jssoup - npm Package Compare versions

Comparing version 0.0.10 to 0.0.11

dist/lib/jssoup.js

@@ -11,2 +11,8 @@ 'use strict';
 
+var _builder = require('./builder.js');
+
+var _builder2 = _interopRequireDefault(_builder);
+
+function _interopRequireDefault(obj) { return obj && obj.__esModule ? obj : { default: obj }; }
+
 function _possibleConstructorReturn(self, call) { if (!self) { throw new ReferenceError("this hasn't been initialised - super() hasn't been called"); } return call && (typeof call === "object" || typeof call === "function") ? call : self; }
@@ -19,2 +25,4 @@
 var htmlparser = require('htmlparser');
+
+
 if (typeof module !== 'undefined' && typeof module.exports !== 'undefined') {
@@ -163,7 +171,7 @@ try {
 
-  function SoupTag(name) {
-    var attrs = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : null;
-    var parent = arguments.length > 2 && arguments[2] !== undefined ? arguments[2] : null;
-    var previousElement = arguments.length > 3 && arguments[3] !== undefined ? arguments[3] : null;
-    var nextElement = arguments.length > 4 && arguments[4] !== undefined ? arguments[4] : null;
+  function SoupTag(name, builder) {
+    var attrs = arguments.length > 2 && arguments[2] !== undefined ? arguments[2] : null;
+    var parent = arguments.length > 3 && arguments[3] !== undefined ? arguments[3] : null;
+    var previousElement = arguments.length > 4 && arguments[4] !== undefined ? arguments[4] : null;
+    var nextElement = arguments.length > 5 && arguments[5] !== undefined ? arguments[5] : null;
 
@@ -178,2 +186,3 @@ _classCallCheck(this, SoupTag);
     _this4.hidden = false;
+    _this4.builder = builder;
     return _this4;
@@ -198,3 +207,3 @@ }
       for (var i = 0; i < children.length; ++i) {
-        var ele = this._transfer(children[i]);
+        var ele = this._transform(children[i]);
         last.nextElement = ele;
@@ -218,4 +227,4 @@ ele.previousElement = last;
   }, {
-    key: '_transfer',
-    value: function _transfer(dom) {
+    key: '_transform',
+    value: function _transform(dom) {
       if (!dom) return null;
@@ -232,3 +241,3 @@ if (dom.type === 'text') {
       }
-      return new SoupTag(dom.name, dom.attribs, this);
+      return new SoupTag(dom.name, this.builder, dom.attribs, this);
     }
@@ -265,3 +274,5 @@ }, {
           var tag = strainer.match(descendants[i]);
-          if (tag) results.push(tag);
+          if (tag) {
+            results.push(tag);
+          }
         }
@@ -273,2 +284,70 @@ }
   }, {
+    key: 'findPreviousSibling',
+    value: function findPreviousSibling() {
+      var name = arguments.length > 0 && arguments[0] !== undefined ? arguments[0] : undefined;
+      var attrs = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : undefined;
+      var string = arguments.length > 2 && arguments[2] !== undefined ? arguments[2] : undefined;
+
+      var results = this.findPreviousSiblings(name, attrs, string);
+      if (results.length > 0) {
+        return results[0];
+      }
+      return undefined;
+    }
+  }, {
+    key: 'findPreviousSiblings',
+    value: function findPreviousSiblings() {
+      var name = arguments.length > 0 && arguments[0] !== undefined ? arguments[0] : undefined;
+      var attrs = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : undefined;
+      var string = arguments.length > 2 && arguments[2] !== undefined ? arguments[2] : undefined;
+
+      var results = [];
+      var cur = this.previousSibling;
+      var strainer = new SoupStrainer(name, attrs, string);
+      while (cur) {
+        if (cur instanceof SoupTag) {
+          var tag = strainer.match(cur);
+          if (tag) {
+            results.push(tag);
+          }
+        }
+        cur = cur.previousSibling;
+      }
+      return results;
+    }
+  }, {
+    key: 'findNextSibling',
+    value: function findNextSibling() {
+      var name = arguments.length > 0 && arguments[0] !== undefined ? arguments[0] : undefined;
+      var attrs = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : undefined;
+      var string = arguments.length > 2 && arguments[2] !== undefined ? arguments[2] : undefined;
+
+      var results = this.findNextSiblings(name, attrs, string);
+      if (results.length > 0) {
+        return results[0];
+      }
+      return undefined;
+    }
+  }, {
+    key: 'findNextSiblings',
+    value: function findNextSiblings() {
+      var name = arguments.length > 0 && arguments[0] !== undefined ? arguments[0] : undefined;
+      var attrs = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : undefined;
+      var string = arguments.length > 2 && arguments[2] !== undefined ? arguments[2] : undefined;
+
+      var results = [];
+      var cur = this.nextSibling;
+      var strainer = new SoupStrainer(name, attrs, string);
+      while (cur) {
+        if (cur instanceof SoupTag) {
+          var tag = strainer.match(cur);
+          if (tag) {
+            results.push(tag);
+          }
+        }
+        cur = cur.nextSibling;
+      }
+      return results;
+    }
+  }, {
     key: 'getText',
@@ -314,11 +393,29 @@ value: function getText() {
         if (attrs) {
-          text += indent.repeat(level) + '<' + this.name + ' ' + attrs + '>' + breakline;
+          text += indent.repeat(level) + '<' + this.name + ' ' + attrs;
         } else {
-          text += indent.repeat(level) + '<' + this.name + '>' + breakline;
+          text += indent.repeat(level) + '<' + this.name;
         }
       }
 
