xml-stream - npm Package Compare versions

Comparing version 0.2.0 to 0.3.0

lib/finite-automata.js

@@ -10,3 +10,4 @@ module.exports = FiniteAutomata;
     leave: {},
-    state: {}
+    state: {},
+    flag: {}
   };
@@ -23,3 +24,3 @@ this._stack = [];
 
-function run(args, type) {
+function run(type, args) {
   var cbs = this._callbacks[type];
@@ -39,3 +40,3 @@ for (var cb in this._state) if (this._state.hasOwnProperty(cb)) {
   return this._deterministic;
-}
+};
 
@@ -52,10 +53,10 @@ FiniteAutomata.prototype.on = function(type, state, cb) {
   return this;
-}
+};
 
 FiniteAutomata.prototype.setState = function(state, args) {
   this._state = state;
-  run.call(this, args, 'enter');
-  run.call(this, args, 'state');
+  run.call(this, 'enter', args);
+  run.call(this, 'state', args);
   return this;
-}
+};
 
@@ -76,3 +77,3 @@ FiniteAutomata.prototype.nextState = function(symbol) {
   return newState;
-}
+};
 
@@ -87,3 +88,3 @@ FiniteAutomata.prototype.go = function(symbol, args) {
   this._stack[this._stackPtr] = undefined;
-  run.call(this, args, 'leave');
+  run.call(this, 'leave', args);
   this._state = this._stack[--this._stackPtr];
@@ -100,9 +101,10 @@ return this;
   this._state = next;
-  run.call(this, args, 'enter');
+  run.call(this, 'flag');
+  run.call(this, 'enter', args);
   return this;
 };
 
-FiniteAutomata.prototype.run = function(args) {
-  run.call(this, args, 'state');
-}
+FiniteAutomata.prototype.run = function(state, args) {
+  run.call(this, state, args);
+};
 
@@ -109,0 +111,0 @@ FiniteAutomata.prototype.transition = function(stateFrom, symbol, stateTo) {

lib/xml-stream.js

@@ -6,3 +6,3 @@ var events         = require('events')
 
