Issues related to the minerU plugin

[ooodwbooo]

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:)
• Please do not modify this template :) and fill in all the required fields.

Content

When using the minerU plugin, small files are processed without any problems; however, large files fail if the processing time exceeds 300 seconds. Additionally, the relevant configuration options cannot be found in the env file. What should be done? Moreover, the task ID changes with each run, making it impossible to reuse the cache from a previous processing session.

pipeline

dependencies:
- current_identifier: null
  type: package
  value:
    plugin_unique_identifier: stvlynn/lmstudio:0.0.2@d581405702d3bee86a245c314a3b209531695a747839822add5ba23cbe626d89
    version: null
- current_identifier: null
  type: marketplace
  value:
    marketplace_plugin_unique_identifier: langgenius/general_chunker:0.0.7@a685cc66820d0471545499d2ff5c87ed7e51525470155dbc2f82e1114cd2a9d6
    version: null
- current_identifier: null
  type: marketplace
  value:
    marketplace_plugin_unique_identifier: langgenius/dify_extractor:0.0.5@ba7e2fd9165eda73bfcc68e31a108855197e88706e5556c058e0777ab08409b3
    version: null
- current_identifier: null
  type: marketplace
  value:
    marketplace_plugin_unique_identifier: langgenius/mineru:0.5.1@ccf545b476c0771707ea21122f9342ebe24d2a44ae040268665b41a9ca9f37fe
    version: null
kind: rag_pipeline
rag_pipeline:
  description: ''
  icon: 📙
  icon_background: '#FFF4ED'
  icon_type: emoji
  icon_url: null
  name: 第三方规格书
version: 0.1.0
workflow:
  conversation_variables: []
  environment_variables: []
  features: {}
  graph:
    edges:
    - data:
        isInIteration: false
        isInLoop: false
        sourceType: datasource
        targetType: if-else
      id: 1764134193043-source-1764211496722-target
      source: '1764134193043'
      sourceHandle: source
      target: '1764211496722'
      targetHandle: target
      type: custom
      zIndex: 0
    - data:
        isInLoop: false
        sourceType: if-else
        targetType: tool
      id: 1764211496722-true-1764134206101-target
      source: '1764211496722'
      sourceHandle: 'true'
      target: '1764134206101'
      targetHandle: target
      type: custom
      zIndex: 0
    - data:
        isInIteration: false
        isInLoop: false
        sourceType: if-else
        targetType: tool
      id: 1764211496722-false-1764211657552-target
      source: '1764211496722'
      sourceHandle: 'false'
      target: '1764211657552'
      targetHandle: target
      type: custom
      zIndex: 0
    - data:
        isInIteration: false
        isInLoop: false
        sourceType: variable-aggregator
        targetType: tool
      id: 1764211916124-source-1764134530811-target
      source: '1764211916124'
      sourceHandle: source
      target: '1764134530811'
      targetHandle: target
      type: custom
      zIndex: 0
    - data:
        isInLoop: false
        sourceType: tool
        targetType: knowledge-index
      id: 1764134530811-source-knowledgeBase-target
      source: '1764134530811'
      sourceHandle: source
      target: knowledgeBase
      targetHandle: target
      type: custom
      zIndex: 0
    - data:
        isInLoop: false
        sourceType: tool
        targetType: variable-aggregator
      id: 1764211657552-source-1764211916124-target
      selected: false
      source: '1764211657552'
      sourceHandle: source
      target: '1764211916124'
      targetHandle: target
      type: custom
      zIndex: 0
    - data:
        isInLoop: false
        sourceType: tool
        targetType: variable-aggregator
      id: 1764134206101-source-1764211916124-target
      source: '1764134206101'
      sourceHandle: source
      target: '1764211916124'
      targetHandle: target
      type: custom
      zIndex: 0
    nodes:
    - data:
        datasource_configurations: {}
        datasource_label: File
        datasource_name: upload-file
        datasource_parameters: {}
        fileExtensions:
        - txt
        - markdown
        - mdx
        - pdf
        - html
        - xlsx
        - xls
        - vtt
        - properties
        - doc
        - docx
        - csv
        - eml
        - msg
        - pptx
        - xml
        - epub
        - ppt
        - md
        plugin_id: langgenius/file
        provider_name: file
        provider_type: local_file
        selected: false
        title: File
        type: datasource
      height: 52
      id: '1764134193043'
      position:
        x: 280
        y: 282
      positionAbsolute:
        x: 280
        y: 282
      selected: false
