[Pytables-users] smart search of data file

[Andre' Walker-L. <wal...@gm...>]

Hi All,

I have a question about reading through a file in a smart way.

I am trying to write a generic executable to read a data file, and with user input, grab a certain array from an hdf5 file.

./my_data_analyzer -i my_data.h5 -other_args


The structure of the files is 

f = tables.openFile('my_data.h5')

f.root.childNode.childNode....dataArray

where I do not know ahead of time how deep the childNodes go.  My general strategy was to query the user (me and a colleague) which particular array we would like to analyze.  So I was thinking to use raw_input to ask which node the user wants - but providing a list of possibilities

for node in f.walkNodes():
	print node

if the user provides the full path to the data array - then read as numpy array - but if the user provides only partial path (maybe just the top level group) then query further until the user finds the full path, and finally read as numpy array.

Also - given the data files, some of the groups (childNodes) likely won't obey the NaturalName conventions - in case that matters.

1) is my goal clear?
2) does anyone have a clever way to do this?  some sort of recursive query with raw_input to get the desired data.


Thanks,

Andre