-// Test if object is empty (has no own properties).
+// Tests if object is empty (has no own properties).
 function isEmpty(obj) {
@@ -52,2 +52,4 @@ for (var key in obj) if (obj.hasOwnProperty(key)) {
   this._emitData = false;
+  this._bufferLevel = 0;
+  this._collect = false;
   parse.call(this);
@@ -68,10 +70,13 @@ }
 // Supported events:
+//
 // * `data` on outgoing data chunk,
 // * `end` when parsing has ended,
 // * `startElement[: selector]` on opening tag for selector match,
+// * `updateElement[: selector]` on finished node for selector match
+//   with its contents buffered,
 // * `endElement[: selector]` on closing tag for selector match,
 // * `text[: selector]` on tag text for selector match.
 //
-// When adding listeners for `startElement` and `text`, the callback
-// can modify the provided node, before it is sent to the consumer.
+// When adding listeners for `startElement`, `updateElement`, and `text` the
+// callback can modify the provided node, before it is sent to the consumer.
 //
@@ -82,43 +87,151 @@ // Selector syntax is CSS-like and currently supports:
 // * `parent > child`
-XmlStream.prototype.on = function(event, listener) {
-  XmlStream.super_.prototype.on.call(this, event, listener);
-  if (event === 'data') {
+XmlStream.prototype.on = function(eventName, listener) {
+  XmlStream.super_.prototype.on.call(this, eventName, listener);
+  if (eventName === 'data') {
     this._emitData = true;
   }
-  var eventParts = event.match(/^(startElement|endElement|text):?(.*)/);
-  if (eventParts !== null) {
+  var event = parseEvent(eventName);
+  if (event !== null) {
     // If we're dealing with a selector event,
     // continue with selector-specific processing logic.
-    var evtType = eventParts[1];
-    var selector = eventParts[2];
-    var parts = selector.match(/[a-z0-9\:]+|>/ig);
-    selector = parts.join(' ');
-    event = evtType + ': ' + selector;
-    var finalState;
-    if (this._finalStates.hasOwnProperty(selector)) {
-      finalState = this._finalStates[selector];
+    var finalState = getFinalState.call(this, event.selector);
+    var self = this;
+    if (event.type === 'updateElement') {
+      this._fa.on('enter', finalState, function() {
+        self._bufferLevel++;
+      });
+      this._fa.on('leave', finalState, function(element, context, trace) {
+        self.emit(event.name, element, context, trace);
+        if (!--self._bufferLevel && self._emitData) {
+           emitElement.call(self, element, element.$name, true);
+        }
+      });
     } else {
-      var n = parts.length;
-      var immediate = false;
-      this._startState[this._lastState] = true;
-      for (var i = 0; i < n; i++) {
-        if (parts[i] === '>') {
-          immediate = true;
+      var fn = function(element, context, trace) {
+        self.emit(event.name, element, context, trace);
+      };
+      this._fa.on(faModes[event.type], finalState, fn);
+    }
+  }
+};
+
+//
+XmlStream.prototype.collect = function(selector) {
+  selector = normalizeSelector(selector);
+  var finalState = getFinalState.call(this, selector);
+  var self = this;
+  this._fa.on('flag', finalState, function() {
+    self._collect = true;
+  });
+};
+
+// Normalizes the selector and returns the new version and its parts.
+function normalizeSelector(selector) {
+  var parts = selector.match(/[a-z0-9\:]+|>/ig);
+  selector = parts.join(' ');
+  return {
+    normalized: parts.join(' '),
+    parts: parts
+  };
+}
+
+// Parses the selector event string and returns event information.
+function parseEvent(event) {
+  var eventParts = event.match(/^((?:start|end|update)Element|text):?(.*)/);
+  if (eventParts === null) {
+    return null;
+  }
+  var eventType = eventParts[1];
+  var selector = normalizeSelector(eventParts[2]);
+  return {
+    selector: selector,
+    type: eventType,
+    name: eventType + ': ' + selector.normalized
+  };
+}
+
+// Compiles a given selector object to a finite automata
+// and returns its last state.
+function getFinalState(selector) {
+  if (this._finalStates.hasOwnProperty(selector.normalized)) {
+    finalState = this._finalStates[selector.normalized];
+  } else {
+    var n = selector.parts.length;
+    var immediate = false;
+    this._startState[this._lastState] = true;
+    for (var i = 0; i < n; i++) {
+      var part = selector.parts[i];
+      if (part === '>') {
+        immediate = true;
+      } else {
+        if (!immediate) {
+          this._fa.transition(this._lastState, '', this._lastState);
+        }
+        this._fa.transition(this._lastState, part, ++this._lastState);
+        immediate = false;
+      }
+    }
+    finalState = this._lastState++;
+    this._finalStates[selector.normalized] = finalState;
+  }
+  return finalState;
+}
+
+// Emits XML for element opening tag.
+function emitStart(name, attrs) {
+  this.emit('data', '<' + name);
+  for (var attr in attrs) if (attrs.hasOwnProperty(attr)) {
+    this.emit('data', ' ' + attr + '="' + escape(attrs[attr]) + '"');
+  }
+  this.emit('data', '>');
+}
+
+// Emits XML for element closing tag.
+function emitEnd(name) {
+  this.emit('data', '</' + name + '>');
+}
+
+// Emits XML for element text.
+function emitText(text) {
+  this.emit('data', escape(text));
+}
+
+// Emits a single element and its descendants, or an array of elements.
+function emitElement(element, name, onLeave) {
+  if (Array.isArray(element)) {
+    var i;
+    for (i = 0; i < element.length - 1; i++) {
+      emitOneElement.call(this, element[i], name);
+    }
+    emitOneElement.call(this, element[i], name, onLeave);
+  } else {
+    emitOneElement.call(this, element, name, onLeave);
+  }
+}
+
+// Recursively emits a given element and its descendants.
+function emitOneElement(element, name, onLeave) {
+  if (typeof element === 'object') {
+    emitStart.call(this, name, element.$);
+    var hasText = false;
+    for (var child in element) {
+      if (element.hasOwnProperty(child) && child !== '$' && child != '$name') {
+        if (child === '$text') {
+          hasText = true;
         } else {
-          if (!immediate) {
-            this._fa.transition(this._lastState, '', this._lastState);
-          }
-          this._fa.transition(this._lastState, parts[i], ++this._lastState);
-          immediate = false;
+          emitElement.call(this, element[child], child);
         }
       }
-      finalState = this._lastState++;
-      this._finalStates[selector] = finalState;
     }
-    var self = this;
-    this._fa.on(faModes[evtType], finalState, function(item, context, trace) {
-      self.emit(event, item, context, trace);
-    });
+    if (hasText) {
+      emitText.call(this, element.$text);
+    }
+  } else {
+    emitStart.call(this, name, element.$);
+    emitText.call(this, element);
   }
-};
+  if(!onLeave) {
+    emitEnd.call(this, name);
+  }
+}
 