      sourcePosition: right
      targetPosition: left
      type: custom
      width: 242
    - data:
        chunk_structure: text_model
        embedding_model: text-embedding-qwen3-embedding-8b
        embedding_model_provider: stvlynn/lmstudio/lmstudio
        index_chunk_variable_selector:
        - '1764134530811'
        - result
        indexing_technique: high_quality
        keyword_number: 10
        retrieval_model:
          reranking_enable: false
          reranking_mode: weighted_score
          reranking_model: null
          score_threshold: 0.5
          score_threshold_enabled: false
          search_method: hybrid_search
          top_k: 100
          weights:
            keyword_setting:
              keyword_weight: 0.3
            vector_setting:
              embedding_model_name: text-embedding-qwen3-embedding-8b
              embedding_provider_name: stvlynn/lmstudio/lmstudio
              vector_weight: 0.7
        selected: false
        title: 知识库
        type: knowledge-index
      height: 114
      id: knowledgeBase
      position:
        x: 2143.8063273885377
        y: 300.47311075082047
      positionAbsolute:
        x: 2143.8063273885377
        y: 300.47311075082047
      selected: false
      sourcePosition: right
      targetPosition: left
      type: custom
      width: 242
    - data:
        is_team_authorization: true
        paramSchemas:
        - auto_generate: null
          default: null
          form: llm
          human_description:
            en_US: The file to be parsed(support pdf, ppt, pptx, doc, docx, png, jpg,
              jpeg)
            ja_JP: 解析するファイル(pdf、ppt、pptx、doc、docx、png、jpg、jpegをサポート)
            pt_BR: The file to be parsed(support pdf, ppt, pptx, doc, docx, png, jpg,
              jpeg)
            zh_Hans: 用于解析的文件(支持 pdf, ppt, pptx, doc, docx, png, jpg, jpeg)
          label:
            en_US: file
            ja_JP: file
            pt_BR: file
            zh_Hans: file
          llm_description: The file to be parsed (support pdf, ppt, pptx, doc, docx,
            png, jpg, jpeg)
          max: null
          min: null
          name: file
          options: []
          placeholder: null
          precision: null
          required: true
          scope: null
          template: null
          type: file
        - auto_generate: null
          default: auto
          form: form
          human_description:
            en_US: (For local deployment v1 and v2) Parsing method, can be auto, ocr,
              or txt. Default is auto. If results are not satisfactory, try ocr
            ja_JP: （ローカルデプロイメントv1とv2用）解析方法は、auto、ocr、またはtxtのいずれかです。デフォルトはautoです。結果が満足できない場合は、ocrを試してください
            pt_BR: (For local deployment v1 and v2) Parsing method, can be auto, ocr,
              or txt. Default is auto. If results are not satisfactory, try ocr
            zh_Hans: （用于本地部署v1和v2版本）解析方法，可以是auto, ocr, 或 txt。默认是auto。如果结果不理想，请尝试ocr
          label:
            en_US: parse method
            ja_JP: 解析方法
            pt_BR: parse method
            zh_Hans: 解析方法
          llm_description: (For local deployment v1 and v2) Parsing method, can be
            auto, ocr, or txt. Default is auto. If results are not satisfactory, try
            ocr
          max: null
          min: null
          name: parse_method
          options:
          - icon: ''
            label:
              en_US: auto
              ja_JP: auto
              pt_BR: auto
              zh_Hans: auto
            value: auto
          - icon: ''
            label:
              en_US: ocr
              ja_JP: ocr
              pt_BR: ocr
              zh_Hans: ocr
            value: ocr
          - icon: ''
            label:
              en_US: txt
              ja_JP: txt
              pt_BR: txt
              zh_Hans: txt
            value: txt
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: select
        - auto_generate: null
          default: 1
          form: form
          human_description:
            en_US: (For official API and local deployment v2) Whether to enable formula
              recognition
            ja_JP: （公式API用とローカルデプロイメントv2用）数式認識を有効にするかどうか
            pt_BR: (For official API and local deployment v2) Whether to enable formula
              recognition
            zh_Hans: （用于官方API和本地部署v2版本）是否开启公式识别
          label:
            en_US: Enable formula recognition
            ja_JP: 数式認識を有効にする
            pt_BR: Enable formula recognition
            zh_Hans: 开启公式识别
          llm_description: (For official API and local deployment v2) Whether to enable
            formula recognition
          max: null
