fix(super-converter): tolerate empty style properties on table styles (#3861)

packages/super-editor/src/editors/v1/core/super-converter/docx-helpers/index.js

@@ -2,3 +2,4 @@
 export * from './docx-helpers.js';
 export * from './docx-constants.js';
 export * from './get-default-style-definition.js';
+export * from './xml-node-access.js';

packages/super-editor/src/editors/v1/core/super-converter/docx-helpers/xml-node-access.js

@@ -0,0 +1,54 @@
+// @ts-check
+
+/**
+ * Safe reads over xml-js element trees.
+ *
+ * AIDEV-NOTE: xml-js omits `elements` and `attributes` entirely rather than
+ * emitting empty ones, so `<w:rPr/>`, `<w:rPr />` and `<w:rPr></w:rPr>` all
+ * parse to `{ type: 'element', name: 'w:rPr' }` with no `elements` key. Every
+ * child of `CT_Style` is optional in ECMA-376, so empty property containers are
+ * schema-valid, but readers that assume `node.elements` exists throw on them
+ * (issue #3861). Reach into parsed OOXML through these helpers rather than
+ * dereferencing `.elements` or `.attributes` directly.
+ */
+
+/**
+ * A node in a parsed xml-js tree.
+ * @typedef {object} XmlNode
+ * @property {string} [type] Node kind, e.g. `element` or `text`.
+ * @property {string} [name] Qualified element name, e.g. `w:rPr`.
+ * @property {Record<string, string>} [attributes] Absent when the element has no attributes.
+ * @property {XmlNode[]} [elements] Absent when the element has no children.
+ * @property {string} [text] Character data, for text nodes.
+ */
+
+/**
+ * Child elements of a parsed OOXML node, or an empty array when it has none.
+ * @param {XmlNode | null | undefined} node
+ * @returns {XmlNode[]}
+ */
+export const childElements = (node) => (Array.isArray(node?.elements) ? node.elements : []);
+
+/**
+ * First child element with the given qualified name.
+ * @param {XmlNode | null | undefined} node
+ * @param {string} name Qualified element name, e.g. `w:rPr`.
+ * @returns {XmlNode | undefined}
+ */
+export const findChild = (node, name) => childElements(node).find((el) => el?.name === name);
+
+/**
+ * Every child element with the given qualified name.
+ * @param {XmlNode | null | undefined} node
+ * @param {string} name Qualified element name, e.g. `w:tblStylePr`.
+ * @returns {XmlNode[]}
+ */
+export const findChildren = (node, name) => childElements(node).filter((el) => el?.name === name);
+
+/**
+ * Attribute value, or undefined when the element carries no attributes.
+ * @param {XmlNode | null | undefined} node
+ * @param {string} name Qualified attribute name, e.g. `w:val`.
+ * @returns {string | undefined}
+ */
+export const attrValue = (node, name) => node?.attributes?.[name];

packages/super-editor/src/editors/v1/core/super-converter/v2/importer/docxImporter.empty-style-properties.test.js

