]> git.pond.sub.org Git - eow/blobdiff - static/dojo-release-1.1.1/dojox/rpc/JsonReferencing.js
Replace Dojo with jQuery
[eow] / static / dojo-release-1.1.1 / dojox / rpc / JsonReferencing.js
diff --git a/static/dojo-release-1.1.1/dojox/rpc/JsonReferencing.js b/static/dojo-release-1.1.1/dojox/rpc/JsonReferencing.js
deleted file mode 100644 (file)
index 618ea06..0000000
+++ /dev/null
@@ -1,265 +0,0 @@
-if(!dojo._hasResource["dojox.rpc.JsonReferencing"]){ //_hasResource checks added by build. Do not use _hasResource directly in your code.
-dojo._hasResource["dojox.rpc.JsonReferencing"] = true;
-dojo.provide("dojox.rpc.JsonReferencing");
-dojo.require("dojo.date.stamp");
-dojo.require("dojo._base.Deferred");
-
-// summary:
-// Adds advanced JSON {de}serialization capabilities to the base json library.
-// This enhances the capabilities of dojo.toJson and dojo.fromJson,
-// adding referencing support, date handling, and other extra format handling.
-// On parsing, references are resolved. When references are made to 
-// ids/objects that have been loaded yet, a Deferred object will be used as the
-// value and as soon as a callback is added to the Deferred object, the target
-// object will be loaded.
-
-
-dojox.rpc._index={}; // the global map of id->object
-dojox.rpc.onUpdate = function(/*Object*/ object,  /* attribute-name-string */ attribute,  /* any */ oldValue,  /* any */ newValue){
-               //      summary:
-               //              This function is called when an existing object in the system is updated. Existing objects are found by id. 
-};
-
-dojox.rpc.resolveJson = function(/*Object*/ root,/*Object?*/ schema){
-       // summary:
-       //              Indexes and resolves references in the JSON object. 
-       // A JSON Schema object that can be used to advise the handling of the JSON (defining ids, date properties, urls, etc)
-       //  
-       // root: 
-       //              The root object of the object graph to be processed
-       //
-       // schema: A JSON Schema object that can be used to advise the parsing of the JSON (defining ids, date properties, urls, etc)   //
-       //              Currently this provides a means for context based id handling
-       //
-       // return:
-       //              An object, the result of the processing
-       var ref,reWalk=[];
-       function makeIdInfo(schema){ // find out what attribute and what id prefix to use
-               if (schema){
-                       var attr;
-                       if (!(attr = schema._idAttr)){
-                               for (var i in schema.properties){
-                                       if (schema.properties[i].unique){
-                                               schema._idAttr = attr = i;
-                                       }
-                               }
-                       }
-                       if (attr || schema._idPrefix){
-                               return {attr:attr || 'id',prefix:schema._idPrefix};
-                       }
-               }
-
-               return false;
-       }
-       function walk(it,stop,schema,idInfo,defaultId){
-               // this walks the new graph, resolving references and making other changes 
-               var val,i;
-               var id = it[idInfo.attr];
-               id = (id && (idInfo.prefix + id)) || defaultId; // if there is an id, prefix it, otherwise inherit 
-               var target = it;
-               
-               if (id){ // if there is an id available...