          min: null
          name: enable_formula
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: boolean
        - auto_generate: null
          default: 1
          form: form
          human_description:
            en_US: (For official API and local deployment v2) Whether to enable table
              recognition
            ja_JP: （公式API用とローカルデプロイメントv2用）表認識を有効にするかどうか
            pt_BR: (For official API and local deployment v2) Whether to enable table
              recognition
            zh_Hans: （用于官方API和本地部署v2版本）是否开启表格识别
          label:
            en_US: Enable table recognition
            ja_JP: 表認識を有効にする
            pt_BR: Enable table recognition
            zh_Hans: 开启表格识别
          llm_description: (For official API and local deployment v2) Whether to enable
            table recognition
          max: null
          min: null
          name: enable_table
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: boolean
        - auto_generate: null
          default: auto
          form: form
          human_description:
            en_US: '(For official API and local deployment v2) Specify document language,
              default ch, can be set to auto(local deployment need to specify the
              language, default ch), other optional value list see: https://paddlepaddle.github.io/PaddleOCR/latest/version3.x/pipeline_usage/OCR.html#5'
            ja_JP: （公式API用とローカルデプロイメントv2用）ドキュメント言語を指定します。デフォルトはchで、autoに設定できます。autoの場合（ローカルデプロイメントでは言語を指定する必要があります。デフォルトはchです）、モデルはドキュメント言語を自動的に識別します。他のオプション値リストについては、次を参照してください：https://paddlepaddle.github.io/PaddleOCR/latest/version3.x/pipeline_usage/OCR.html#5
            pt_BR: '(For official API and local deployment v2) Specify document language,
              default ch, can be set to auto(local deployment need to specify the
              language, default ch), other optional value list see: https://paddlepaddle.github.io/PaddleOCR/latest/version3.x/pipeline_usage/OCR.html#5'
            zh_Hans: （仅限官方api和本地部署v2版本）指定文档语言，默认 ch，可以设置为auto，当为auto时模型会自动识别文档语言（本地部署需要指定明确的语言，默认ch），其他可选值列表详见：https://paddlepaddle.github.io/PaddleOCR/latest/version3.x/pipeline_usage/OCR.html#5
          label:
            en_US: Document language
            ja_JP: ドキュメント言語
            pt_BR: Document language
            zh_Hans: 文档语言
          llm_description: '(For official API and local deployment v2) Specify document
            language, default ch, can be set to auto(local deployment need to specify
            the language, default ch), other optional value list see: https://paddlepaddle.github.io/PaddleOCR/latest/version3.x/pipeline_usage/OCR.html#5'
          max: null
          min: null
          name: language
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: string
        - auto_generate: null
          default: 0
          form: form
          human_description:
            en_US: (For official API) Whether to enable OCR recognition
            ja_JP: （公式API用）OCR認識を有効にするかどうか
            pt_BR: (For official API) Whether to enable OCR recognition
            zh_Hans: （用于官方API）是否开启OCR识别
          label:
            en_US: Enable OCR recognition
            ja_JP: OCR認識を有効にする
            pt_BR: Enable OCR recognition
            zh_Hans: 开启OCR识别
          llm_description: (For official API) Whether to enable OCR recognition
          max: null
          min: null
          name: enable_ocr
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: boolean
        - auto_generate: null
          default: '[]'
          form: form
          human_description:
            en_US: '(For official API) Example: ["docx","html"], markdown, json are
              the default export formats, no need to set, this parameter only supports
              one or more of docx, html, latex'
            ja_JP: （公式API用）例：["docx","html"]、markdown、jsonはデフォルトのエクスポート形式であり、設定する必要はありません。このパラメータは、docx、html、latexの3つの形式のいずれかまたは複数のみをサポートします
            pt_BR: '(For official API) Example: ["docx","html"], markdown, json are
              the default export formats, no need to set, this parameter only supports
              one or more of docx, html, latex'
            zh_Hans: （用于官方API）示例：["docx","html"],markdown、json为默认导出格式，无须设置，该参数仅支持docx、html、latex三种格式中的一个或多个
          label:
            en_US: Extra export formats
            ja_JP: 追加のエクスポート形式
            pt_BR: Extra export formats
            zh_Hans: 额外导出格式
          llm_description: '(For official API) Example: ["docx","html"], markdown,
            json are the default export formats, no need to set, this parameter only
            supports one or more of docx, html, latex'
          max: null
