mineru-4.0.6-released

neondatabase/neonmineru-4.0.6-releasedSep 22, 2026by myhloli

Upgrading?

AI Summary

Introduces EPUB online preview with chapter navigation and metadata. Fixes a regression where PDF text blocks were dropped and a crash when exporting PDFs with CJK paragraphs containing formulas.

Key Highlights

  • New EPUB viewer in WebUI with chapter navigation
  • Fixes regression for PDF text blocks missing line info
  • Fixes crash in PDF export for CJK paragraphs with inline formulas

Breaking Changes

  • Dependency update requires docvortex >= 0.4.22

New Features

  • EPUB online preview with section navigation
  • Improved metadata handling for EPUBs
  • Regression test for text block normalization

Full Release Notes

## 更新内容 / What's Changed

**新增 / Features**
- WebUI 新增 EPUB 在线预览:支持章节导航与元数据展示,静态资源经浏览器代理加载 / EPUB viewer in the WebUI with section navigation and improved metadata handling

**改进与修复 / Improvements & Fixes**
- 修复缺少行信息(lines)的 PDF 文本块在归一化时被整块丢弃的回归,增加 bbox 兜底(#5565)/ Fix regression where PDF text blocks missing line info were dropped entirely during normalization, with a bbox fallback (#5565)
- 修复中文段落含行内公式时导出 PDF 的崩溃(reportlab CJK 断行缺陷),并新增回归测试 / Fix crash when exporting PDFs with CJK paragraphs containing inline formulas (reportlab CJK line-breaking), with a regression test

**依赖 / Dependencies**
- docvortex 最低版本提升至 0.4.22 / Require docvortex>=0.4.22

**Full Changelog**: https://github.com/opendatalab/MinerU/compare/mineru-4.0.5-released...mineru-4.0.6-released

Related Repositories