🌐 Mailman 3 July 2007 - Pythonmac-SIG - python.org

py2app warning: may need to include email.Util
by Russell E Owen Aug. 6, 2007

Aug. 6, 2007

My py2app-bundled Mac application recently started failing with "no module named email.Util" when trying to display images. It turns out pyfits uses urllib which uses email.Util, which apparently has been renamed email.util with some kind of clever lazy import code so one can still use email.Util. Apparently the cleverness is too much for py2app's module finder. Anyway, the easy workaround is to specify email.Util as a necessary module to import. But I'll report a python bug or submit a urllib patch. -- Russell

5 6

Accessing Resource forks
by Donna and Chris Barker July 30, 2007

July 30, 2007

HI all, Somehow over all these years, I've managed to use Python on MAcs without having it deal with resource forks and sll that -- but ironically, not that we're all OS-X all the time, I need to do ti now. What I'm trying to do is simple check for the existance of a resource fork, and if it's there, the size. This is what I've tried: import Carbon.File as File files = ["junk.txt", "junk_rsc.txt"] for filename in files: f = File.FSRef(filename) print "file path is:", f.FSRefMakePath() i = f.FSGetCatalogInfo(0) print "logical size is:", i[0].rsrcLogicalSize print "Physical size is:", i[0].rsrcPhysicalSize In this case, junk.txt was created at the command line, so it shouldn't have a resource fork. junk_rsc.txt I opened up in resedit, which should have given it a small resource fork. Runnign this script, I get: file path is: /Users/cbarker/temp/AFP-SMB/junk.txt logical size is: 65543 Physical size is: 30457855 file path is: /Users/cbarker/temp/AFP-SMB/junk_rsc.txt logical size is: -593231776 Physical size is: -210763775 The paths are right, so it looks like I've got the FSRef right, but rsrcLogicalSize and rsrcPhysicalSize make no sense. How should I be doing this -- the docs are sparse, to say the least! By the way, I might as well tell you the real goal, maybe one of you will have a better idea. We have a Windows file server that is about to be retired. over the years, people of have put files on it using both the AFP and SMB protocols (the server is running MS services for macintosh). These are all mixed up. However, any file put up with one protocol loses it's resource fork (and type and creator) if brought down with the other protocol. Our goal is to clean up this mess automatically. The idea at hand is that a file brought down with AFP will either: Have a nice resource fork, in which case it was put up with AFP, and we're happy Have no resource fork, in which case it never had one, and we dont' care what protocol was used, or it was put up with SMB, and we can then bring it down that way instead. The other obvious option is to look for the ._* files, which is where the resource fork is stored with the SMB protocol. However, we're concerned that that may not be reliable -- if a file were put up with SMB, then replaced with AFP, there may be a ._* file, but it won't work right. This actually seems pretty likely as while we have this mixed system, there have been a lot of "that didn't work, please put the file back up with APF" interations. Thanks for your thoughts. -Chris

3 2

Compiling cx_Oracle
by stv July 27, 2007

July 27, 2007

I am trying to compile cx_Oracle for my Intel Mac 10.4.9 / Python 2.4.4 installation, but when I import the compile .so file I get ImportError: Failure linking new module (Detail below). I have the Oracle instant client installed, including the SDK package for the header files. In order to pacify the compiler warnings, I created symbolic links for the two dylibs that ended in 10.1 /Applications/instantclient/libclntsh.dylib@ -> libclntsh.dylib.10.1 /Applications/instantclient/libocci.dylib@ -> libocci.dylib.10.1 The following builds the cx_Oracle.so and installs it into site-packages, but the import fails as follows: $ sudo python setup.py install running install running build running build_ext running install_lib $ python Python 2.4.4 (#1, May 21 2007, 11:26:22) [GCC 4.0.1 (Apple Computer, Inc. build 5367)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> import cx_Oracle Traceback (most recent call last): File "<stdin>", line 1, in ? ImportError: Failure linking new module: /opt/local/Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site-packages/cx_Oracle.so: Symbol not found: _OCINumberFromInt Referenced from: /opt/local/Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site-packages/cx_Oracle.so Expected in: dynamic lookup >>> I've stepped out of my aging compiler expertis.... er, memories. Dynamically linked libraries, statically linked libraries, it's all a blur now. Tis what I love about Python : ) Any tips are appreciated. I found only one page Googling about (which gave me symbolic link tip) http://pedro.emanuel.familiasalgado.org/?p=103 --stv

