degausser - npm Package Compare versions

Comparing version 2.2.1 to 2.3.0

dist/degausser.js

@@ -7,2 +7,18 @@ (function (global, factory) {
 
+  function _typeof(obj) {
+    "@babel/helpers - typeof";
+
+    if (typeof Symbol === "function" && typeof Symbol.iterator === "symbol") {
+      _typeof = function (obj) {
+        return typeof obj;
+      };
+    } else {
+      _typeof = function (obj) {
+        return obj && typeof Symbol === "function" && obj.constructor === Symbol && obj !== Symbol.prototype ? "symbol" : typeof obj;
+      };
+    }
+
+    return _typeof(obj);
+  }
+
   function _classCallCheck(instance, Constructor) {
@@ -195,6 +211,91 @@ if (!(instance instanceof Constructor)) {
   var phrasingConstructs = ['a', 'abbr', 'audio', 'b', 'bdo', 'br', 'button', 'canvas', 'cite', 'code', 'command', 'data', 'datalist', 'dfn', 'em', 'embed', 'i', 'iframe', 'img', 'input', 'kbd', 'keygen', 'label', 'mark', 'math', 'meter', 'noscript', 'object', 'output', 'progress', 'q', 'ruby', 'samp', 'script', 'select', 'small', 'span', 'strong', 'sub', 'sup', 'svg', 'textarea', 'time', 'var', 'video', 'wbr', // special cases
-  'map', 'area'];
+  'map', 'area']; // copied from readium-cfi-js library
+  // original function called "isElementBlacklisted"
 
+  var isElementBlacklisted = function isElementBlacklisted(element, classBlacklist, elementBlacklist, idBlacklist) {
+    if (classBlacklist && classBlacklist.length) {
+      var classList = getClassNameArray(element);
+
+      if (classList.length === 1 && classBlacklist.includes(classList[0])) {
+        return true;
+      }
+
+      if (classList.length && intersection(classBlacklist, classList).length) {
+        return true;
+      }
+    }
+
+    if (elementBlacklist && elementBlacklist.length) {
+      if (element.tagName) {
+        var isElementInBlacklist = elementBlacklist.find(function (blacklistedTag) {
+          return matchesLocalNameOrElement(element, blacklistedTag.toLowerCase());
+        });
+
+        if (isElementInBlacklist) {
+          return true;
+        }
+      }
+    }
+
+    if (idBlacklist && idBlacklist.length) {
+      var id = element.id;
+
+      if (id && id.length && idBlacklist.includes(id)) {
+        return true;
+      }
+    }
+
+    return false;
+  };
+
+  var intersection = function intersection(array1, array2) {
+    var intersectionArray = [];
+
+    var _iterator2 = _createForOfIteratorHelper(array1),
+        _step2;
+
+    try {
+      for (_iterator2.s(); !(_step2 = _iterator2.n()).done;) {
+        var value = _step2.value;
+        var index = array2.indexOf(value);
+
+        if (index !== -1) {
+          intersectionArray.push(value);
+        }
+      }
+    } catch (err) {
+      _iterator2.e(err);
+    } finally {
+      _iterator2.f();
+    }
+
+    return intersectionArray;
+  };
+
+  var getClassNameArray = function getClassNameArray(element) {
+    var className = element.className;
+
+    if (typeof className === 'string') {
+      return className.split(/\s/);
+    }
+
+    if (_typeof(className) === 'object' && 'baseVal' in className) {
+      return className.baseVal.split(/\s/);
+    }
+
+    return [];
+  };
+
+  var matchesLocalNameOrElement = function matchesLocalNameOrElement(element, otherNameOrElement) {
+    if (typeof otherNameOrElement === 'string') {
+      return (element.localName || element.nodeName) === otherNameOrElement;
+    }
+
+    return element === otherNameOrElement;
+  };
+
   var StringCollector = /*#__PURE__*/function () {
     function StringCollector() {
+      var options = arguments.length > 0 && arguments[0] !== undefined ? arguments[0] : {};
+
       _classCallCheck(this, StringCollector);
@@ -204,2 +305,3 @@
       this.text = [];
+      this.options = options;
       this.hasEncounteredFirstCell = false;
@@ -270,2 +372,6 @@ this.lastBreak = null;
       value: function processElementNode(node, isOpening) {
+        if (isElementBlacklisted(node, this.options.classBlacklist, this.options.elementBlacklist, this.options.idBlacklist)) {
+          return true;
+        }
+
         var tag = node.tagName.toLowerCase(); // Special case for Preformatted
@@ -374,2 +480,4 @@
     function MapCollector() {
+      var options = arguments.length > 0 && arguments[0] !== undefined ? arguments[0] : {};
+
       _classCallCheck(this, MapCollector);
@@ -379,2 +487,3 @@
       this.text = [];
+      this.options = options;
       this.hasEncounteredFirstCell = false;
@@ -502,2 +611,6 @@ this.lastBreak = null;
       value: function processElementNode(node, isOpening) {
+        if (isElementBlacklisted(node, this.options.classBlacklist, this.options.elementBlacklist, this.options.idBlacklist)) {
+          return true;
+        }
+
         var tag = node.tagName.toLowerCase(); // Special case for Preformatted
@@ -743,6 +856,6 @@
     var options = arguments.length > 1 && arguments[1] !== undefined ? arguments[1] : {};
-    var collector = new StringCollector();
+    var collector = new StringCollector(options);
 
