@adobe/helix-md2docx 2.1.5 → 2.1.6

package/CHANGELOG.md

@@ -1,3 +1,10 @@
+## [2.1.6](https://github.com/adobe/helix-md2docx/compare/v2.1.5...v2.1.6) (2023-06-26)
+
+
+### Bug Fixes
+
+* error with newlines and <u>, <sub>, <sup> ([52821fa](https://github.com/adobe/helix-md2docx/commit/52821fa1d1999eaa81a8114127452d118a23ad4f)), closes [#276](https://github.com/adobe/helix-md2docx/issues/276)
+
 ## [2.1.5](https://github.com/adobe/helix-md2docx/compare/v2.1.4...v2.1.5) (2023-05-18)
 
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@adobe/helix-md2docx",
-  "version": "2.1.5",
+  "version": "2.1.6",
   "description": "Helix Service that converts markdown to word documents",
   "main": "src/index.js",
   "type": "module",
@@ -39,19 +39,19 @@
     "mime": "3.0.0",
     "rehype-parse": "8.0.4",
     "remark-gfm": "3.0.1",
-    "remark-parse": "10.0.1",
+    "remark-parse": "10.0.2",
     "unified": "10.1.2",
     "unist-util-visit": "4.1.2"
   },
   "devDependencies": {
     "@adobe/eslint-config-helix": "2.0.2",
-    "@adobe/helix-mediahandler": "2.1.11",
+    "@adobe/helix-mediahandler": "2.2.2",
     "@semantic-release/changelog": "6.0.3",
     "@semantic-release/exec": "6.0.3",
     "@semantic-release/git": "10.0.1",
-    "c8": "7.13.0",
-    "dotenv": "16.0.3",
-    "eslint": "8.40.0",
+    "c8": "8.0.0",
+    "dotenv": "16.3.1",
+    "eslint": "8.43.0",
     "eslint-import-resolver-exports": "1.0.0-beta.5",
     "eslint-plugin-header": "3.1.1",
     "eslint-plugin-import": "2.27.5",
@@ -61,7 +61,7 @@
     "mocha": "10.2.0",
     "mocha-multi-reporters": "1.5.1",
     "nock": "13.3.1",
-    "semantic-release": "21.0.2",
+    "semantic-release": "21.0.5",
     "unist-util-inspect": "7.0.2",
     "yauzl": "2.10.0"
   },

package/src/mdast2docx/handlers/index.js

@@ -40,8 +40,8 @@ export default {
   emphasis: characterStyle('italics'),
   strong: characterStyle('bold'),
   underline: characterStyle('underline'),
-  subScript: characterStyle('subScript'),
-  superScript: characterStyle('superScript'),
+  subscript: characterStyle('subScript'),
+  superscript: characterStyle('superScript'),
   heading,
   html,
   image,

package/src/mdast2docx/mdast-sanitize-html.js

@@ -12,7 +12,7 @@
 import { visit } from 'unist-util-visit';
 import { unified } from 'unified';
 import parse from 'rehype-parse';
-import { toMdast } from 'hast-util-to-mdast';
+import { defaultHandlers, toMdast } from 'hast-util-to-mdast';
 // import inspect from 'unist-util-inspect';
 import tableHandler from './hast-table-handler.js';
 import tableCellHandler from './hast-table-cell-handler.js';
@@ -22,7 +22,32 @@ import tableCellHandler from './hast-table-cell-handler.js';
  * @param type
  */
 function formatHandler(type) {
-  return (state, { children }) => ({ type, children });
+  return (state, node) => {
+    const result = { type, children: state.all(node) };
+    state.patch(node, result);
+    return result;
+  };
+}
+
+/**
+ * removes paragraphs from the child nodes recursively.
+ * @param  node
+ */
+function unwrapParagraphs(node) {
+  if (!node.children) {
+    return node;
+  }
+  for (let idx = 0; idx < node.children.length; idx += 1) {
+    const child = node.children[idx];
+    if (child.type === 'paragraph') {
+      node.children.splice(idx, 1, ...child.children);
+      idx += child.children.length - 1;
+    } else {
+      // eslint-disable-next-line no-param-reassign
+      node.children[idx] = unwrapParagraphs(child);
+    }
+  }
+  return node;
 }
 
 /**
@@ -32,10 +57,27 @@ function formatHandler(type) {
 function mdHandler(mdasts) {
   return (state, node) => {
     const { idx } = node.properties;
-    return mdasts[idx];
+    return mdasts[+idx];
   };
 }
 
+function isPhrasingParent(node) {
+  return [
+    'paragraph',
+    'underline',
+    'subscript',
+    'superscript',
+    'heading',
+    'emphasis',
+    'strong',
+    'link',
+    'linkReference',
+    'tableCell',
+    'delete',
+    'footnote',
+  ].includes(node.type);
+}
+
 /**
  * Sanitizes html:
  * - collapses consecutive html content (simply concat all nodes until the last html sibling)
@@ -72,12 +114,16 @@ export default function sanitizeHtml(tree) {
           if (n.type === 'html' || n.type === 'text') {
             html += n.value;
           } else {
-            html += `<markdown idx="${mdInserts.length}"></markdown>`;
+            html += `<markdown idx="${mdInserts.length}">foo</markdown>`;
+            mdInserts.push(n);
           }
-          mdInserts.push(n);
         });
       }
 
+      if (isPhrasingParent(parent)) {
+        html = `<p>${html}</p>`;
+      }
+
       // try parse html
       const hast = unified()
         .use(parse, { fragment: true })
@@ -85,22 +131,46 @@ export default function sanitizeHtml(tree) {
 
       // convert to mdast with extra handlers
       const mdast = toMdast(hast, {
+        document: false,
         handlers: {
+          ...defaultHandlers,
           u: formatHandler('underline'),
-          sub: formatHandler('subScript'),
-          sup: formatHandler('superScript'),
+          sub: formatHandler('subscript'),
+          sup: formatHandler('superscript'),
           table: tableHandler,
           markdown: mdHandler(mdInserts),
           th: tableCellHandler,
           td: tableCellHandler,
         },
       });
+      // clear inserts
+      mdInserts.length = 0;
 
-      // console.log('************************************');
-      // // console.log('>>>>', html);
-      // process.stdout.write(inspect(hast));
-      // process.stdout.write('\n');
-      // console.log('************************************');
+      // ensure that flow nodes are in phrasing context
+      if (!isPhrasingParent(parent)) {
+        let lastParagraph;
+        for (let idx = 0; idx < mdast.children.length; idx += 1) {
+          const child = mdast.children[idx];
+          if (child.type === 'underline' || child.type === 'subscript' || child.type === 'superscript') {
+            unwrapParagraphs(child);
+            if (!lastParagraph) {
+              lastParagraph = {
+                type: 'paragraph',
+                children: [child],
+              };
+              mdast.children.splice(idx, 1, lastParagraph);
+            } else {
+              lastParagraph.children.push(child);
+              mdast.children.splice(idx, 1);
+              idx -= 1;
+            }
+          } else {
+            lastParagraph = null;
+          }
+        }
+      } else {
+        unwrapParagraphs(mdast);
+      }
 
       // inject children of parsed tree
       siblings.splice(index, 1, ...mdast.children);