3 2

PIL (1.1.6) working with py2app on intel macs?
by Dav Clark July 27, 2007

July 27, 2007

Hi there, I have a properly working copy of PIL 1.1.6 - I used this generic recipe: http://wiki.python.org/moin/MacPython/UniversalLibrariesAndExtensions I can load jpegs and see them in preview with im.show(). I installed py2app via easy_install, and also by downloading the tarball and installing... I see old posts about rebuilding the bootstrap, but no idea if this is still current, and if so what the bootstrap even _is_, let alone how to build it. But when I run an 'app', I still get the ol': ImportError: The _imaging C module is not installed Under Contents/Resources/lib/python2.3/lib-dynload/, I see both a _imaging.so AND a PIL/_imaging.so (among other things). I am using the system python2.3. Is anyone successful with this configuration? Any help would be appreciated! Dav Clark

2 5

Easy way to avoid Fatal Python error: Interpreter not initialized (version mismatch?)
by Ivan Vilata i Balaguer July 26, 2007

July 26, 2007

I've been looking for some hours for a more-or-less easy solution or workaround to the Fatal Python error: Interpreter not initialized (version mismatch?) Abort trap error that didn't imply uninstalling additional Pythons and the like, so I'm posting this here so at least people can find it when googling for it. ;) Quite surely not the best solution, but it looks quite simple. The problem comes from an extension being linked against the *default* Python (i.e. the one which pops up when running ``python``) instead of the requested one. When using GCC with ``-framework Python``, it accesses the Python framework in ``/Library/Frameworks/Python.framework``, and it uses some components right under it which happen to be links to paths under ``Versions/Current``, which is linked to the default version. In fact, when you inspect the extension which fails to load with ``otool -L EXTENSION.so`` you can see it's linked with the wrong Python library. The workaround for this is kludgy as I warned but at least is quite simple. It consists in *changing the default Python version* before building the failing extension:: $ cd /Library/Frameworks/Python.framework/Versions $ sudo rm Current && sudo ln -s 2.4 Current $ cd your_project_path $ python2.4 setup.py build_or_whatever Of course, change "2.4" to whatever Python version you need -- as long as it is installed. ;) If you want to recover your default version, simply run the first two lines with the proper version number. Just my 2 euro cents, :: Ivan Vilata i Balaguer >qo< http://www.carabos.com/ Cárabos Coop. V. V V Enjoy Data ""

2 2

Trouble installing Python Imaging Library on Mac OS X
by Ritesh Nadhani July 24, 2007

