python: added missing Python bindings.

[cristi-iacob]

Added the missing Python bindings. More details about these can be found in this discussion.

Signed-off-by: Cristi Iacob [email protected]

[rgetz]

https://app.codacy.com/gh/analogdevicesinc/libiio/pullRequest?prid=5393691

[rgetz]

tip : if you add the issue number #443 , it will automatically link things,

https://guides.github.com/features/issues/#notifications

[rgetz]

I ran the script in the issue on the branch, I get:

0 iio_create_xml_context_mem
0 iio_context_get_attr_value
0 iio_context_find_device
0 iio_device_find_channel
0 iio_device_find_attr
0 iio_device_find_buffer_attr
0 iio_device_set_data
0 iio_device_get_data
0 iio_channel_find_attr
0 iio_channel_set_data
0 iio_buffer_first
0 iio_buffer_foreach_sample
0 iio_buffer_set_data
0 iio_channel_convert_inverse
0 iio_device_find_debug_attr
0 iio_device_identify_filename

which is much less - thanks.

If we add these names in a comment in any file in that directory, (even a the doc directory), it will not show up...

slight tweak to look in all files, and doc directory.

for funct in $(grep ^__api iio.h | tr " " "\n" | grep "(" | sed 's/(.*$//' | grep -v -e _bool$ -e _double$ -e _longlong$ -e _all$ -e _raw$);    do echo "$(grep -c -R $funct ./bindings/python/* | awk -F: '{sum += $2} END {print sum}') $funct" ;   done | grep "^0"

[cristi-iacob]

I have corrected the lines with codacy-issues, but it seems that the issue I told you about here still persists.
Even stranger is the fact that I have made a draft PR to test Codacy #446 by copying the iio.py file on another branch and letting Codacy analyze it and it finds the same issues.

Should I add a (hidden) file or comment containing the name of the missing bindings? If yes, where should I add them?

[cristi-iacob]

I have managed to avoid this behavior of Codacy. I have added a new commit containing the modifications requested by Codacy. The strange thing is that if I squash these 2 commits, Codacy reports the issues again. Exactly the same happens on the codacy-test branch (#446), which is the one I have made by copy-paste-ing the iio.py file.

[rgetz]

I would add the on purpose un-implemented functions into the doc (rst) in a separate section. Or make a list of C to python equivalents, and just put the python equivalents of these functions.

@tfcollins : any thoughts?

[tfcollins]

If you have a list of non-implemented functions sure. A mapping between the C and python I think is a bit overkill since they are used in different ways. With the python API you can do autodiscovery of the functions by just using it and not looking at the doc

[tfcollins]

Code looks good, formatting consistent, and everything is documented