Release 1.1.0

10 September 2026

This release focuses on clearer review, safer handling of missing content, and an offline workflow. It extracts editable content from PDFs with selectable text.

What changed

• Original PDF and extracted content can be compared page by page. On small screens, switch between the two views.
• Pages without selectable text are flagged. Exporting a partly readable document requires acknowledgement of the missing content.
• Clearer damaged-file and password guidance, extraction cancellation, and protection against results from a previous file.
• Explicit Word table column widths prevent collapsed cells in compatible document viewers.
• Compact file controls, optional conversion settings, accessible status messages and light/dark themes.

What was tested

The offline converter was opened directly from disk in Chrome 152 on macOS with browser networking disabled before opening it. The sample converted, its original page rendered, and a Word file downloaded without external requests.

Regression fixtures cover a report, an invoice, a short contract, Cyrillic text, an 80-page text document, mixed text/scan pages, image-only pages, corrupt files, password protection, cancellation and overlapping file changes. The sample Word file was also checked for real paragraph, heading and table structures.

Responsive layouts were checked at widths from 320 to 1280 pixels. These checks do not establish compatibility with every browser, mobile file-opening workflow or Word editor. Safari, Firefox and Edge have not been independently validated for this release.

External checks also used W3C’s tagged table example and two-column example. The five-row, five-column table was reconstructed. The two-column example was misclassified into tables, confirming that multi-column reading order remains unsupported. These independent fixtures are not included as product samples.

The generated sample DOCX was rendered in macOS Quick Look and visually checked after the table-width correction. This is not a Microsoft Word or cross-editor compatibility certification.

Limits that still matter

• No OCR: scanned text must be recognized in another PDF app first.
• No image or drawing export, clickable hyperlink preservation, or original font-size preservation.
• Lists keep their text, not automatic Word numbering.
• Complex tables, merged cells, columns and unusual writing systems can need substantial correction. A completed conversion does not guarantee correct reading order.
• Large-file capacity depends on document complexity and available browser memory.

Check every page and inspect the Word file before using it. The sample is a demonstration, not a promise of identical results on other PDFs.