Issue 29 in pylibnidaqmx: nidaqmx.h not available when using the core runtime

[pylibn...@googlecode.com]

Status: New
Owner: ----
Labels: Type-Defect Priority-Medium

New issue 29 by tolis...@gmail.com: nidaqmx.h not available when using the
core runtime
http://code.google.com/p/pylibnidaqmx/issues/detail?id=29

What steps will reproduce the problem?
1. install the nidaqmx core runtime
(http://joule.ni.com/nidu/cds/view/p/id/2891/lang/en) rather than the full
installation (166MB rather than the 1.5GB or so for the full driver -- a
huge time saver!).
2. the nidaqmx.h file is not installed
3. attempts to use pylibnidaqmx fail

What is the expected output? What do you see instead?

import nidaqmx
File "C:\Python27\lib\site-packages\nidaqmx\__init__.py", line 83, in
<module>

from .libnidaqmx import AnalogInputTask, AnalogOutputTask,\
File "C:\Python27\lib\site-packages\nidaqmx\libnidaqmx.py", line 122, in
<modu
le>
assert os.path.isfile (include_nidaqmx_h), `include_nidaqmx_h`
AssertionError: u'C:\\Program Files (x86)\\National
Instruments\\NI-DAQ\\DAQmx A
NSI C Dev\\include\\NIDAQmx.h'


What version of the product are you using? On what operating system?

windows 7 64-bit
the latest pylibnidaqmx from source
nidaqmx 9.5.1 core runtime

Please provide any additional information below.

it would appear that pylibnidaqmx parses this header file to obtain what
is/isn't supported by the driver version - would it be possible for you to
include a default header file to use if one is not found in the NI
folders? (if their licensing permits). How likely is it to become
out-of-date?

[pylibn...@googlecode.com]

Comment #1 on issue 29 by pearu.peterson: nidaqmx.h not available when

using the core runtime
http://code.google.com/p/pylibnidaqmx/issues/detail?id=29

pylibnidaqmx does not need to use header file if nidaqmx version is one of
the following: 8.0, 8.9, 9.0, 9.1, 9.3.
To add new versions to this list, somebody has to import nidaqmx once
that will generate a file nidaqmx_h_X_Y.py using an existing header file.
This file can be added to pylibnidaqmx tree so that others with the same
version can reuse it without the need to install the actual header files.

I don't think that having a default header file will solve this issue as the
content of nidaqmx header files can vary quite a bit in between different
versions.