node-html-parser 6.1.6 → 6.1.8

package/CHANGELOG.md

@@ -2,6 +2,22 @@
 
 All notable changes to this project will be documented in this file. See [standard-version](https://github.com/conventional-changelog/standard-version) for commit guidelines.
 
+### [6.1.8](https://github.com/taoqf/node-fast-html-parser/compare/v6.1.7...v6.1.8) (2023-09-08)
+
+
+### Bug Fixes
+
+* [#248](https://github.com/taoqf/node-fast-html-parser/issues/248) void tag ([01ff9f4](https://github.com/taoqf/node-fast-html-parser/commit/01ff9f41e416d0ef0c47bc40c282d6aeb405e100))
+
+### [6.1.7](https://github.com/taoqf/node-fast-html-parser/compare/v6.1.6...v6.1.7) (2023-09-08)
+
+
+### Bug Fixes
+
+* block text elements [#242](https://github.com/taoqf/node-fast-html-parser/issues/242) ([97e593c](https://github.com/taoqf/node-fast-html-parser/commit/97e593c2bdefbfefd35c2c9ccaef914517d60fe7))
+* docs [#245](https://github.com/taoqf/node-fast-html-parser/issues/245) ([f656c7b](https://github.com/taoqf/node-fast-html-parser/commit/f656c7b2de2a62a9a363f797c6f03a2477e7312e))
+* remove test code ([d3980c5](https://github.com/taoqf/node-fast-html-parser/commit/d3980c5fb7744d6fda4b270619bb581a19a4cb18))
+
 ### [6.1.6](https://github.com/taoqf/node-fast-html-parser/compare/v6.1.5...v6.1.6) (2023-08-17)
 
 

package/README.md

@@ -10,7 +10,6 @@ no closing `<li>`, `<td>` etc).
 
 ## Install
 
-
 ```shell
 npm install --save node-html-parser
 ```
@@ -86,7 +85,7 @@ Parse the data provided, and return the root of the generated DOM.
     comment: false,            // retrieve comments (hurts performance slightly)
     voidTag:{
       tags: ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr'],	// optional and case insensitive, default value is ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr']
-      addClosingSlash: true     // optional, default false. void tag serialisation, add a final slash <br/>
+      closingSlash: true     // optional, default false. void tag serialisation, add a final slash <br/>
     },
     blockTextElements: {
       script: true,	// keep text content when parsing
@@ -293,7 +292,7 @@ Get class names.
 
 Clone a node.
 
-#### getElementById(id: string): HTMLElement;
+#### getElementById(id: string): HTMLElement
 
 Get element by it's ID.
 

package/dist/main.js

@@ -222,12 +222,12 @@ define("void-tag", ["require", "exports"], function (require, exports) {
             this.addClosingSlash = addClosingSlash;
             if (Array.isArray(tags)) {
                 this.voidTags = tags.reduce(function (set, tag) {
-                    return set.add(tag.toLowerCase());
+                    return set.add(tag.toLowerCase()).add(tag.toUpperCase()).add(tag);
                 }, new Set());
             }
             else {
                 this.voidTags = ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr'].reduce(function (set, tag) {
-                    return set.add(tag);
+                    return set.add(tag.toLowerCase()).add(tag.toUpperCase()).add(tag);
                 }, new Set());
             }
         }
@@ -1294,36 +1294,6 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
     // https://html.spec.whatwg.org/multipage/custom-elements.html#valid-custom-element-name
     var kMarkupPattern = /<!--[\s\S]*?-->|<(\/?)([a-zA-Z][-.:0-9_a-zA-Z]*)((?:\s+[^>]*?(?:(?:'[^']*')|(?:"[^"]*"))?)*)\s*(\/?)>/g;
     var kAttributePattern = /(?:^|\s)(id|class)\s*=\s*((?:'[^']*')|(?:"[^"]*")|\S+)/gi;
-    var kSelfClosingElements = {
-        area: true,
-        AREA: true,
-        base: true,
-        BASE: true,
-        br: true,
-        BR: true,
-        col: true,
-        COL: true,
-        hr: true,
-        HR: true,
-        img: true,
-        IMG: true,
-        input: true,
-        INPUT: true,
-        link: true,
-        LINK: true,
-        meta: true,
-        META: true,
-        source: true,
-        SOURCE: true,
-        embed: true,
-        EMBED: true,
-        param: true,
-        PARAM: true,
-        track: true,
-        TRACK: true,
-        wbr: true,
-        WBR: true,
-    };
     var kElementsClosedByOpening = {
         li: { li: true, LI: true },
         LI: { li: true, LI: true },
@@ -1381,7 +1351,9 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
             style: true,
             pre: true,
         };
-        var element_names = Object.keys(elements);
+        var element_names = Object.keys(elements).filter(function (name) {
+            return Boolean(elements[name]);
+        });
         var kBlockTextElements = element_names.map(function (it) { return new RegExp("^".concat(it, "$"), 'i'); });
         var kIgnoreElements = element_names.filter(function (it) { return elements[it]; }).map(function (it) { return new RegExp("^".concat(it, "$"), 'i'); });
         function element_should_be_ignore(tag) {
@@ -1488,7 +1460,7 @@ define("nodes/html", ["require", "exports", "css-select", "he", "back", "matcher
                 }
             }
             // Handle closing tags or self-closed elements (ie </tag> or <br>)
-            if (leadingSlash || closingSlash || kSelfClosingElements[tagName]) {
+            if (leadingSlash || closingSlash || voidTag.isVoidElement(tagName)) {
                 while (true) {
                     if (noNestedTagIndex != null && (tagName === 'a' || tagName === 'A'))
                         noNestedTagIndex = undefined;