@@ -133,8 +246,6 @@ // Starts parsing the source stream and emitting various events.
   var curr = {
-    node: {
-      element: {},
-      text: '',
-      name: 'root'
-    },
+    element: {},
+    collect: this._collect,
     fullText: '',
+    space: 0,
     path: '',
@@ -151,28 +262,36 @@ context: {}
     stack.push(curr);
-    trace[curr.path] = curr.node.element;
+    trace[curr.path] = curr.element;
     var context = Object.create(curr.context);
-    var element = {$: attr};
-    curr.node.element[name] = element;
+    var element = {
+      $: attr,
+      $name: name,
+      $text: ''
+    };
+    var parent = curr.element;
     curr = {
-      node: {
-        element: element,
-        text: '',
-        name: name
-      },
+      element: element,
+      collect: false,
       fullText: '',
-      path: curr.path + '/' + name
+      space: 0,
+      path: curr.path + '/' + name,
+      context: context
     };
-    context[name] = curr.node.element;
-    curr.context = context;
-    fa.enter(name, [curr.node, curr.context, trace]);
-    if (self._emitData) {
-      self.emit('data', '<' + curr.node.name);
-      var attrs = curr.node.element.$;
-      for (var attr in attrs) if (attrs.hasOwnProperty(attr)) {
-        self.emit('data', ' ' + attr + '="' + escape(attrs[attr]) + '"');
+    fa.enter(name, [element, context, trace]);
+    name = element.$name;
+    curr.collect = self._collect;
+    if (curr.collect) {
+      var container;
+      if (Object.hasOwnProperty.call(parent, name)) {
+        container = parent[name];
+        container.push(element);
+      } else {
+        container = [element];
+        parent[name] = container;
       }
-      self.emit('data', '>');
+    } else {
+      parent[name] = element;
+      context[name] = element;
     }
-    if (isEmpty(curr.node.element.$)) {
-      delete curr.node.element.$;
+    if (self._bufferLevel === 0 && self._emitData) {
+      emitStart.call(self, element.$name, element.$);
     }
@@ -187,22 +306,37 @@ });
     var prev = stack.pop();
-    var node = curr.node;
-    node.text = curr.fullText;
-    var val;
-    if (isEmpty(node.element)) {
-      val = node.text;
-    } else if (node.text !== '') {
-      val = new String(node.text);
-      for(var prop in node.element) if (node.element.hasOwnProperty(prop)) {
-        val[prop] = node.element[prop];
+    var element = curr.element;
+    var text = curr.fullText;
+    var attr = element.$;
+    if (typeof attr !== 'object') {
+      attr = {};
+    }
+    var name = element.$name;
+    delete element.$;
+    delete element.$text;
+    delete element.$name;
+    var val = element;
+    if (isEmpty(element) && isEmpty(attr)) {
+      val = text;
+    } else {
+      if (text !== '') {
+        element.$text = text;
       }
-    } else {
-      val = node.element;
+      if (attr !== null) {
+        element.$ = attr;
+      }
     }
+    element.$name = name;
     curr.context[name] = val;
-    prev.node.element[node.name] = val;
-    fa.leave([node, curr.context, trace]);
-    if (self._emitData) {
-      self.emit('data', '</' + node.name + '>');
+    if (this._collect) {
+      var container = prev.element[name];
+      container[container.length - 1] = val;
+    } else {
+      prev.element[name] = val;
     }
+    fa.leave([element, curr.context, trace]);
+    if (self._bufferLevel === 0 && self._emitData) {
+      emitEnd.call(self, name);
+    }
     curr = prev;
+    this._collect = curr.collect;
   });
@@ -213,8 +347,23 @@
   xml.on('text', function(text) {
-    curr.node.text = text;
-    fa.run([curr.node, curr.context, trace]);
-    if (self._emitData) {
-      self.emit('data', escape(curr.node.text));
+    curr.element.$text = text;
+    fa.run('state', [curr.element, curr.context, trace]);
+    if (self._bufferLevel === 0 && self._emitData) {
+      emitText.call(self, text);
     }
-    curr.fullText += curr.node.text.trim();
+    var trimmed = curr.element.$text.trim();
+    if (curr.space == 0) {
+      if (trimmed != '') {
+          curr.space = 1;
+      }
+    } else if (curr.space == 1) {
+      if (trimmed == '') {
+        curr.space = 2;
+      }
+    } else if (curr.space == 2) {
+      if (trimmed != '') {
+        curr.space = 1;
+        curr.fullText += ' ';
+      }
+    }
+    curr.fullText += trimmed;
   });
@@ -232,2 +381,2 @@
   });
-};
+}

package.json

@@ -9,3 +9,3 @@ {
   ],
-  "version": "0.2.0",
+  "version": "0.3.0",
   "author": "AssistUnion <[e-mail]>",
@@ -23,5 +23,2 @@ "maintainers": [
   },
-  "devDependencies": {
-    "vows": "~0.5.8"
-  },
   "engines": {
@@ -34,2 +31,2 @@ "node": ">=0.4.7"
   "main": "index"
-}
+}

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

No README

Quality

Package does not have a README. This may indicate a failed publish or a low quality package.

Found 1 instance in 1 package

Filesystem access

Supply chain risk

Accesses the file system, and could potentially read sensitive data.

Found 1 instance in 1 package

Improved metrics

Dev dependency count

0

decreased by-100%

Lines of code

500

increased by18.76%

Number of medium quality alerts

0

decreased by-100%

Number of lines in readme file

39

increased byInfinity%

Number of low supply chain risk alerts

0

decreased by-100%

Worsened metrics

Total package byte prevSize

17910

decreased by-32.66%

Number of package files

8

decreased by-27.27%

No dependency changes