July 24, 2007

Hello So I want to install Python Imaging Library on my Mac OS X. I followed the tutorial steps by step. When I executed: IOError: decoder jpeg not available I got the error. Further Googling pointed me to the website: http://www.kyngchaos.com/macosx/install/libjpeg I downloaded the libtool updated linked there and unzipped it. It says that: Then, you need to update the libtool file in the source. Though I have no clue what does that mean. Unzipping it gave me two file ltconfig and ltconfig.sh. Running the sh file gave me libtool version mismatch error. Doing a search of system I found that there is another ltconfig in /usr/share/libtool. Assuming that this file needs to be updated with the one provided in the zip, I copied the original ltconfig.sh to ltconfig.sh.bak and copied the new one to the folder. The I followed the same steps as mentioned up there and the compilation and installation seems to have gone correctly. But again when I do selftest.py, I get: Exception raised: Traceback (most recent call last): File "./doctest.py", line 499, in _run_examples_inner exec compile(source, "<string>", "single") in globs File "<string>", line 1, in <module> File "./selftest.py", line 22, in _info im.load() File "PIL/ImageFile.py", line 180, in load d = Image._getdecoder(self.mode, d, a, self.decoderconfig) File "PIL/Image.py", line 375, in _getdecoder raise IOError("decoder %s not available" % decoder_name) IOError: decoder jpeg not available 1 items had failures: 1 of 57 in selftest.testimage ***Test Failed*** 1 failures. *** 1 tests of 57 failed. Moreover, sudo python setup.py build_ext -i, give me: -------------------------------------------------------------------- PIL 1.1.6 BUILD SUMMARY -------------------------------------------------------------------- version 1.1.6 platform darwin 2.5 (r25:51918, Sep 19 2006, 08:49:13) [GCC 4.0.1 (Apple Computer, Inc. build 5341)] -------------------------------------------------------------------- --- TKINTER support ok --- JPEG support ok --- ZLIB (PNG/ZIP) support ok --- FREETYPE2 support ok -------------------------------------------------------------------- To check the build, run the selftest.py script. So it seems that the libraries were correctly found. -- Ritesh http://www.riteshn.com

3 7

Semi-standalone Qt py2app can't find qt.so
by Ivan Vilata i Balaguer July 22, 2007

July 22, 2007

Hi people, I'm attaching an utterly trivial project (the hello world example from the PyQt tutorial) which shows that py2app-generated *semi-standalone* apps using Qt fail to run, apparently because of the path to the Qt libs included in the appdir not being placed in ``sys.path``. To check this yourselves, simply place ``MyApp.py`` and ``setup.py`` in the same directory, run ``python setup.py py2app`` and try to open the resulting app bundle which is left in the ``dist`` directory with Finder. This is the error dialog that I get:: MyApp Error MyApp Error An unexpected error has occurred during execution of the main script ImportError: '/Library/Frameworks/Python.framework/Versions/2.5/lib/ python2.5/lib-dynload/qt.so' not found And these are the console messages:: Traceback (most recent call last): File "/Users/ivan/MyApp/dist/MyApp.app/Contents/Resources/__boot__.py", line 31, in <module> _run('MyApp.py') File "/Users/ivan/MyApp/dist/MyApp.app/Contents/Resources/__boot__.py", line 28, in _run execfile(path, globals(), globals()) File "/Users/ivan/MyApp/dist/MyApp.app/Contents/Resources/MyApp.py", line 6, in <module> import qt File "qt.pyc", line 18, in <module> File "qt.pyc", line 15, in __load ImportError: '/Library/Frameworks/Python.framework/Versions/2.5/lib/ python2.5/lib-dynload/qt.so' not found 2007-07-20 20:11:48.233 MyApp[286] MyApp Error 2007-07-20 20:11:48.235 MyApp[286] MyApp Error An unexpected error has occurred during execution of the main script ImportError: '/Library/Frameworks/Python.framework/Versions/2.5/lib/ python2.5/lib-dynload/qt.so' not found Please note how the app is trying to load ``qt.so`` *from the system* instead of from the appdir. Running ``python MyApp.py`` works fine. I'm using Universal MacPython 2.5, PyQt 3.17 and py2app 0.3.6. :: Ivan Vilata i Balaguer >qo< http://www.carabos.com/ Cárabos Coop. V. V V Enjoy Data ""