          min: null
          name: extra_formats
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: string
        - auto_generate: null
          default: pipeline
          form: form
          human_description:
            en_US: '(For local deployment v2 or v2.5) Example: pipeline, vlm-transformers,
              vlm-sglang-engine, vlm-sglang-client, default is pipeline'
            ja_JP: （ローカルデプロイメント v2/v2.5 用）例：pipeline、vlm-transformers、vlm-sglang-engine、vlm-sglang-client、デフォルトはpipeline
            pt_BR: '(For local deployment v2 or v2.5) Example: pipeline, vlm-transformers,
              vlm-sglang-engine, vlm-sglang-client, default is pipeline'
            zh_Hans: （用于本地部署 v2 或 v2.5 版本）示例：pipeline、vlm-transformers、vlm-sglang-engine、vlm-sglang-client，默认值为pipeline
          label:
            en_US: Backend type
            ja_JP: バックエンドタイプ
            pt_BR: Backend type
            zh_Hans: 解析后端
          llm_description: '(For local deployment v2/v2.5) Example: pipeline, vlm-transformers,
            vlm-sglang-engine, vlm-sglang-client, default is pipeline'
          max: null
          min: null
          name: backend
          options:
          - icon: ''
            label:
              en_US: pipeline
              ja_JP: pipeline
              pt_BR: pipeline
              zh_Hans: pipeline
            value: pipeline
          - icon: ''
            label:
              en_US: vlm-transformers
              ja_JP: vlm-transformers
              pt_BR: vlm-transformers
              zh_Hans: vlm-transformers
            value: vlm-transformers
          - icon: ''
            label:
              en_US: vlm-sglang-engine (v2.2)
              ja_JP: vlm-sglang-engine (v2.2)
              pt_BR: vlm-sglang-engine (v2.2)
              zh_Hans: vlm-sglang-engine (v2.2)
            value: vlm-sglang-engine
          - icon: ''
            label:
              en_US: vlm-sglang-client (v2.2)
              ja_JP: vlm-sglang-client (v2.2)
              pt_BR: vlm-sglang-client (v2.2)
              zh_Hans: vlm-sglang-client (v2.2)
            value: vlm-sglang-client
          - icon: ''
            label:
              en_US: vlm-http-client (v2.5)
              ja_JP: vlm-http-client (v2.5)
              pt_BR: vlm-http-client (v2.5)
              zh_Hans: vlm-http-client (v2.5)
            value: vlm-http-client
          - icon: ''
            label:
              en_US: vlm-vllm-async-engine (v2.5)
              ja_JP: vlm-vllm-async-engine (v2.5)
              pt_BR: vlm-vllm-async-engine (v2.5)
              zh_Hans: vlm-vllm-async-engine (v2.5)
            value: vlm-vllm-async-engine
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: select
        - auto_generate: null
          default: ''
          form: form
          human_description:
            en_US: '(For local deployment v2 when backend is vlm-sglang-client or
              vlm-http-client) Example: http://127.0.0.1:30000, default is empty'
            ja_JP: （ローカルデプロイメントv2用 解析後端が vlm-sglang-client 或 vlm-http-client の場合）例：http://127.0.0.1:30000、デフォルトは空
            pt_BR: '(For local deployment v2 when backend is vlm-sglang-client or
              vlm-http-client) Example: http://127.0.0.1:30000, default is empty'
            zh_Hans: （用于本地部署v2版本 解析后端为 vlm-sglang-client 或 vlm-http-client 时）示例：http://127.0.0.1:30000，默认值为空
          label:
            en_US: server url
            ja_JP: serverアドレス
            pt_BR: server url
            zh_Hans: server地址
          llm_description: '(For local deployment v2 when backend is vlm-sglang-client
            or vlm-http-client) Example: http://127.0.0.1:30000, default is empty'
          max: null
          min: null
          name: server_url
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: string
        params:
          backend: ''
          enable_formula: ''
          enable_ocr: ''
          enable_table: ''
          extra_formats: ''
          file: ''
          language: ''
          parse_method: ''
          server_url: ''
        provider_id: langgenius/mineru/mineru
        provider_name: langgenius/mineru/mineru
        provider_type: builtin
        retry_config:
          max_retries: 2
          retry_enabled: false
          retry_interval: 1000
        selected: false
        title: Parse File
        tool_configurations:
          backend:
            type: constant
            value: pipeline
          enable_formula:
            type: constant
            value: false
          enable_ocr:
            type: constant
            value: 0
          enable_table:
            type: constant
            value: 1
