Merge pull request #4 from md2docx/handle-edge-cases

Handle edge cases

Author: mayank1513

examples/vite/tsconfig.json

@@ -5,5 +5,5 @@
     "rootDir": ".",
     "outDir": "dist"
   },
-  "include": ["src", "vite.config.ts"]
+  "include": ["src", "vite.config.mts"]
 }

examples/vite/vite.config.ts renamed to examples/vite/vite.config.mts

@@ -1,5 +1,11 @@
 # @m2d/html
 
+## 1.1.8
+
+### Patch Changes
+
+- 4a811aa: Fix HTML parsing issues in case of empty tags.
+
 ## 1.1.7
 
 ### Patch Changes

lib/CHANGELOG.md

@@ -2,7 +2,7 @@
   "name": "@m2d/html",
   "author": "Mayank Kumar Chaudhari (https://mayank-chaudhari.vercel.app)",
   "private": false,
-  "version": "1.1.7",
+  "version": "1.1.8",
   "description": "Extend MDAST by parsing embedded HTML in Markdown. Converts HTML into structured MDAST nodes compatible with @m2d/core for DOCX generation.",
   "license": "MPL-2.0",
   "main": "./dist/index.js",

lib/package.json

@@ -10,6 +10,7 @@ import {
   BlockContent,
   TableRow,
   Html,
+  Literal,
 } from "@m2d/core";
 import { standardizeColor } from "./utils";
 import { AlignmentType, BorderStyle, IBorderOptions } from "docx";
@@ -69,6 +70,8 @@ const INLINE_TAGS = [
   "WBR",
 ] as const;
 
+const EMPTY_TAGS = ["br", "hr", "img", "input"];
+
 /**
  * Mapping of DOM tag names to MDAST node types.
  */
@@ -499,8 +502,8 @@ const preprocess = (pNode: Parent, isRoot = true) => {
   for (const node of pNode.children) {
     if ((node as Parent).children?.length) preprocess(node as Parent, false);
     // match only inline non-self-closing html nodes.
-    if (node.type === "html" && /^<[^>]*[^/]>$/.test(node.value)) {
-      const tag = node.value.split(" ")[0].replace(/^<|>$/g, "");
+    const tag = (node as Literal).value?.split(" ")[0].replace(/^<|\/?>$/g, "");
+    if (node.type === "html" && !EMPTY_TAGS.includes(tag) && /^<[^>]*[^/]>$/.test(node.value)) {
       // ending tag
       if (tag[0] === "/") {
         const hNode = htmlNodeStack.shift();
@@ -516,7 +519,9 @@ const preprocess = (pNode: Parent, isRoot = true) => {
       children.push(node);
     }
 
-    const isSelfClosingTag = node.type === "html" && /^<[^>]*\/>$/.test(node.value);
+    const isSelfClosingTag =
+      node.type === "html" && (EMPTY_TAGS.includes(tag) || /^<[^>]*\/>$/.test(node.value));
+
     // self closing tags
     if (isSelfClosingTag && !isRoot) {
       // @ts-expect-error -- ok

lib/src/index.ts

@@ -1,5 +1,12 @@
 # @repo/shared
 
+## 0.0.14
+
+### Patch Changes
+
+- Updated dependencies [4a811aa]
+  - @m2d/html@1.1.8
+
 ## 0.0.13
 
 ### Patch Changes

packages/shared/CHANGELOG.md

@@ -1,6 +1,6 @@
 {
   "name": "@repo/shared",
-  "version": "0.0.13",
+  "version": "0.0.14",
   "private": true,
   "sideEffects": false,
   "main": "./dist/index.js",

packages/shared/package.json

@@ -1,6 +1,12 @@
 ## 13. HTML Elements
 
-Using HTML for additional interactivity.
+### Table with html
+
+| Left-aligned             | Center-aligned | Right-aligned |
+| :----------------------- | :------------: | ------------: |
+| Text 1 <br>- second line |     Text 2     |        Text 3 |
+
+### Using HTML for additional interactivity.
 
 <details>
   <summary>Click to expand</summary>