3 3

NSKeyedArchiver and NSMutableArray
by Jim Matthews July 17, 2007

July 17, 2007

I'm trying to learn PyObjC by translating the examples in Aaron Hillegass's "Cocoa Programming for Mac OS X" to Python. I've run into a problem with the archiving feature described in chapter 8. I have an NSMutableArray of Person objects, where Person is a subclass of NSObject. I can turn the array into an NSData object with NSKeyedArchiver.archvedDataWithRootObject_(), but when I call NSKeyedUnarchiver.unarchiveObjectWithData_() I get back an empty array. Any suggestions? Here's simplified code that demonstrates the problem: from Foundation import * class Person(NSObject): def init(self): self = super(Person, self).init() if self is None: return self self.setPersonName_(u"New Person") return self def initWithCoder_(self, inCoder): self = super(Person, self).init() if self is None: return self self.setPersonName_(inCoder.decodeObjectForKey_('personName')) print 'Person.initWithCoder_ personName =', self.personName() def encodeWithCoder_(self, inCoder): print 'Person.encodeWithCoder_ personName =', self.personName() inCoder.encodeObject_forKey_(self.personName(), 'personName') def description(self): return super(Person, self).description() + '(' + self.personName() + ')' def personName(self): return self._personName def setPersonName_(self, inPersonName): self._personName = inPersonName ma = NSMutableArray.alloc().init() p1 = Person.alloc().init() p1.setPersonName_('John Doe') ma.insertObject_atIndex_(p1, 0) print 'ma =', ma.description() madata = NSKeyedArchiver.archivedDataWithRootObject_(ma) decodedma = NSKeyedUnarchiver.unarchiveObjectWithData_(madata) print 'decodedma =', decodedma.description() -- Expected output: ma = (<Person: 0x62e810>(John Doe)) Person.encodeWithCoder_ personName = John Doe Person.initWithCoder_ personName = John Doe decodedma = (<Person: 0x62e810>(John Doe)) Seen: ma = (<Person: 0x62e810>(John Doe)) Person.encodeWithCoder_ personName = John Doe Person.initWithCoder_ personName = John Doe decodedma = () Thanks for any suggestions, -- Jim Matthews Fetch Softworks http://fetchsoftworks.com

1 1

Re: [Pythonmac-SIG] NSKeyedArchiver and NSMutableArray
by Nehemiah Dacres July 17, 2007

July 17, 2007

if im not mistaken, coder is a parameter also. I don have my book with me, if it works in objective c post the equivilant code in objc and i can work from there. I remember most of this excersize. I thought coder is the datatype of the object you're coding fore (id). "lalalalala! it's not broken because I can use it" http://linux.slashdot.org/comments.pl?sid=194281&threshold=1&commentsort=0&…

1 0

Re: [Pythonmac-SIG] Accessing Resource forks
by Christopher Barker July 17, 2007

July 17, 2007

Hamish Allan wrote: > Let me start by saying that I've never used MS services for Macintosh, > so this is all guesswork. But the first thing I wondered was, if files > uploaded over AFP aren't using ._* files, what are they using? The server is storing the resource forks somewhere, but not anywhere we can see them - who knows? > Presumably the Windows machine isn't actually mounting an HFS or HFS+ > filesystem, so they're being stored somewhere "non-standard", in which > case all bets for using FSRef are off anyway. I'm only trying to use those after downloading the files via AFP. In my test code, I'm using files created on the Mac they're being tested on, so it should work, if I knew how. > Proceeding with this assumption, could you perhaps compare the > modification times of the ._* files with the files for which they > store the resource fork, and if the latter has a more recent > modification time the chances are that it has been overwritten using > AFP? Good idea, why didn't I think of that? > That is to say, there are three cases: > > 1) No ._* file is present: Either the resource fork is stored in > non-standard place due to AFP or the file never had one -- download > via AFP > 2) ._* file is present and has same modification time as its > counterpart -- download via SMB > 3) ._* file is present and has later modification time than its > counterpart -- download via AFP Yup, that looks like a good option. I'll give it a try. I'd still like to know how to get info in the Resource fork, however... Thanks, -Chris -- Christopher Barker, Ph.D. Oceanographer Emergency Response Division NOAA/NOS/OR&R (206) 526-6959 voice 7600 Sand Point Way NE (206) 526-6329 fax Seattle, WA 98115 (206) 526-6317 main reception Chris.Barker(a)noaa.gov

2 2