          extra_formats:
            type: mixed
            value: '[]'
          language:
            type: mixed
            value: auto
          parse_method:
            type: constant
            value: auto
          server_url:
            type: mixed
            value: http://host.docker.internal:8009
        tool_description: 一个用于解析文本，表格和图片的工具，支持pdf,pptx,docx等多种格式。支持英语，中文等多种语言
        tool_label: Parse File
        tool_name: parse-file
        tool_node_version: '2'
        tool_parameters:
          file:
            type: variable
            value:
            - '1764134193043'
            - file
        type: tool
      height: 270
      id: '1764134206101'
      position:
        x: 882
        y: 188.31493833512462
      positionAbsolute:
        x: 882
        y: 188.31493833512462
      selected: true
      sourcePosition: right
      targetPosition: left
      type: custom
      width: 242
    - data:
        is_team_authorization: true
        paramSchemas:
        - auto_generate: null
          default: null
          form: llm
          human_description:
            en_US: The text you want to chunk.
            ja_JP: The text you want to chunk.
            pt_BR: The text you want to chunk.
            zh_Hans: 你想要分块的文本。
          label:
            en_US: Input Content
            ja_JP: Input Content
            pt_BR: Input Content
            zh_Hans: 输入变量
          llm_description: The text you want to chunk.
          max: null
          min: null
          name: input_variable
          options: []
          placeholder: null
          precision: null
          required: true
          scope: null
          template: null
          type: string
        - auto_generate: null
          default: null
          form: llm
          human_description:
            en_US: The delimiter of the chunks.
            ja_JP: The delimiter of the chunks.
            pt_BR: The delimiter of the chunks.
            zh_Hans: 块的分隔符。
          label:
            en_US: Delimiter
            ja_JP: Delimiter
            pt_BR: Delimiter
            zh_Hans: 分隔符
          llm_description: The delimiter of the chunks, the format of the delimiter
            must be a string.
          max: null
          min: null
          name: delimiter
          options: []
          placeholder: null
          precision: null
          required: true
          scope: null
          template: null
          type: string
        - auto_generate: null
          default: null
          form: llm
          human_description:
            en_US: The maximum chunk length.
            ja_JP: The maximum chunk length.
            pt_BR: The maximum chunk length.
            zh_Hans: 最大块的长度。
          label:
            en_US: Maximum Chunk Length
            ja_JP: Maximum Chunk Length
            pt_BR: Maximum Chunk Length
            zh_Hans: 最大块的长度
          llm_description: The maximum chunk length, the format of the chunk size
            must be an integer.
          max: null
          min: null
          name: max_chunk_length
          options: []
          placeholder: null
          precision: null
          required: true
          scope: null
          template: null
          type: number
        - auto_generate: null
          default: null
          form: llm
          human_description:
            en_US: The chunk overlap length.
            ja_JP: The chunk overlap length.
            pt_BR: The chunk overlap length.
            zh_Hans: 块的重叠长度。
          label:
            en_US: Chunk Overlap Length
            ja_JP: Chunk Overlap Length
            pt_BR: Chunk Overlap Length
            zh_Hans: 块的重叠长度
          llm_description: The chunk overlap length, the format of the chunk overlap
            length must be an integer.
          max: null
          min: null
          name: chunk_overlap_length
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: number
        - auto_generate: null
          default: null
          form: llm
          human_description:
            en_US: Replace consecutive spaces, newlines and tabs
            ja_JP: Replace consecutive spaces, newlines and tabs
            pt_BR: Replace consecutive spaces, newlines and tabs
            zh_Hans: 替换连续的空格、换行符和制表符
          label:
            en_US: Replace consecutive spaces, newlines and tabs
            ja_JP: Replace consecutive spaces, newlines and tabs
            pt_BR: Replace consecutive spaces, newlines and tabs
            zh_Hans: 替换连续的空格、换行符和制表符
          llm_description: Replace consecutive spaces, newlines and tabs, the format
            of the replace must be a boolean.