@@ -0,0 +1,84 @@
+// @ts-check
+import { describe, expect, it } from 'vitest';
+import * as xmljs from 'xml-js';
+import { createDocumentJson } from './docxImporter.js';
+
+/**
+ * End-to-end import cover for empty OOXML property containers (issue #3861).
+ *
+ * AIDEV-NOTE: Fixtures are parsed from XML strings because xml-js omits the
+ * `elements` key on empty elements. Object literals using `elements: []` do not
+ * reproduce the failing shape.
+ */
+
+const WORDPROCESSING_NS = 'http://schemas.openxmlformats.org/wordprocessingml/2006/main';
+const RELATIONSHIPS_NS = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships';
+
+const parse = (xml) => xmljs.xml2js(xml, { compact: false });
+
+const DOCUMENT_XML = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<w:document xmlns:w="${WORDPROCESSING_NS}" xmlns:r="${RELATIONSHIPS_NS}">
+  <w:body>
+    <w:p><w:r><w:t>before</w:t></w:r></w:p>
+    <w:tbl>
+      <w:tblPr><w:tblStyle w:val="TableGrid"/></w:tblPr>
+      <w:tblGrid><w:gridCol w:w="2880"/></w:tblGrid>
+      <w:tr><w:tc><w:p><w:r><w:t>cell</w:t></w:r></w:p></w:tc></w:tr>
+    </w:tbl>
+    <w:p><w:r><w:t>after</w:t></w:r></w:p>
+    <w:sectPr><w:pgSz w:w="12240" w:h="15840"/></w:sectPr>
+  </w:body>
+</w:document>`;
+
+/** Styles.xml where Table Grid inherits from a bare base and carries an empty w:rPr. */
+const STYLES_XML = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<w:styles xmlns:w="${WORDPROCESSING_NS}">
+  <w:style w:type="paragraph" w:default="1" w:styleId="Normal"><w:name w:val="Normal"/></w:style>
+  <w:style w:type="table" w:styleId="TableNormal"/>
+  <w:style w:type="table" w:styleId="TableGrid">
+    <w:name w:val="Table Grid"/>
+    <w:basedOn w:val="TableNormal"/>
+    <w:rPr/>
+  </w:style>
+</w:styles>`;
+
+const buildDocx = () => ({
+  'word/document.xml': parse(DOCUMENT_XML),
+  'word/styles.xml': parse(STYLES_XML),
+  'word/_rels/document.xml.rels': parse(
+    `<?xml version="1.0"?><Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships"/>`,
+  ),
+});
+
+const buildConverter = () => ({
+  headerIds: {},
+  headers: {},
+  footers: {},
+  docHiglightColors: new Set(),
+  trackedChangesOptions: {},
+  convertedXml: {},
+});
+
+const textOf = (node) => {
+  if (!node) return '';
+  if (node.type === 'text') return node.text ?? '';
+  return (node.content ?? []).map(textOf).join('');
+};
+
+describe('DOCX import with empty style property containers', () => {
+  it('keeps a table whose style has an empty w:rPr and a bare basedOn target', () => {
+    const exceptions = [];
+    const editor = {
+      emit: (name, payload) => {
+        if (name === 'exception') exceptions.push(payload.error);
+      },
+      options: {},
+    };
+
+    const result = createDocumentJson(buildDocx(), buildConverter(), editor);
+
+    const body = (result?.pmDoc?.content ?? []).map((node) => `${node.type}:${textOf(node)}`);
+    expect(body).toEqual(['paragraph:before', 'table:cell', 'paragraph:after']);
+    expect(exceptions).toEqual([]);
+  });
+});

packages/super-editor/src/editors/v1/core/super-converter/v3/handlers/w/tbl/tbl-translator.empty-style-properties.test.js

