regex - Extracting data between curly brackets in text file in MATLAB -

i have text file has multiple lines , hierarchies this:

...   ... ...  hierarchy {    {      b {       name            "b1"       type            virtual        }       c {     name            "c1"     parent          "b1"       }       d {     name            "d1"     type            virtual       } } {      b {       name            "b2"       type            virtual        }       c {     name            "c2"     parent          "b1"       }       d {     name            "d2"     type            virtual       } } }     ...     ... 

i looking @ extracting text "a {" till end of quotes matching "a" in matlab. content inside "a" vary in number of elements (there more name , contains fields). there way extract "a" , possibly fields cell array?

any appreciated!

i guess want code work number of object , name.

i'd recommend store objects in structs (but change code full cell array if wanted), , guess want code work n objects in pattern you've specified. regexp , cell arrays wouldn't best thing use here, recursion would. but, code below should work example , number of objects. comments should explain what's going on.

% open file fid = fopen('input.txt');  % line file tline = fgetl(fid);  % assign tokens starttoken = '{'; endtoken = '}';  % initialise counters bracketcount = 0; objectcount = 0;  % initialise root object store parsed file data = {};  % initialise children cell array store names of objects children = {};  % while not finished file while ischar(tline)     % check {     indexstart = regexp(tline, starttoken);     % check }     indexend = regexp(tline, endtoken);     % if { found     if ~isempty(indexstart)         % increment bracket counter         bracketcount = bracketcount + 1;         % extract item name         item = strtrim(tline(1:indexstart-1));         % if root          if bracketcount == 1             % increase object count             objectcount = objectcount + 1;             % initialise object in cell array             data{objectcount}.name = item;         % if not root object (e.g. a)         else             % initialise item field in object struct             data{objectcount}.(item) = [];         end         % keep track of item name         children{bracketcount} = item;         % optional display line         disp(['start: ' tline])     % if } found     elseif ~isempty(indexend)         % decrement bracket counter         bracketcount = bracketcount - 1;         % optional display line         disp(['end: ' tline])     % no brackets name value pair assumed     else         % split line 2 -> {name, value}         namevalue = strsplit(strtrim(tline));         % if line isnt empty , there 2 items          if ~isempty(namevalue) && numel(namevalue) == 2             % assign value (namevalue{2}) field (namevalue{1})             % of current child (children{bracketcount}) of current             % object (doc{objectcount})             data{objectcount}.(children{bracketcount}).(namevalue{1}) = namevalue{2};             % optional display line             disp(['name:' namevalue{1} ', value:' namevalue{2}])         end     end     % next line file     tline = fgetl(fid);  end  % close file fclose(fid);