Mercurial > hg > camir-aes2014

annotate core/tools/xml2struct.m @ 0:e9a9cd732c1e tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
first hg version after svn
author wolffd
date Tue, 10 Feb 2015 15:05:51 +0000
parents
children
rev   line source
wolffd@0 1 function [ s ] = xml2struct( file )
wolffd@0 2 %Convert xml file into a MATLAB structure
wolffd@0 3 % [ s ] = xml2struct( file )
wolffd@0 4 %
wolffd@0 5 % A file containing:
wolffd@0 6 % <XMLname attrib1="Some value">
wolffd@0 7 % <Element>Some text</Element>
wolffd@0 8 % <DifferentElement attrib2="2">Some more text</Element>
wolffd@0 9 % <DifferentElement attrib3="2" attrib4="1">Even more text</DifferentElement>
wolffd@0 10 % </XMLname>
wolffd@0 11 %
wolffd@0 12 % Will produce:
wolffd@0 13 % s.XMLname.Attributes.attrib1 = "Some value";
wolffd@0 14 % s.XMLname.Element.Text = "Some text";
wolffd@0 15 % s.XMLname.DifferentElement{1}.Attributes.attrib2 = "2";
wolffd@0 16 % s.XMLname.DifferentElement{1}.Text = "Some more text";
wolffd@0 17 % s.XMLname.DifferentElement{2}.Attributes.attrib3 = "2";
wolffd@0 18 % s.XMLname.DifferentElement{2}.Attributes.attrib4 = "1";
wolffd@0 19 % s.XMLname.DifferentElement{2}.Text = "Even more text";
wolffd@0 20 %
wolffd@0 21 % Please note that the following characters are substituted
wolffd@0 22 % '-' by '_dash_', ':' by '_colon_' and '.' by '_dot_'
wolffd@0 23 %
wolffd@0 24 % Written by W. Falkena, ASTI, TUDelft, 21-08-2010
wolffd@0 25 % Attribute parsing speed increased by 40% by A. Wanner, 14-6-2011
wolffd@0 26 % Added CDATA support by I. Smirnov, 20-3-2012
wolffd@0 27 %
wolffd@0 28 % Modified by X. Mo, University of Wisconsin, 12-5-2012
wolffd@0 29
wolffd@0 30 if (nargin < 1)
wolffd@0 31 clc;
wolffd@0 32 help xml2struct
wolffd@0 33 return
wolffd@0 34 end
wolffd@0 35
wolffd@0 36 if isa(file, 'org.apache.xerces.dom.DeferredDocumentImpl') || isa(file, 'org.apache.xerces.dom.DeferredElementImpl')
wolffd@0 37 % input is a java xml object
wolffd@0 38 xDoc = file;
wolffd@0 39 else
wolffd@0 40 %check for existance
wolffd@0 41 if (exist(file,'file') == 0)
wolffd@0 42 %Perhaps the xml extension was omitted from the file name. Add the
wolffd@0 43 %extension and try again.
wolffd@0 44 if (isempty(strfind(file,'.xml')))
wolffd@0 45 file = [file '.xml'];
wolffd@0 46 end
wolffd@0 47
wolffd@0 48 if (exist(file,'file') == 0)
wolffd@0 49 error(['The file ' file ' could not be found']);
wolffd@0 50 end
wolffd@0 51 end
wolffd@0 52 %read the xml file
wolffd@0 53 xDoc = xmlread(file);
wolffd@0 54 end
wolffd@0 55
wolffd@0 56 %parse xDoc into a MATLAB structure
wolffd@0 57 s = parseChildNodes(xDoc);
wolffd@0 58
wolffd@0 59 end
wolffd@0 60
wolffd@0 61 % ----- Subfunction parseChildNodes -----
wolffd@0 62 function [children,ptext,textflag] = parseChildNodes(theNode)
wolffd@0 63 % Recurse over node children.