@@ -0,0 +1,175 @@
+// @ts-check
+import { describe, expect, it } from 'vitest';
+import * as xmljs from 'xml-js';
+import { _getReferencedTableStyles } from './tbl-translator.js';
+
+/**
+ * Empty property containers in `word/styles.xml` used to abort table import (issue #3861).
+ *
+ * AIDEV-NOTE: These fixtures parse real XML strings on purpose. xml-js omits the
+ * `elements` key for empty elements, and hand-authored nodes using `elements: []`
+ * do NOT reproduce that shape, so object-literal fixtures pass against the bug.
+ * Keep building these from XML.
+ */
+
+const WORDPROCESSING_NS = 'http://schemas.openxmlformats.org/wordprocessingml/2006/main';
+
+/**
+ * Build a `docx` bundle whose Table Grid style carries the given child XML.
+ * @param {string} tableGridChildren Raw XML placed inside `<w:style w:styleId="TableGrid">`.
+ * @param {string} [extraStyles] Raw XML for sibling `<w:style>` elements.
+ */
+const docxWithTableGrid = (tableGridChildren, extraStyles = '') => ({
+  'word/styles.xml': xmljs.xml2js(
+    `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+     <w:styles xmlns:w="${WORDPROCESSING_NS}">
+       <w:style w:type="table" w:styleId="TableNormal"><w:name w:val="Normal Table"/></w:style>
+       ${extraStyles}
+       <w:style w:type="table" w:styleId="TableGrid">
+         <w:name w:val="Table Grid"/>
+         ${tableGridChildren}
+       </w:style>
+     </w:styles>`,
+    { compact: false },
+  ),
+});
+
+const resolve = (docx, styleId = 'TableGrid') => _getReferencedTableStyles(styleId, { docx });
+
+describe('_getReferencedTableStyles with empty style properties', () => {
+  describe('empty run properties (the reported case)', () => {
+    it.each([
+      ['self-closing', '<w:rPr/>'],
+      ['self-closing with space', '<w:rPr />'],
+      ['paired', '<w:rPr></w:rPr>'],
+      ['whitespace only', '<w:rPr>\n  </w:rPr>'],
+    ])('treats a %s w:rPr as a no-op', (_label, rPr) => {
+      const styles = resolve(docxWithTableGrid(rPr));
+
+      expect(styles).not.toBeNull();
+      expect(styles?.fonts).toBeUndefined();
+      expect(styles?.fontSize).toBeUndefined();
+    });
+
+    it('still reads populated run properties', () => {
+      const styles = resolve(
+        docxWithTableGrid('<w:rPr><w:rFonts w:ascii="Arial" w:hAnsi="Arial" w:cs="Arial"/><w:sz w:val="24"/></w:rPr>'),
+      );
+
+      expect(styles?.fonts).toEqual({ ascii: 'Arial', hAnsi: 'Arial', cs: 'Arial' });
+      expect(styles?.fontSize).toBe('12pt');
+    });
+
+    it('treats a w:rFonts without attributes as a no-op', () => {
+      const styles = resolve(docxWithTableGrid('<w:rPr><w:rFonts/></w:rPr>'));
+
+      expect(styles?.fonts).toBeUndefined();
+    });
+  });
+
+  describe('empty paragraph properties', () => {
+    it.each([
+      ['self-closing', '<w:pPr/>'],
+      ['paired', '<w:pPr></w:pPr>'],
+    ])('treats a %s w:pPr as a no-op', (_label, pPr) => {
+      const styles = resolve(docxWithTableGrid(pPr));
+
+      expect(styles?.justification).toBeUndefined();
+    });
+
+    it('still reads justification from a populated w:pPr', () => {
+      const styles = resolve(docxWithTableGrid('<w:pPr><w:jc w:val="center"/></w:pPr>'));
+
+      expect(styles?.justification).toBe('center');
+    });
+  });
+
+  describe('inherited styles', () => {
+    it('tolerates a w:basedOn target that has no children', () => {
+      const docx = docxWithTableGrid(
+        '<w:basedOn w:val="BareBase"/><w:rPr/>',
+        '<w:style w:type="table" w:styleId="BareBase"/>',
+      );
+
+      expect(() => resolve(docx)).not.toThrow();
+    });
+
+    it('tolerates a w:basedOn target that does not exist', () => {
+      expect(() => resolve(docxWithTableGrid('<w:basedOn w:val="Missing"/>'))).not.toThrow();
+    });
+
+    it('still inherits table properties from a populated base style', () => {
+      const docx = docxWithTableGrid(
+        '<w:basedOn w:val="BorderedBase"/><w:tblPr><w:tblStyleRowBandSize w:val="1"/></w:tblPr>',
+        `<w:style w:type="table" w:styleId="BorderedBase">
+           <w:tblPr><w:tblBorders><w:top w:val="single" w:sz="4"/></w:tblBorders></w:tblPr>
+         </w:style>`,
+      );
+
+      expect(resolve(docx)?.borders?.top).toMatchObject({ val: 'single' });
+    });
+  });
+
+  describe('conditional formatting', () => {
+    it('skips a w:tblStylePr that has no w:type to key it by', () => {
+      const styles = resolve(docxWithTableGrid('<w:tblStylePr><w:tcPr/></w:tblStylePr>'));
+
+      expect(styles).not.toBeNull();
+      expect(Object.keys(styles ?? {})).not.toContain('undefined');
+    });
+
+    it('still reads a typed w:tblStylePr', () => {
+      const styles = resolve(docxWithTableGrid('<w:tblStylePr w:type="firstRow"><w:rPr><w:b/></w:rPr></w:tblStylePr>'));
+
+      expect(styles?.firstRow).toBeDefined();
+    });
+
+    it('treats empty conditional property containers as no-ops', () => {
+      const styles = resolve(
+        docxWithTableGrid('<w:tblStylePr w:type="firstRow"><w:tcPr/><w:rPr/><w:pPr/><w:trPr/></w:tblStylePr>'),
+      );
+
+      expect(styles?.firstRow).toBeDefined();
+    });
+  });
+
+  describe('malformed style records', () => {
+    it('tolerates a matched style element with no children', () => {
+      const docx = {
+        'word/styles.xml': xmljs.xml2js(
+          `<?xml version="1.0"?><w:styles xmlns:w="${WORDPROCESSING_NS}">
+             <w:style w:type="table" w:styleId="TableGrid"/>
+           </w:styles>`,
+          { compact: false },
+        ),
+      };
+
+      expect(() => resolve(docx)).not.toThrow();
+    });
+
+    it('ignores style elements that carry no attributes', () => {
+      const docx = {
+        'word/styles.xml': xmljs.xml2js(
+          `<?xml version="1.0"?><w:styles xmlns:w="${WORDPROCESSING_NS}">
+             <w:style><w:name w:val="Orphan"/></w:style>
+             <w:style w:type="table" w:styleId="TableGrid"><w:name w:val="Table Grid"/></w:style>
+           </w:styles>`,
+          { compact: false },
+        ),
+      };
+
+      expect(resolve(docx)?.name).toBeDefined();
+    });
+
+    it('returns null when styles.xml is absent', () => {
+      expect(resolve({})).toBeNull();
+    });
+
+    it('treats empty w:tblPr, w:tcPr and w:trPr as no-ops', () => {
+      const styles = resolve(docxWithTableGrid('<w:tblPr/><w:tcPr/><w:trPr/>'));
+
+      expect(styles?.borders).toBeUndefined();
+      expect(styles?.cellMargins).toBeUndefined();
+    });
+  });
+});