ncbi taxonomy in matlab (incomplete -- see DANGER)

traverse.m

function traverse(node_id, iter)
    if nargin < 2
        iter = 0;
    end

% > head names.dmp
% 1	|	all	|		|	synonym	|
% 1	|	root	|		|	scientific name	|
% 2	|	Bacteria	|	Bacteria <prokaryote>	|	scientific name	|

% > head nodes.dmp
% 1   |   1   |   no rank |       |   8   |   0   |   1   |   0   |   0   |   0   |   0   |   0   |       |
% 2   |   131567  |   superkingdom    |       |   0   |   0   |   11  |   0   |   0   |   0   |   0   |   0   |       |

    persistent parentmap;
    persistent genusmap;
    persistent namemap;

    if isempty(genusmap);
        disp('loading files...');
        t0 = now;
        F = fopen('nodes.dmp');
        rawnodes = textscan(F, '%d | %d | %s %*[^\n]');
        fclose(F);

        parentmap = containers.Map(rawnodes{1}, rawnodes{2});
        genusmap = containers.Map(rawnodes{1}, rawnodes{3});

        % FIXME
        % DANGER: id -> name is not 1:1 !!!
        F = fopen('names.dmp');
        rawnames = textscan(F, '%d | %s %*[^\n]');
        fclose(F);

        namemap = containers.Map(rawnames{1}, rawnames{2});

        disp('files loaded in');
        disp(now-t0);
    else
        ;
        % disp('files already loaded');
    end



    if iter > 10;
        disp('big loop! break');
        return;
    end

    if ~genusmap.isKey(node_id);
        disp('no such id');
        return;
    end

    category = genusmap(node_id);
    if strcmp(category, 'genus')
        disp(sprintf('found genus at %d\n', node_id));
    else
        if ~parentmap.isKey(node_id);
            disp('warning: no parent');
        elseif node_id == parentmap(node_id);
            disp('end of tree!')
        else
            traverse(parentmap(node_id), iter+1);
        end
    end

end