     if (options.map) {
-      collector = new MapCollector();
+      collector = new MapCollector(options);
     }
@@ -749,0 +862,0 @@

package.json

 {
     "name": "degausser",
-    "version": "2.2.1",
+    "version": "2.3.0",
     "description": "Transforms HTML to plain text by eliminating tags from a document.",
@@ -32,9 +32,9 @@ "author": "FlowPub",
     "devDependencies": {
-        "@babel/core": "7.12.9",
-        "@babel/preset-env": "7.12.7",
+        "@babel/core": "7.13.15",
+        "@babel/preset-env": "7.13.15",
         "@rollup/plugin-node-resolve": "8.4.0",
         "glob": "7.1.6",
         "jest": "26.6.3",
-        "prettier": "2.1.2",
-        "rollup": "2.23.0",
+        "prettier": "2.2.1",
+        "rollup": "2.39.0",
         "rollup-plugin-babel": "4.4.0"
@@ -41,0 +41,0 @@ },

src/degausser.js

@@ -6,6 +6,6 @@ import { StringCollector } from './stringCollector'
 export const degausser = (parentNode, options = {}) => {
-  let collector = new StringCollector()
+  let collector = new StringCollector(options)
 
   if (options.map) {
-    collector = new MapCollector()
+    collector = new MapCollector(options)
   }
@@ -16,8 +16,3 @@
 
-export const getRangeFromOffset = (
-  start,
-  end,
-  doc = document,
-  map = null,
-) => {
+export const getRangeFromOffset = (start, end, doc = document, map = null) => {
   const docType = doc.nodeType
@@ -24,0 +19,0 @@ if (

src/mapCollector.js

@@ -8,2 +8,3 @@ import {
   phrasingConstructs,
+  isElementBlacklisted,
 } from './util'
@@ -17,6 +18,8 @@
 export class MapCollector {
-  constructor() {
+  constructor(options = {}) {
     this.map = []
     this.text = []
 
+    this.options = options
+
     this.hasEncounteredFirstCell = false
@@ -131,2 +134,13 @@ this.lastBreak = null
   processElementNode(node, isOpening) {
+    if (
+      isElementBlacklisted(
+        node,
+        this.options.classBlacklist,
+        this.options.elementBlacklist,
+        this.options.idBlacklist,
+      )
+    ) {
+      return true
+    }
+
     const tag = node.tagName.toLowerCase()
@@ -252,11 +266,25 @@
 
-          if (entity.node.nodeType === Node.TEXT_NODE || entity.node.tagName === 'img') {
-            const nodeContent = entity.node.tagName === 'img' ?
-            entity.node.getAttribute('alt').normalize() : 
-            entity.node.textContent.normalize()
+          if (
+            entity.node.nodeType === Node.TEXT_NODE ||
+            entity.node.tagName === 'img'
+          ) {
+            const nodeContent =
+              entity.node.tagName === 'img'
+                ? entity.node.getAttribute('alt').normalize()
+                : entity.node.textContent.normalize()
 