-                       it._id = id;
-                       if (dojox.rpc._index[id]){ // if the id already exists in the system, we should use the existing object, and just update it
-                               target = dojox.rpc._index[id];
-                               delete target.$ref; // remove this artifact
-                       }
-                       dojox.rpc._index[id] = target; // add the prefix, set _id, and index it
-                       if (schema && dojox.validate && dojox.validate.jsonSchema){ // if json schema is activated, we can load it in the registry of instance schemas map
-                               dojox.validate.jsonSchema._schemas[id] = schema;
-                       }
-                       
-               }               
-               for (i in it){
-                       if (it.hasOwnProperty(i) && (typeof (val=it[i]) =='object') && val){
-                               ref=val.$ref;
-                               if (ref){ // a reference was found
-                                       var stripped = ref.replace(/\\./g, '@').replace(/"[^"\\\n\r]*"/g, '');// trim it
-                                       if(/[\w\[\]\.\$ \/\r\n\t]/.test(stripped) && !/=|((^|\W)new\W)/.test(stripped)){ // make sure it is a safe reference
-                                               var path = ref.match(/(^\.*[^\.\[]+)([\.\[].*)?/); // divide along the path
-                                               if ((ref=path[1]=='$' ? root:dojox.rpc._index[new dojo._Url(idInfo.prefix,path[1])]) &&  // a $ indicates to start with the root, otherwise start with an id 
-                                                       (ref = path[2] ? eval('ref' + path[2]) : ref)){// starting point was found, use eval to resolve remaining property references
-                                                       // otherwise, no starting point was found (id not found), if stop is set, it does not exist, we have 
-                                                       // unloaded reference, if stop is not set, it may be in a part of the graph not walked yet, 
-                                                       // we will wait for the second loop
-                                                       val = ref;
-                                               }
-                                               else{
-                                                       if (!stop){
-                                                               if (!rewalking) 
-                                                                       reWalk.push(it); // we need to rewalk it to resolve references
-                                                               var rewalking = true; // we only want to add it once
-                                                       }
-                                                       else {
-                                                               ref = val.$ref;
-                                                               val = new dojo.Deferred();
-                                                               val._id = idInfo.prefix + ref;
-                                                               (function(val,ref){
-                                                                       var connectId = dojo.connect(val,"addCallbacks",function(){
-                                                                               dojo.disconnect(connectId);
-                                                                               dojox.rpc.services[idInfo.prefix.substring(0,idInfo.prefix.length-1)](ref) // otherwise call by looking up the service 
-                                                                                       .addCallback(dojo.hitch(val,val.callback));
-                                                                               
-                                                                       });
-                                                               })(val,ref);
-                                                       }
-                                               }
-                                       }
-                               }
-                               else {
-                                       if (!stop){ // if we are in stop, that means we are in the second loop, and we only need to check this current one,
-                                                                               // further walking may lead down circular loops
-                                               var valSchema = val.schema || // a schema can be self-defined by the object,  
-                                                                               (schema && schema.properties && schema.properties[i]);  // or it can from the schema sub-object definition 
-                                               if (valSchema){
-                                                       idInfo = makeIdInfo(valSchema)||idInfo;
-                                               }
-                                               val = walk(val,reWalk==it,valSchema,idInfo,id && (id + ('[' + dojo._escapeString(i) + ']')));
-                                       }
-                               }
-                       }
-                       if (dojo.isString(val) && schema && schema.properties && schema.properties[i] && schema.properties[i].format=='date-time'){// parse the date string
-                               val = dojo.date.stamp.fromISOString(val); // create a date object
-                       }
-                       it[i] = val;
-                       var old = target[i];  
-                       if (val !== old){ // only update if it changed
-                               target[i] = val; // update the target
-                               propertyChange(i,old,val);
-                       }
-               }
-               function propertyChange(key,old,newValue){
-                       setTimeout(function(){
-                               dojox.rpc.onUpdate(target,i,old,newValue); // call the listener for each update
-                       });
-               }
-               if (target != it){ // this means we are updating, we need to remove deleted
-                       for (i in target){
-                               if (!it.hasOwnProperty(i) && i != '_id' && !(target instanceof Array && isNaN(i))){
-                                       propertyChange(i,target[i],undefined);
-                                       delete target[i];
-                               }
-                       }
-               }
-               return target;
-       }
-       var idInfo = makeIdInfo(schema)||{attr:'id',prefix:''};
-       if (!root){ return root; } 
-       root = walk(root,false,schema,idInfo,dojox._newId && (new dojo._Url(idInfo.prefix,dojox._newId) +'')); // do the main walk through
-       walk(reWalk,false,schema,idInfo); // re walk any parts that were not able to resolve references on the first round
-       return root;
-};
-dojox.rpc.fromJson = function(/*String*/ str,/*Object?*/ schema){
-       // summary:
-       //              evaluates the passed string-form of a JSON object. 
-       // A JSON Schema object that can be used to advise the parsing of the JSON (defining ids, date properties, urls, etc)
-       // which may defined by setting dojox.currentSchema to the current schema you want to use for this evaluation
-       //  
-       // json: 
-       //              a string literal of a JSON item, for instance:
-       //                      '{ "foo": [ "bar", 1, { "baz": "thud" } ] }'
-       // schema: A JSON Schema object that can be used to advise the parsing of the JSON (defining ids, date properties, urls, etc)   //
-       //              Currently this provides a means for context based id handling
-       //
-       // return:
-       //              An object, the result of the evaluation
-       root = eval('(' + str + ')'); // do the eval
-       if (root){
-               return this.resolveJson(root,schema);
-       }
-       return root;
-}
-dojox.rpc.toJson = function(/*Object*/ it, /*Boolean?*/ prettyPrint, /*Object?*/ schema){
-       // summary:
-       //              Create a JSON serialization of an object. 
