Decode numeric entities and quotes. Fixes #99

oozcitak · Jul 30, 2021 · aba7aa0 · aba7aa0
1 parent 7146d34
commit aba7aa0
Show file tree

Hide file tree

Showing 6 changed files with 29 additions and 17 deletions.
diff --git a/src/readers/BaseReader.ts b/src/readers/BaseReader.ts
@@ -2,11 +2,18 @@ import { XMLBuilderOptions, ExpandObject, XMLBuilder } from "../interfaces"
 import { sanitizeInput } from "../builder/dom"
 
 /**
- * Pre-serializes XML nodes.
+ * Parses XML nodes.
  */
 export abstract class BaseReader<U extends string | ExpandObject> {
 
   protected _builderOptions: XMLBuilderOptions
+  private static _entityTable: { [key: string]: string } = {
+    "lt": "<",
+    "gt": ">",
+    "amp": "&",
+    "quot": '"',
+    "apos": "'",
+  }
 
   /**
    * Initializes a new instance of `BaseReader`.
@@ -60,9 +67,13 @@ export abstract class BaseReader<U extends string | ExpandObject> {
    * @param text - text value to serialize
    */
   _decodeText(text: string): string {
-    return text == null ? text : text.replace(/&lt;/g, '<')
-      .replace(/&gt;/g, '>')
-      .replace(/&amp;/g, '&')
+    if (text == null) return text
+
+    return text.replace(/&(quot|amp|apos|lt|gt);/g, (_match, tag) =>
+      BaseReader._entityTable[tag]
+    ).replace(/&#(?:x([a-fA-F0-9]+)|([0-9]+));/g, (_match, hexStr, numStr) =>
+      String.fromCodePoint(parseInt(hexStr || numStr, hexStr ? 16 : 10))
+    )
   }
 
   /**
@@ -71,9 +82,7 @@ export abstract class BaseReader<U extends string | ExpandObject> {
    * @param text - attribute value to serialize
    */
   _decodeAttributeValue(text: string): string {
-    return text == null ? text : text.replace(/&lt;/g, '<')
-      .replace(/&gt;/g, '>')
-      .replace(/&amp;/g, '&')
+    return this._decodeText(text)
   }
 
   /**

diff --git a/src/writers/BaseWriter.ts b/src/writers/BaseWriter.ts
@@ -5,7 +5,6 @@ import {
 } from "@oozcitak/dom/lib/dom/interfaces"
 import { LocalNameSet } from "@oozcitak/dom/lib/serializer/LocalNameSet"
 import { NamespacePrefixMap } from "@oozcitak/dom/lib/serializer/NamespacePrefixMap"
-import { InvalidStateError } from "@oozcitak/dom/lib/dom/DOMException"
 import { namespace as infraNamespace } from "@oozcitak/infra"
 import { xml_isName, xml_isLegalChar, xml_isPubidChar } from "@oozcitak/dom/lib/algorithm"
 
@@ -929,7 +928,7 @@ export abstract class BaseWriter<T extends BaseWriterOptions, U extends XMLSeria
      * 5. Replace any occurrences of ">" in markup by "&gt;".
      * 6. Return the value of markup.
      */
-    const markup = node.data.replace(/(?!&(lt|gt|amp|apos|quot);)&/g, '&amp;')
+    const markup = node.data.replace(/(?!&([^&;]*);)&/g, '&amp;')
       .replace(/</g, '&lt;')
       .replace(/>/g, '&gt;')
 
@@ -1598,7 +1597,7 @@ export abstract class BaseWriter<T extends BaseWriterOptions, U extends XMLSeria
      * grammar requirement in the XML specification's AttValue production by
      * also replacing ">" characters.
      */
-    return value.replace(/(?!&(lt|gt|amp|apos|quot);)&/g, '&amp;')
+    return value.replace(/(?!&([^&;]*);)&/g, '&amp;')
       .replace(/</g, '&lt;')
       .replace(/>/g, '&gt;')
       .replace(/"/g, '&quot;')

diff --git a/test/callback/basic.test.ts b/test/callback/basic.test.ts
@@ -173,8 +173,8 @@ describe('basic callback API tests', () => {
     xmlStream.ele(obj).end()
 
     $$.expectCBResult(xmlStream,
-      '<root att="attribute value with &amp; and &amp;#38;">' +
-      'XML entities for ampersand are &amp; and &amp;#38;.' +
+      '<root att="attribute value with &amp; and &amp;">' +
+      'XML entities for ampersand are &amp; and &amp;.' +
       '</root>', done)
   })
 

diff --git a/test/issues/issue-016.test.ts b/test/issues/issue-016.test.ts
@@ -9,7 +9,7 @@ describe("Replicate issue", () => {
         { format: "xml" }
       )
     ).toBe(
-      '<?xml version="1.0"?><example>&lt;p&gt;Hello&amp;nbsp;World&lt;/p&gt;</example>'
+      '<?xml version="1.0"?><example>&lt;p&gt;Hello&nbsp;World&lt;/p&gt;</example>'
     );
     expect(
       $$.convert(

diff --git a/test/issues/issue-099.test.ts b/test/issues/issue-099.test.ts
@@ -3,10 +3,14 @@ import $$ from "../TestHelpers";
 describe("Replicate issue", () => {
   // https://github.com/oozcitak/xmlbuilder2/issues/99
   test(`#99 - Most XML entities not being decoded by reader (with unknown entity)`, () => {
-    expect(() => $$.create('<ssid>Me &amp; Myself&apos;s WiFi &copy;&#x1D306;</ssid>')).toThrow()
+    const xmlResponse = $$.create('<ssid>Me &amp; Myself&apos;s WiFi &copy;&#x1D306;</ssid>');
+    const networkResponse = xmlResponse.end({format: 'object'});
+    expect(networkResponse).toEqual({
+      ssid: `Me &amp; Myself's WiFi &copy;𝌆`
+    })
   })
 
-  test(`#99 - Most XML entities not being decoded by reader`, () => {
+  test(`#99 - Most XML entities not being decoded by reader (without unknown entity)`, () => {
     const xmlResponse = $$.create('<ssid>Me &amp; Myself&apos;s WiFi &#x1D306;</ssid>');
     const networkResponse = xmlResponse.end({format: 'object'});
     expect(networkResponse).toEqual({

diff --git a/test/writers/XMLWriter.test.ts b/test/writers/XMLWriter.test.ts
@@ -596,8 +596,8 @@ describe('XMLWriter', () => {
 
     expect($$.create(obj).end()).toBe(
       '<?xml version="1.0"?>' +
-      '<root att="attribute value with &amp; and &amp;#38;">' +
-      'XML entities for ampersand are &amp; and &amp;#38;.' +
+      '<root att="attribute value with &amp; and &amp;">' +
+      'XML entities for ampersand are &amp; and &amp;.' +
       '</root>')
   })