-            for (let charInMap = 0, charInNode = 0; charInNode < nodeContent.length; ++charInNode) {
-              const isEqual = entity.content.charAt(charInMap) === nodeContent.charAt(charInNode)
-              const isMapWhitespace = isCharWhitespace(entity.content.charCodeAt(charInMap))
-              const isNodeWhitespace = isCharWhitespace(nodeContent.charCodeAt(charInNode))
+            for (
+              let charInMap = 0, charInNode = 0;
+              charInNode < nodeContent.length;
+              ++charInNode
+            ) {
+              const isEqual =
+                entity.content.charAt(charInMap) ===
+                nodeContent.charAt(charInNode)
+              const isMapWhitespace = isCharWhitespace(
+                entity.content.charCodeAt(charInMap),
+              )
+              const isNodeWhitespace = isCharWhitespace(
+                nodeContent.charCodeAt(charInNode),
+              )
 
@@ -268,7 +296,9 @@ if (isEqual || (isMapWhitespace && isNodeWhitespace)) {
                   after: charInMap - 1,
-                  position: charInNode
+                  position: charInNode,
                 }
                 whitespace.push(skips)
               } else {
-                throw new Error(`Degauss error, character mismatch and not a whitespace`)
+                throw new Error(
+                  `Degauss error, character mismatch and not a whitespace`,
+                )
               }
@@ -275,0 +305,0 @@ }

src/stringCollector.js

@@ -7,8 +7,10 @@ import {
   phrasingConstructs,
+  isElementBlacklisted,
 } from './util'
 
 export class StringCollector {
-  constructor() {
+  constructor(options = {}) {
     this.runs = []
     this.text = []
+    this.options = options
 
@@ -74,2 +76,13 @@ this.hasEncounteredFirstCell = false
   processElementNode(node, isOpening) {
+    if (
+      isElementBlacklisted(
+        node,
+        this.options.classBlacklist,
+        this.options.elementBlacklist,
+        this.options.idBlacklist,
+      )
+    ) {
+      return true
+    }
+
     const tag = node.tagName.toLowerCase()
@@ -76,0 +89,0 @@

src/util.js

@@ -147,2 +147,72 @@ function autoBind() {
 
+// copied from readium-cfi-js library
+// original function called "isElementBlacklisted"
+const isElementBlacklisted = (
+  element,
+  classBlacklist,
+  elementBlacklist,
+  idBlacklist,
+) => {
+  if (classBlacklist && classBlacklist.length) {
+    const classList = getClassNameArray(element)
+    if (classList.length === 1 && classBlacklist.includes(classList[0])) {
+      return true
+    }
+    if (classList.length && intersection(classBlacklist, classList).length) {
+      return true
+    }
+  }
+
+  if (elementBlacklist && elementBlacklist.length) {
+    if (element.tagName) {
+      const isElementInBlacklist = elementBlacklist.find((blacklistedTag) =>
+        matchesLocalNameOrElement(element, blacklistedTag.toLowerCase()),
+      )
+
+      if (isElementInBlacklist) {
+        return true
+      }
+    }
+  }
+
+  if (idBlacklist && idBlacklist.length) {
+    const { id } = element
+    if (id && id.length && idBlacklist.includes(id)) {
+      return true
+    }
+  }
+
+  return false
+}
+
+const intersection = (array1, array2) => {
+  const intersectionArray = []
+  for (let value of array1) {
+    const index = array2.indexOf(value)
+    if (index !== -1) {
+      intersectionArray.push(value)
+    }
+  }
+
+  return intersectionArray
+}
+
+const getClassNameArray = (element) => {
+  const { className } = element
+  if (typeof className === 'string') {
+    return className.split(/\s/)
+  }
+  if (typeof className === 'object' && 'baseVal' in className) {
+    return className.baseVal.split(/\s/)
+  }
+  return []
+}
+
+const matchesLocalNameOrElement = (element, otherNameOrElement) => {
+  if (typeof otherNameOrElement === 'string') {
+    return (element.localName || element.nodeName) === otherNameOrElement
+  }
+  return element === otherNameOrElement
+}
+
 export {
@@ -155,3 +225,4 @@ autoBind,
   phrasingConstructs,
-  isCharWhitespace
+  isElementBlacklisted,
+  isCharWhitespace,
 }

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

48355

increased by13.57%

Lines of code

1491

increased by14.08%

No dependency changes