-       //              This has support for referencing, including circular references, duplicate references, and out-of-message references
-       //              id and path-based referencing is supported as well and is based on http://www.json.com/2007/10/19/json-referencing-proposal-and-library/.
-       //
-       // it:
-       //              an object to be serialized. 
-       //
-       // prettyPrint:
-       //              if true, we indent objects and arrays to make the output prettier.
-       //              The variable dojo.toJsonIndentStr is used as the indent string 
-       //              -- to use something other than the default (tab), 
-       //              change that variable before calling dojo.toJson().
-       //
-       // schema: A JSON Schema object that can be used to advise the parsing of the JSON (defining ids, date properties, urls, etc)   //
-       //              Currently this provides a means for context based id handling
-       //              
-       // return:
-       //              a String representing the serialized version of the passed object.
-       
-       var idPrefix = (schema&& schema._idPrefix) || ''; // the id prefix for this context 
-       var paths={};
-       function serialize(it,path,_indentStr){ 
-               if (it && dojo.isObject(it)){
-                       var value;
-                       if (it instanceof Date){ // properly serialize dates
-                               return '"' + dojo.date.stamp.toISOString(it,{zulu:true}) + '"';
-                       }
-                       var id = it._id;
-                       if (id){ // we found an identifiable object, we will just serialize a reference to it... unless it is the root
-                               
-                               if (path != '$'){
-                                       return serialize({$ref:id.charAt(0)=='$' ? id : // a pure path based reference, leave it alone
-                                                                       id.substring(0,idPrefix.length)==idPrefix ?  // see if the reference is in the current context
-                                                                               id.substring(idPrefix.length): // a reference with a prefix matching the current context, the prefix should be removed
-                                                                                       '../' + id});// a reference to a different context, assume relative url based referencing
-                               }
-                               path = id;
-                       }
-                       else {
-                               it._id = path; // we will create path ids for other objects in case they are circular 
-                               paths[path] = it;// save it here so they can be deleted at the end
-                       }
-                       _indentStr = _indentStr || "";
-                       var nextIndent = prettyPrint ? _indentStr + dojo.toJsonIndentStr : "";
-                       var newLine = prettyPrint ? "\n" : "";
-                       var sep = prettyPrint ? " " : "";
-                       
-                       if (it instanceof Array){
-                               var res = dojo.map(it, function(obj,i){
-                                       var val = serialize(obj, path + '[' + i + ']', nextIndent);
-                                       if(!dojo.isString(val)){
-                                               val = "undefined";
-                                       }
-                                       return newLine + nextIndent + val;
-                               });
-                               return "[" + res.join("," + sep) + newLine + _indentStr + "]";
-                       } 
-                       
-                       var output = [];
-                       for(var i in it){
-                               var keyStr;
-                               if(typeof i == "number"){
-                                       keyStr = '"' + i + '"';
-                               }else if(dojo.isString(i) && i != '_id'){
-                                       keyStr = dojo._escapeString(i);
-                               }else{
-                                       // skip non-string or number keys
-                                       continue;
-                               }
-                               var val = serialize(it[i],path+(i.match(/^[a-zA-Z]\w*$/) ? // can we use simple .property syntax? 
-                                                                                                       ('.' + i) : // yes, otherwise we have to escape it
-                                                                                                       ('[' + dojo._escapeString(i) + ']')),nextIndent);
-                               if(!dojo.isString(val)){
-                                       // skip non-serializable values
-                                       continue;
-                               }
-                               output.push(newLine + nextIndent + keyStr + ":" + sep + val);
-                       }
-                       return "{" + output.join("," + sep) + newLine + _indentStr + "}";
-               }
-               
-               return dojo.toJson(it); // use the default serializer for primitives
-       }
-       var json = serialize(it,'$','');
-       for (i in paths){  // cleanup the temporary path-generated ids
-               delete paths[i]._id;
-       }
-       return json;
-}
-
-}