+      // is an element doesn't have any contents, it's a self closing element
+      if (!this.hidden) {
+        if (this._isEmptyElement() && this.builder.canBeEmptyElement(this.name)) {
+          text += ' />' + breakline;
+          return text;
+        } else {
+          text += '>' + breakline;
+        }
+      }
+
       for (var i = 0; i < this.contents.length; ++i) {
         if (this.contents[i] instanceof SoupString) {
-          text += indent.repeat(level + 1) + this.contents[i].toString() + breakline;
+          var curText = this.contents[i].toString();
+          curText = curText.trim();
+          if (curText.length != 0) {
+            if (curText.substring(curText.length - 1) == "\n") {
+              text += indent.repeat(level + 1) + curText;
+            } else {
+              text += indent.repeat(level + 1) + curText + breakline;
+            }
+          }
         } else {
@@ -332,5 +429,7 @@ if (this.contents[i] instanceof SoupComment) {
       }
+
       if (!this.hidden) {
         text += indent.repeat(level) + '</' + this.name + '>' + breakline;
       }
+
       return text;
@@ -383,2 +482,7 @@ }
   }, {
+    key: '_isEmptyElement',
+    value: function _isEmptyElement() {
+      return this.contents.length == 0;
+    }
+  }, {
     key: 'string',
@@ -429,5 +533,7 @@ get: function get() {
   function JSSoup(text) {
+    var ignoreWhitespace = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : true;
+
     _classCallCheck(this, JSSoup);
 
-    var _this5 = _possibleConstructorReturn(this, (JSSoup.__proto__ || Object.getPrototypeOf(JSSoup)).call(this, ROOT_TAG_NAME, null));
+    var _this5 = _possibleConstructorReturn(this, (JSSoup.__proto__ || Object.getPrototypeOf(JSSoup)).call(this, ROOT_TAG_NAME, new _builder2.default(), null));
 
@@ -438,3 +544,3 @@ var handler = new htmlparser.DefaultHandler(function (error, dom) {
       } else {}
-    }, { verbose: false, ignoreWhitespace: true });
+    }, { verbose: false, ignoreWhitespace: ignoreWhitespace });
 
@@ -441,0 +547,0 @@ var parser = new htmlparser.Parser(handler);

package.json

 {
   "name": "jssoup",
-  "version": "0.0.10",
+  "version": "0.0.11",
   "description": "JSSoup is a BeautifulSoup style HTML parser library.",
@@ -15,3 +15,3 @@ "main": "./dist/lib/jssoup.js",
     "prebuild": "npm run clean && npm run init",
-    "build": "babel ./lib/jssoup.js -d dist --presets es2015 "
+    "build": "babel ./lib/*.js -d dist --presets es2015 "
   },
@@ -45,4 +45,4 @@ "repository": {
     "htmlparser": "^1.7.7",
-    "lodash": "^4.17.5"
+    "lodash": "^4.17.15"
   }
 }

README.md

@@ -37,2 +37,8 @@ JSSoup
 ```
+> The text element only contains whitespace will be ignored by default. To disable this feature, set second parameter 
+of JSSoup to false. This parameter is "ignoreWhitespace" and will be passed into htmlparser.
+```javascript
+var soup = new JSSoup('<html><head>hello</head></html>', false);
+```
+
 ### Name
@@ -140,2 +146,64 @@ ```javascript
 ```
+#### .findNextSibling()
+```javascript
+var data = `
+<div>
+  <span> test </span>
+  <div> div </div>
+  <p> hello </p>
+  <p> world </p>
+</div>
+`
+var soup = new JSSoup(data);
+var span = soup.find('span');
+span.findNextSibling('p')
+// <p> hello </p>
+```
+#### .findNextSiblings()
+```javascript
+var data = `
+<div>
+  <span> test </span>
+  <div> div </div>
+  <p> hello </p>
+  <p> world </p>
+</div>
+`
+var soup = new JSSoup(data);
+var span = soup.find('span');
+span.findNextSiblings('p')
+// <p> hello </p>
+// <p> world </p>
+```
+#### .findPreviousSibling()
+```javascript
+var data = `
+<div>
+  <p> hello </p>
+  <p> world </p>
+  <div> div </div>
+  <span> test </span>
+</div>
+`
+var soup = new JSSoup(data);
+var span = soup.find('span');
+span.findPreviousSibling('p')
+// <p> world </p>
+```
+#### .findPreviousSiblings()
+```javascript
+var data = `
+<div>
+  <p> hello </p>
+  <p> world </p>
+  <div> div </div>
+  <span> test </span>
+</div>
+`
+var soup = new JSSoup(data);
+var span = soup.find('span');
+span.findPreviousSiblings('p')
+// <p> hello </p>
+// <p> world </p>
+```
 ### Output
@@ -142,0 +210,0 @@ #### .prettify()

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

Improved metrics

Total package byte prevSize

28803

increased by29.18%

Number of package files

4

increased by33.33%

Lines of code

584

increased by25.86%

Number of lines in readme file

239

increased by39.77%

Dependency changes

• Updatedlodash@^4.17.15