wolffd@0 64 children = struct;
wolffd@0 65 ptext = struct; textflag = 'Text';
wolffd@0 66 if hasChildNodes(theNode)
wolffd@0 67 childNodes = getChildNodes(theNode);
wolffd@0 68 numChildNodes = getLength(childNodes);
wolffd@0 69
wolffd@0 70 for count = 1:numChildNodes
wolffd@0 71 theChild = item(childNodes,count-1);
wolffd@0 72 [text,name,attr,childs,textflag] = getNodeData(theChild);
wolffd@0 73
wolffd@0 74 if (~strcmp(name,'#text') && ~strcmp(name,'#comment') && ~strcmp(name,'#cdata_dash_section'))
wolffd@0 75 %XML allows the same elements to be defined multiple times,
wolffd@0 76 %put each in a different cell
wolffd@0 77 if (isfield(children,name))
wolffd@0 78 if (~iscell(children.(name)))
wolffd@0 79 %put existsing element into cell format
wolffd@0 80 children.(name) = {children.(name)};
wolffd@0 81 end
wolffd@0 82 index = length(children.(name))+1;
wolffd@0 83 %add new element
wolffd@0 84 children.(name){index} = childs;
wolffd@0 85 textFieldNames = fieldnames(text);
wolffd@0 86 for t = 1:length(textFieldNames)
wolffd@0 87 textFieldName = textFieldNames{t};
wolffd@0 88 children.(name){index}.(textFieldName) = text.(textFieldName);
wolffd@0 89 end
wolffd@0 90 if(~isempty(attr))
wolffd@0 91 children.(name){index}.('Attributes') = attr;
wolffd@0 92 end
wolffd@0 93 else
wolffd@0 94 %add previously unknown (new) element to the structure
wolffd@0 95 children.(name) = childs;
wolffd@0 96 if(~isempty(text) && ~isempty(fieldnames(text)))
wolffd@0 97 textFieldNames = fieldnames(text);
wolffd@0 98 numTextFieldNames = length( textFieldNames );
wolffd@0 99 for i = 1:numTextFieldNames
wolffd@0 100 thisFieldName = textFieldNames{i};
wolffd@0 101 children.(name).(thisFieldName) = text.(thisFieldName);
wolffd@0 102 end
wolffd@0 103 end
wolffd@0 104 if(~isempty(attr))
wolffd@0 105 children.(name).('Attributes') = attr;
wolffd@0 106 end
wolffd@0 107 end
wolffd@0 108 else
wolffd@0 109 ptextflag = 'Text';
wolffd@0 110 if (strcmp(name, '#cdata_dash_section'))
wolffd@0 111 ptextflag = 'CDATA';
wolffd@0 112 elseif (strcmp(name, '#comment'))
wolffd@0 113 ptextflag = 'Comment';
wolffd@0 114 end
wolffd@0 115
wolffd@0 116 %this is the text in an element (i.e., the parentNode)
wolffd@0 117 if (~isempty(regexprep(text.(textflag),'[\s]*','')))
wolffd@0 118 if (~isfield(ptext,ptextflag) || isempty(ptext.(ptextflag)))
wolffd@0 119 ptext.(ptextflag) = text.(textflag);
wolffd@0 120 else
wolffd@0 121 %what to do when element data is as follows:
wolffd@0 122 %<element>Text <!--Comment--> More text</element>
wolffd@0 123
wolffd@0 124 %put the text in different cells:
wolffd@0 125 % if (~iscell(ptext)) ptext = {ptext}; end
wolffd@0 126 % ptext{length(ptext)+1} = text;
wolffd@0 127
wolffd@0 128 %just append the text
wolffd@0 129 ptext.(ptextflag) = [ptext.(ptextflag) text.(textflag)];
wolffd@0 130 end
wolffd@0 131 end
wolffd@0 132 end
wolffd@0 133
wolffd@0 134 end
wolffd@0 135 end
wolffd@0 136 end
wolffd@0 137
wolffd@0 138 % ----- Subfunction getNodeData -----
wolffd@0 139 function [text,name,attr,childs,textflag] = getNodeData(theNode)
wolffd@0 140 % Create structure of node info.
wolffd@0 141
wolffd@0 142 %make sure name is allowed as structure name
wolffd@0 143 name = toCharArray(getNodeName(theNode))';
wolffd@0 144 name = strrep(name, '-', '_dash_');
wolffd@0 145 name = strrep(name, ':', '_colon_');
wolffd@0 146 name = strrep(name, '.', '_dot_');
wolffd@0 147
wolffd@0 148 attr = parseAttributes(theNode);
wolffd@0 149 if (isempty(fieldnames(attr)))
wolffd@0 150 attr = [];
wolffd@0 151 end
wolffd@0 152
wolffd@0 153 %parse child nodes
wolffd@0 154 [childs,text,textflag] = parseChildNodes(theNode);
wolffd@0 155
wolffd@0 156 if (isempty(fieldnames(childs)) && isempty(fieldnames(text)))
wolffd@0 157 %get the data of any childless nodes
wolffd@0 158 % faster than if any(strcmp(methods(theNode), 'getData'))
wolffd@0 159 % no need to try-catch (?)
wolffd@0 160 % faster than text = char(getData(theNode));
wolffd@0 161 text.(textflag) = toCharArray(getTextContent(theNode))';
wolffd@0 162 end
wolffd@0 163
wolffd@0 164 end
wolffd@0 165
wolffd@0 166 % ----- Subfunction parseAttributes -----
wolffd@0 167 function attributes = parseAttributes(theNode)
wolffd@0 168 % Create attributes structure.
wolffd@0 169
wolffd@0 170 attributes = struct;
wolffd@0 171 if hasAttributes(theNode)
wolffd@0 172 theAttributes = getAttributes(theNode);
wolffd@0 173 numAttributes = getLength(theAttributes);
wolffd@0 174
wolffd@0 175 for count = 1:numAttributes
wolffd@0 176 %attrib = item(theAttributes,count-1);
wolffd@0 177 %attr_name = regexprep(char(getName(attrib)),'[-:.]','_');
wolffd@0 178 %attributes.(attr_name) = char(getValue(attrib));
wolffd@0 179
wolffd@0 180 %Suggestion of Adrian Wanner
wolffd@0 181 str = toCharArray(toString(item(theAttributes,count-1)))';
wolffd@0 182 k = strfind(str,'=');
wolffd@0 183 attr_name = str(1:(k(1)-1));
wolffd@0 184 attr_name = strrep(attr_name, '-', '_dash_');
wolffd@0 185 attr_name = strrep(attr_name, ':', '_colon_');
wolffd@0 186 attr_name = strrep(attr_name, '.', '_dot_');
wolffd@0 187 attributes.(attr_name) = str((k(1)+2):(end-1));
wolffd@0 188 end
wolffd@0 189 end
wolffd@0 190 end