Here are a few of the sample data files that we have worked on and their xml translations.  I believe that these are examples of files where it would be difficult to achieve the desired translation without some form of layering.

premix.bin - a binary file from the chemkin program.  Contains three different sections, determined by a header value.  In the solution section there is a sequence of x values, the values were combined with a sequence of points containing a temp value and massFrac values.

test.prp - a text-based file based on a nwChem format.  Contains some header info, then a list of field names, then a series of frames with values that should correspond with the field names.  When translating, the list of field names is read as a layer and later referenced from the frames.

clayers.bin - same as example of layers shown at last f2f, creates a sequence of 100 complex numbers from a file containing a sequence of 100 real floats then 100 imaginary floats. 

Tara

The raw data files and their xml translations
<<premix.bin>> <<chemkinbin.xml>>
<<test.prp>> <<prp.xml>>
<<clayers.bin>> <<clayer.xml>>