Merge pull request #93 from susanwere/fix-pasting-markdown-with-html

fix pasting of markdown containing links

Author: keithamus

Diff for: src/paste-markdown-html.ts

@@ -37,7 +37,7 @@ function onPaste(event: ClipboardEvent) {
   // Generate DOM tree from HTML string
   const parser = new DOMParser()
   const doc = parser.parseFromString(textHTMLClean, 'text/html')
-  const walker = doc.createTreeWalker(doc.body, NodeFilter.SHOW_ELEMENT, node =>
+  const walker = doc.createTreeWalker(doc.body, NodeFilter.SHOW_ALL, node =>
     node.parentNode && isLink(node.parentNode) ? NodeFilter.FILTER_REJECT : NodeFilter.FILTER_ACCEPT,
   )
 
@@ -64,15 +64,17 @@ function convertToMarkdown(plaintext: string, walker: TreeWalker): string {
     index++
     const text = isLink(currentNode)
       ? (currentNode.textContent || '').replace(/[\t\n\r ]+/g, ' ')
-      : (currentNode.firstChild as Text)?.wholeText || ''
+      : (currentNode as Text)?.wholeText || ''
+
+    // No need to transform whitespace
+    if (isEmptyString(text)) {
+      currentNode = walker.nextNode()
+      continue
+    }
 
     // update value of markdownIgnoreBeforeIndex with current index if the current node is not a link
     if (!isLink(currentNode)) {
       markdownIgnoreBeforeIndex += text.replace(/[\t\n\r ]+/g, ' ').trimStart().length
-    }
-
-    // No need to transform whitespace
-    if (isEmptyString(text)) {
       currentNode = walker.nextNode()
       continue
     }
@@ -81,14 +83,11 @@ function convertToMarkdown(plaintext: string, walker: TreeWalker): string {
     const markdownFoundIndex = markdown.indexOf(text, markdownIgnoreBeforeIndex)
 
     if (markdownFoundIndex >= 0) {
-      if (isLink(currentNode)) {
-        const markdownLink = linkify(currentNode, text)
-        // Transform 'example link plus more text' into 'example [link](example link) plus more text'
-        // Method: 'example [link](example link) plus more text' = 'example ' + '[link](example link)' + ' plus more text'
-        markdown =
-          markdown.slice(0, markdownFoundIndex) + markdownLink + markdown.slice(markdownFoundIndex + text.length)
-        markdownIgnoreBeforeIndex = markdownFoundIndex + markdownLink.length
-      }
+      const markdownLink = linkify(currentNode, text)
+      // Transform 'example link plus more text' into 'example [link](example link) plus more text'
+      // Method: 'example [link](example link) plus more text' = 'example ' + '[link](example link)' + ' plus more text'
+      markdown = markdown.slice(0, markdownFoundIndex) + markdownLink + markdown.slice(markdownFoundIndex + text.length)
+      markdownIgnoreBeforeIndex = markdownFoundIndex + markdownLink.length
     }
 
     currentNode = walker.nextNode()

Diff for: test/test.js

@@ -443,6 +443,45 @@ describe('paste-markdown', function () {
       paste(textarea, {'text/html': sentence, 'text/plain': plaintextSentence})
       assert.equal(textarea.value, markdownSentence)
     })
+
+    it('finds the right link when identical labels are present chrome and edge', function () {
+      // eslint-disable-next-line github/unescaped-html-literal
+      const sentence = `<meta charset='utf-8'>
+      <p>foo bar<br>bar<span> </span><a href="https://www.abcxyz.org/">bar</a></p><p>bar<span> </span><a href="https://www.abcxyz.com/">bar</a><span> </span>foo</p>`
+
+      const markdownSentence = `foo bar
+bar [bar](https://www.abcxyz.org/)
+
+bar [bar](https://www.abcxyz.com/) foo`
+
+      const plaintextSentence = `foo bar
+bar bar
+
+bar bar foo`
+
+      paste(textarea, {'text/html': sentence, 'text/plain': plaintextSentence})
+      assert.equal(textarea.value, markdownSentence)
+    })
+
+    it('finds the right link when identical labels are present firefox', function () {
+      // eslint-disable-next-line github/unescaped-html-literal
+      const sentence = `<meta charset='utf-8'>
+      <div><div><p>foo bar<br>
+      bar <a href="https://www.abcxyz.org/">bar</a></p>
+      <p>bar <a href="https://www.abcxyz.com/">bar</a> foo</p></div></div>`
+      const markdownSentence = `foo bar
+bar [bar](https://www.abcxyz.org/)
+
+bar [bar](https://www.abcxyz.com/) foo`
+
+      const plaintextSentence = `foo bar
+bar bar
+
+bar bar foo`
+
+      paste(textarea, {'text/html': sentence, 'text/plain': plaintextSentence})
+      assert.equal(textarea.value, markdownSentence)
+    })
   })
 })