... case $ac_sys_system in ! darwin|next) ;; BeOS) ;; *) LIBM=-lm esac

1 0

FW: Python2.0 beta 2 on MacOSX/Darwin
by Jonathan Wight Sept. 29, 2000

Sept. 29, 2000

Forwarding to the PythonMac SIG... > > ------ Forwarded Message > From: "Steven D. Majewski" <sdm7g(a)minsky.med.virginia.edu> > Date: Fri, 29 Sep 2000 12:07:35 -0400 (EDT) > To: Darwin-Development(a)public.lists.apple.com > Subject: Python2.0 beta 2 on MacOSX/Darwin (fwd) > > > > > [ Same message being forwarded (NOT cross-posted!) to: > python-list(a)python.org > macosx-dev(a)omnigroup.com > Darwin-Development(a)public.lists.apple.com ] > > > If you want to do Mac OSX development without the developers CD, > "Darwin Tools on Mac OSX" at <http://www.darwinfo.org/> > has a link to the article: > <http://www.macaddict.com/content/news/2000/09/13/18709> > containing instructions and a shell script to merege the > Darwin development tools from the Darwin 1.0 distribution > into MacOSX. > > > Python 2.0 beta 2 > <http://www.beopen.com/products/python2.0/BeOpen-Python-2.0b2.tar.gz> > builds on OSX or Darwin > > configure options: > > --with-dyld # otherwise you get errors trying to include rld.h > # I think thats for the old style next libraries. > > --with-suffix=.app # or ".exe" or whatever > # this solves the case insensitivity problem on HFS+ > # The only clash is when make tries to move the > # executable in "Python/python" to the parent dir. > # this renames the executable with a suffix. > > --host=Darwin # Not sure about this, but it may have stopped some > # errors with trying to build Next platform specific code. > > ( maybe that's not doing what I thought as sys.platform still= 'next') > > > > > Modules: > > bsdbm gets automagically added by configure (See notes in Modules/Setup) > > threads are also automatically configured by default. > > crypt can be added in Modules/Setup -- no added library switches needed. > > pyexpat: expat will build, but building a library using 'ar' , as in > the note in Modules/Setup will give link errors. -- Use libtool instead: > "libtool -static -o libexpat.a xmltol/*.o xmlparse/*.o" > > (CoreFoundation, which is one of the OSX Frameworks also included in > Darwin, also has functions to convert to and from XML property lists, > but I don't believe that it exposes a full XML parser api. ) > > > > Limitations: > > tests seem to run ok except that I get a : > test_re > Running tests on re.search and re.match > Running tests on re.sub > Running tests on symbolic references > Running tests on re.subn > Running tests on re.split > Running tests on re.findall > Running tests on re.match > Running tests on re.escape > Pickling a RegexObject instance > Test engine limitations > Segmentation fault > > > > using configure --with-next-framework along with --with-dyld didn't work. > (Don't remember the error -- I'll try again) > > Just uncommenting the "*shared*" in Modules/Setup gives this build error: > > cc -bundle -prebind arraymodule.o -o arraymodule.so > /usr/bin/ld: warning -prebind has no effect with -bundle > /usr/bin/ld: Undefined symbols: > _PyArg_Parse > _PyArg_ParseTuple > _PyDict_SetItemString > _PyErr_BadArgument > _PyErr_Clear > _PyErr_Format > _PyErr_NoMemory > _PyErr_Occurred > > > > To Do: > > Need to get dynamic loading of shared libraries working. > ( dynamic loading + threading + pyexpat => Zope ought to port! ) > > Then need to work on Darwin|MacOSX specific libraries. > > Tkinter, pyGtk and/or wxWindows under XFREE86 would be nice. > gtk has been ported to Darwin/XFREE86 > Tcl has been ported but not, as far as I know Tk. > Is the Mac port of Tk carbonized ? If it were, it would allow > Tk on both Darwin under X11 and MacOSX under Carbon. > ( We really want Quartz, but it would be nice to have something > that runs anywhere, and TK is probably the closest thing. > It was kind of strange to type "open PythonInterpreter" cd-ed to > the OS9 MacPython folder and see OS9 boot up and run Python. > I get an error message that my preferences are corrupt -- > I wonder if this is from the OSX install on top of OS9 > moving the system folder. But otherwise, things seem to > work -- I only tried EasyDialogs.test() and started a Tkinter > window -- not a very complete test. ) > > > > > > ---| Steven D. Majewski (804-982-0831) <sdm7g(a)Virginia.EDU> |--- > ---| Department of Molecular Physiology and Biological Physics |--- > ---| University of Virginia Health Sciences Center |--- > ---| P.O. Box 10011 Charlottesville, VA 22906-0011 |--- > "All operating systems want to be unix, > All programming languages want to be lisp." > > > _______________________________________________ > MacOSX-dev mailing list > MacOSX-dev(a)omnigroup.com > http://www.omnigroup.com/mailman/listinfo/macosx-dev > > > > ==================================================================== > Unsubscribe: <darwin-development-unsubscribe(a)public.lists.apple.com> > > Help: <http://www.lists.apple.com/darwin-development.html> > or <darwin-development-info(a)public.lists.apple.com> > > ------ End of Forwarded Message

1 0

PythonStandAlone + Numeric
by David Andersen Sept. 28, 2000

Sept. 28, 2000

I'm building PythonStandAlone and PythonInterpreterPPC from the 1.5.2c1 source - things basically seem to work, but I'm not out of the woods yet. The PythonStand Alone application is built to the Python 1.5.2c1 folder. In PythonInterpreterPPC, "import Numeric" works correctly, so it appears the Python installation and preferences are correct. However, when I try "import Numeric" in PythonStandAlone I get the following traceback: File "<stdin>", line 1, in ? File "Trollhaus:Python 1.5.2c1:Extensions:Numerical:Lib:Numeric.py", line 6,in ? import multiarray ImportError: No module named multiarray Why would "import Numeric" work in PythonInterpreterPPC, but not in PythonStandAlone? Do I need to build Numeric into PythonStandAlone? If so, how do I do that?

1 0

Re-- AppleScript guru needed
by Vincent Marchetti Sept. 26, 2000

Sept. 26, 2000

IANAG but, Jack Jensen asked about legality of having classes and properties sharing the same name in AppleScript. I don't know if is legal (or even if there is a "controlling legal authority" for AppleScript and OSA naming), but I always had trouble trying to do this in scriptable applications I have written. It seems to me the problem is how the AppleScript interpreter converts a phrase like "get project document of target 1" to an object specifier. My experience is that in this case the correct specifier is made when "project document" is both the name of a class and the name of a property (of class target); however the interpreter gags when trying to make a more complex object reference, for example using whose -clauses i.e. "get targets whose (name of project document is "parrot") " In short, within AppleScript and OSA, sharing names might be legal, but not very robust. However, it seems that MetroWerks did it in setting up the scripting interface to CodeWarrior IDE. Under MacPython, the script writer should know whether the class project_document or the property project_document is required. It seems to me the best solution is to rename the Python class property to project_document_property, and a corresponding change to element target._propdict . As for the call to make, I've been using simply app.make(new = classname.want, with_properties = etc..) Vince Marchetti

1 0