当文档中存在跨页的长json内容时，多个后端模式的效果不好，会被交错识别为文本和代码或者是算法描述

[tiansusama]

🔎 Search before asking | 提交之前请先搜索

• I have searched the MinerU Readme and found no similar bug report.
• I have searched the MinerU Issues and found no similar bug report.
• I have searched the MinerU Discussions and found no similar bug report.

🤖 Consult the online AI assistant for assistance | 在线 AI 助手咨询

• I have consulted the online AI assistant but was unable to obtain a solution to the issue.

Description of the bug | 错误描述

在解析PDF时，当文档中存在跨页的长json内容时，多个后端模式的效果不好，会被交错识别为文本和代码或者是算法描述

How to reproduce the bug | 如何复现

可以使用该文档进行复现

两种解析后端的表现都不好

CN121996801A.PDF

Operating System Mode | 操作系统类型

Windows

Operating System Version| 操作系统版本

Windows 11

Python version | Python 版本

3.12

Software version | 软件版本 (mineru --version)

>=2.5

Backend name | 解析后端

vlm

Device mode | 设备模式

cuda

[dosubot]

Hi @tiansusama! I'm Dosu and I'm helping the OpenDataLab team.

This is a known limitation in MinerU — code blocks and JSON content lack dedicated cross-page merging logic, unlike tables which have robust cross-page handling [1]. This directly explains the behavior you're seeing where long JSON spanning multiple pages gets alternately classified as text, code, or algorithm descriptions.

Root cause: The pipeline backend uses the Magika library for language detection. When it fails to confidently identify JSON as a programming language, it defaults to sub_type = "algorithm" [2]. Combined with the fact that cross-page text merging is heuristic-based and not designed for structured content like JSON [3], the content gets fragmented and inconsistently classified across pages.

For the VLM backend, the issue is compounded by context window contamination across pages and the loss of indentation/newlines during paragraph merging [4].

Possible workarounds:

1. Upgrade to version 2.6.0+ if you haven't already, as cross-page merging was improved [1]
2. Consider custom post-processing to detect and reassemble fragmented JSON blocks in the output — the built-in merging is limited for this content type
3. For this specific document type (patent PDFs with long JSON), you may need to implement downstream logic to identify and concatenate adjacent code/algorithm/text blocks that are actually part of the same JSON structure

Unfortunately there's no built-in configuration to fully resolve this issue currently, as dedicated cross-page code/JSON merging hasn't been implemented yet.

To reply, just mention @dosu.

---

Share context across your team and agents. Try Dosu.

[tiansusama]

即使是在同一页的内容也会被连续且交替的错误识别