AFAIK you don't need to do anything special (i.e. write "ctypes
definitions") to call the CF functions with ctypes. Just call them.
Here's a snippet of code that I use to open a PDF document and get
the number of pages:
from ctypes import cdll, byref, c_void_p, c_int, c_uint, c_float,
c_double
from ctypes.util import find_library
cf = cdll.LoadLibrary(find_library("ApplicatonServices"))
filename = "/path/to/document.pdf"
kCFAllocatorDefault = None
kCFURLPOSIXPathStyle = 0
# open pdf document
fileref = None
fileurl = None
try:
fileref = cf.CFStringCreateWithCString(None, filename, 0)
fileurl = cf.CFURLCreateWithFileSystemPath(
kCFAllocatorDefault,
fileref,
kCFURLPOSIXPathStyle,
0
)
pdf = cf.CGPDFDocumentCreateWithURL(fileurl)
if not pdf:
raise PrintError("cannot open PDF document: %s" % filename)
finally:
if fileurl:
cf.CFRelease(fileurl)
if fileref:
cf.CFRelease(fileref)
numpages = cf.CGPDFDocumentGetNumberOfPages(pdf)
if pdf:
cf.CGPDFDocumentRelease(pdf)
My experience with developing code like this is that it takes a lot
of patience and trial and error to get it stable. Remember that
you're dealing with C functions and pointers (it's called ctypes for
a reason), which are not nearly as forgiving as normal Python
functions and objects. For example, calling a C function with the
wrong argument types will generally cause a cold hard segfault rather
than give you a nice stack trace with some indication of what caused
the error. When all else fails read the documentation for the library
you're calling...again.
~ Daniel
_______________________________________________
Pythonmac-SIG maillist - Pythonmac-SIG@python.org
http://mail.python.org/mailman/listinfo/pythonmac-sig
