degausser - npm Package Compare versions

Comparing version 2.2.0 to 2.2.1

dist/degausser.js

 (function (global, factory) {
   typeof exports === 'object' && typeof module !== 'undefined' ? factory(exports) :
   typeof define === 'function' && define.amd ? define(['exports'], factory) :
-  (global = global || self, factory(global.degausser = {}));
+  (global = typeof globalThis !== 'undefined' ? globalThis : global || self, factory(global.degausser = {}));
 }(this, (function (exports) { 'use strict';
@@ -46,5 +46,8 @@
 
-  function _createForOfIteratorHelper(o) {
+  function _createForOfIteratorHelper(o, allowArrayLike) {
+    var it;
+
     if (typeof Symbol === "undefined" || o[Symbol.iterator] == null) {
-      if (Array.isArray(o) || (o = _unsupportedIterableToArray(o))) {
+      if (Array.isArray(o) || (it = _unsupportedIterableToArray(o)) || allowArrayLike && o && typeof o.length === "number") {
+        if (it) o = it;
         var i = 0;
@@ -75,4 +78,3 @@
 
-    var it,
-        normalCompletion = true,
+    var normalCompletion = true,
         didErr = false,
@@ -190,21 +192,2 @@ err;
 
-  var collectTrimmedWhitespace = function collectTrimmedWhitespace(string) {
-    var spaces = [];
-    var encounteredWhitespace = false;
-
-    for (var index = 0; index < string.length; ++index) {
-      if (isCharWhitespace(string.charCodeAt(index))) {
-        if (encounteredWhitespace) {
-          spaces.push(index);
-        } else {
-          encounteredWhitespace = true;
-        }
-      } else {
-        encounteredWhitespace = false;
-      }
-    }
-
-    return spaces;
-  };
-
   var blacklist = ['base', 'command', 'link', 'meta', 'noscript', 'script', 'style', 'title', // special cases
@@ -482,3 +465,2 @@ // "html",
 
-            var whitespace = collectTrimmedWhitespace(textMap.node.textContent);
             blockMap.push({
@@ -489,4 +471,3 @@ type: MapType.TEXT,
               length: shrunkText.length,
-              content: shrunkText,
-              whitespace: whitespace
+              content: shrunkText
             });
@@ -645,6 +626,31 @@ fullText = fullText.slice(index + shrunkText.length);
               case MapType.TEXT:
+                // TODO: Tests
+                var whitespace = [];
+
+                if (entity.node.nodeType === Node.TEXT_NODE || entity.node.tagName === 'img') {
+                  var nodeContent = entity.node.tagName === 'img' ? entity.node.getAttribute('alt').normalize() : entity.node.textContent.normalize();
+
+                  for (var charInMap = 0, charInNode = 0; charInNode < nodeContent.length; ++charInNode) {
+                    var isEqual = entity.content.charAt(charInMap) === nodeContent.charAt(charInNode);
+                    var isMapWhitespace = isCharWhitespace(entity.content.charCodeAt(charInMap));
+                    var isNodeWhitespace = isCharWhitespace(nodeContent.charCodeAt(charInNode));
+
+                    if (isEqual || isMapWhitespace && isNodeWhitespace) {
+                      ++charInMap;
+                    } else if (isMapWhitespace || isNodeWhitespace) {
+                      var skips = {
+                        after: charInMap - 1,
+                        position: charInNode
+                      };
+                      whitespace.push(skips);
+                    } else {
+                      throw new Error("Degauss error, character mismatch and not a whitespace");
+                    }
+                  }
+                }
+
                 result.push({
                   node: entity.node,
                   content: entity.content,
-                  whitespace: entity.whitespace,
+                  whitespace: whitespace,
                   start: runningIndex,
@@ -778,3 +784,3 @@ length: entity.length
 
-              if (whitespaceEntry <= adjustedStart) {
+              if (whitespaceEntry.after < adjustedStart) {
                 ++skips;
@@ -807,3 +813,3 @@ }
 
-              if (_whitespaceEntry <= adjustedEnd) {
+              if (_whitespaceEntry.after < adjustedEnd) {
                 ++_skips;
@@ -810,0 +816,0 @@ }

package.json

 {
     "name": "degausser",
-    "version": "2.2.0",
+    "version": "2.2.1",
     "description": "Transforms HTML to plain text by eliminating tags from a document.",
@@ -32,4 +32,4 @@ "author": "FlowPub",
     "devDependencies": {
-        "@babel/core": "7.12.3",
-        "@babel/preset-env": "7.12.1",
+        "@babel/core": "7.12.9",
+        "@babel/preset-env": "7.12.7",
         "@rollup/plugin-node-resolve": "8.4.0",
@@ -36,0 +36,0 @@ "glob": "7.1.6",

src/degausser.js

@@ -46,3 +46,3 @@ import { StringCollector } from './stringCollector'
         for (const whitespaceEntry of entry.whitespace) {
-          if (whitespaceEntry <= adjustedStart) {
+          if (whitespaceEntry.after < adjustedStart) {
             ++skips
@@ -64,3 +64,3 @@ }
         for (const whitespaceEntry of entry.whitespace) {
-          if (whitespaceEntry <= adjustedEnd) {
+          if (whitespaceEntry.after < adjustedEnd) {
             ++skips
@@ -67,0 +67,0 @@ }

src/mapCollector.js

@@ -6,3 +6,3 @@ import {
   collapseWhitespace,
-  collectTrimmedWhitespace,
+  isCharWhitespace,
   phrasingConstructs,
@@ -98,4 +98,2 @@ } from './util'
 
-      const whitespace = collectTrimmedWhitespace(textMap.node.textContent)
-
       blockMap.push({
@@ -107,3 +105,2 @@ type: MapType.TEXT,
         content: shrunkText,
-        whitespace
       })
@@ -251,6 +248,34 @@
         case MapType.TEXT:
+          // TODO: Tests
+
+          const whitespace = []
+
+          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))
+
+              if (isEqual || (isMapWhitespace && isNodeWhitespace)) {
+                ++charInMap
+              } else if (isMapWhitespace || isNodeWhitespace) {
+                const skips = {
+                  after: charInMap - 1,
+                  position: charInNode
+                }
+                whitespace.push(skips)
+              } else {
+                throw new Error(`Degauss error, character mismatch and not a whitespace`)
+              }
+            }
+          }
+
           result.push({
             node: entity.node,
             content: entity.content,
-            whitespace: entity.whitespace,
+            whitespace: whitespace,
             start: runningIndex,
@@ -257,0 +282,0 @@ length: entity.length,

src/util.js

@@ -81,21 +81,2 @@ function autoBind() {
 
-const collectTrimmedWhitespace = (string) => {
-  const spaces = []
-
-  let encounteredWhitespace = false
-  for (let index = 0; index < string.length; ++index) {
-    if (isCharWhitespace(string.charCodeAt(index))) {
-      if (encounteredWhitespace) {
-        spaces.push(index)
-      } else {
-        encounteredWhitespace = true
-      }
-    } else {
-      encounteredWhitespace = false
-    }
-  }
-
-  return spaces
-}
-
 const blacklist = [
@@ -173,4 +154,4 @@ 'base',
   collapseWhitespace,
-  collectTrimmedWhitespace,
   phrasingConstructs,
+  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

42576

increased by3.66%

Lines of code

1307

increased by0.69%

No dependency changes