package/dist/nodes/html.js

@@ -960,36 +960,6 @@ exports.default = HTMLElement;
 // https://html.spec.whatwg.org/multipage/custom-elements.html#valid-custom-element-name
 var kMarkupPattern = /<!--[\s\S]*?-->|<(\/?)([a-zA-Z][-.:0-9_a-zA-Z]*)((?:\s+[^>]*?(?:(?:'[^']*')|(?:"[^"]*"))?)*)\s*(\/?)>/g;
 var kAttributePattern = /(?:^|\s)(id|class)\s*=\s*((?:'[^']*')|(?:"[^"]*")|\S+)/gi;
-var kSelfClosingElements = {
-    area: true,
-    AREA: true,
-    base: true,
-    BASE: true,
-    br: true,
-    BR: true,
-    col: true,
-    COL: true,
-    hr: true,
-    HR: true,
-    img: true,
-    IMG: true,
-    input: true,
-    INPUT: true,
-    link: true,
-    LINK: true,
-    meta: true,
-    META: true,
-    source: true,
-    SOURCE: true,
-    embed: true,
-    EMBED: true,
-    param: true,
-    PARAM: true,
-    track: true,
-    TRACK: true,
-    wbr: true,
-    WBR: true,
-};
 var kElementsClosedByOpening = {
     li: { li: true, LI: true },
     LI: { li: true, LI: true },
@@ -1047,7 +1017,9 @@ function base_parse(data, options) {
         style: true,
         pre: true,
     };
-    var element_names = Object.keys(elements);
+    var element_names = Object.keys(elements).filter(function (name) {
+        return Boolean(elements[name]);
+    });
     var kBlockTextElements = element_names.map(function (it) { return new RegExp("^".concat(it, "$"), 'i'); });
     var kIgnoreElements = element_names.filter(function (it) { return elements[it]; }).map(function (it) { return new RegExp("^".concat(it, "$"), 'i'); });
     function element_should_be_ignore(tag) {
@@ -1154,7 +1126,7 @@ function base_parse(data, options) {
             }
         }
         // Handle closing tags or self-closed elements (ie </tag> or <br>)
-        if (leadingSlash || closingSlash || kSelfClosingElements[tagName]) {
+        if (leadingSlash || closingSlash || voidTag.isVoidElement(tagName)) {
             while (true) {
                 if (noNestedTagIndex != null && (tagName === 'a' || tagName === 'A'))
                     noNestedTagIndex = undefined;

package/dist/void-tag.js

@@ -6,12 +6,12 @@ var VoidTag = /** @class */ (function () {
         this.addClosingSlash = addClosingSlash;
         if (Array.isArray(tags)) {
             this.voidTags = tags.reduce(function (set, tag) {
-                return set.add(tag.toLowerCase());
+                return set.add(tag.toLowerCase()).add(tag.toUpperCase()).add(tag);
             }, new Set());
         }
         else {
             this.voidTags = ['area', 'base', 'br', 'col', 'embed', 'hr', 'img', 'input', 'link', 'meta', 'param', 'source', 'track', 'wbr'].reduce(function (set, tag) {
-                return set.add(tag);
+                return set.add(tag.toLowerCase()).add(tag.toUpperCase()).add(tag);
             }, new Set());
         }
     }

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "node-html-parser",
-	"version": "6.1.6",
+	"version": "6.1.8",
 	"description": "A very fast HTML parser, generating a simplified DOM, with basic element query support.",
 	"main": "dist/index.js",
 	"types": "dist/index.d.ts",