I thought of a way to use ODS-"data" in octave (ODS being the open document spreadsheet).
loadods("data.ods"); # for loading vars in all tables loadods("data.ods#table1"); # loading data from just "table1"
To
address data in cells you have to put "headings" just above the very
first field where data begins. Therefore data can theoretically be spread
anywhere around in the document. Sample "headings" are (separated by ,): \float f1, \bool b1 [2,3], \int32 i1 [3], \list l1 [3,2], \struct s1 [2], \string str2 [2,3], \range r1
f1 would be a single "scalar", b1 a "bool matrix" (since it includes dimensions), i1 would be converted to an int32 matrix with dimensions 3x1, l1 would end up as a cell array that tries to find more \type -definitions below to fill, s1 would just be as a l1 including "names", str2 would be a char matrix with 2x3xN (whereby N is the length of the largest string), r1 is a range with never having dimensions
When writing octave-data back to a spreadsheet it'll overwrite previous contents and then list "headings" and data sequentially line by line. For
now it is implemented as oct-function. It includes a simple xml-parser
(self written, just to full fill the basic needs) and zip-archive
(http://www.artpol-software.com/ZipArchive/) for extracting data from those ODS-files.
Maybe there are some thoughts to this.
best regards, stefan Live Earth is coming. Learn more about the hottest summer event - only on MSN. Check it out!