          max: null
          min: null
          name: replace_consecutive_spaces_newlines_tabs
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: boolean
        - auto_generate: null
          default: null
          form: llm
          human_description:
            en_US: Delete all URLs and email addresses
            ja_JP: Delete all URLs and email addresses
            pt_BR: Delete all URLs and email addresses
            zh_Hans: 删除所有URL和电子邮件地址
          label:
            en_US: Delete all URLs and email addresses
            ja_JP: Delete all URLs and email addresses
            pt_BR: Delete all URLs and email addresses
            zh_Hans: 删除所有URL和电子邮件地址
          llm_description: Delete all URLs and email addresses, the format of the
            delete must be a boolean.
          max: null
          min: null
          name: delete_all_urls_and_email_addresses
          options: []
          placeholder: null
          precision: null
          required: false
          scope: null
          template: null
          type: boolean
        params:
          chunk_overlap_length: ''
          delete_all_urls_and_email_addresses: ''
          delimiter: ''
          input_variable: ''
          max_chunk_length: ''
          replace_consecutive_spaces_newlines_tabs: ''
        provider_id: langgenius/general_chunker/general_chunker
        provider_name: langgenius/general_chunker/general_chunker
        provider_type: builtin
        selected: false
        title: 通用文本分块
        tool_configurations: {}
        tool_description: 一个用于通用文本分块模式的工具，检索和召回的块是相同的。
        tool_label: 通用文本分块
        tool_name: general_chunker
        tool_node_version: '2'
        tool_parameters:
          chunk_overlap_length:
            type: variable
            value:
            - rag
            - '1764134193043'
            - Overlap
          delete_all_urls_and_email_addresses:
            type: constant
            value: null
          delimiter:
            type: mixed
            value: '{{#rag.1764134193043.Delimiter#}}'
          input_variable:
            type: mixed
            value: '{{#1764211916124.output#}}'
          max_chunk_length:
            type: variable
            value:
            - rag
            - '1764134193043'
            - Maximum_Parent_Length
          replace_consecutive_spaces_newlines_tabs:
            type: constant
            value: null
        type: tool
      height: 52
      id: '1764134530811'
      position:
        x: 1761.377696197556
        y: 300.47311075082047
      positionAbsolute:
        x: 1761.377696197556
        y: 300.47311075082047
      selected: false
      sourcePosition: right
      targetPosition: left
      type: custom
      width: 242
    - data:
        cases:
        - case_id: 'true'
          conditions:
          - comparison_operator: contains
            id: 86eb2aa6-66c1-4200-a9bc-cb3e5cfff8e2
            value: pdf
            varType: string
            variable_selector:
            - '1764134193043'
            - file
            - extension
          id: 'true'
          logical_operator: and
        selected: false
        title: 条件分支
        type: if-else
      height: 124
      id: '1764211496722'
      position:
        x: 582
        y: 282
      positionAbsolute:
        x: 582
        y: 282
      selected: false
      sourcePosition: right
      targetPosition: left
      type: custom
      width: 242
    - data:
        is_team_authorization: true
        paramSchemas:
        - auto_generate: null
          default: null
          form: llm
          human_description:
            en_US: the file to be parsed (support pdf, xls, xlsx, md, markdown, mdx,
              htm, html, pptx, docx, csv, json, yaml, yml)
            ja_JP: the file to be parsed (support pdf, xls, xlsx, md, markdown, mdx,
              htm, html, pptx, docx, csv, json, yaml, yml)
            pt_BR: o arquivo a ser analisado (suporta pdf, xls, xlsx, md, markdown,
              mdx, htm, html, pptx, docx, csv, json, yaml, yml)
            zh_Hans: 用于解析的文件 (支持 pdf, xls, xlsx, md, markdown, mdx, htm, html, pptx,
              docx, csv, json, yaml, yml)
          label:
            en_US: file
            ja_JP: file
            pt_BR: file
            zh_Hans: file
          llm_description: the file to be parsed (support pdf, xls, xlsx, md, markdown,
            mdx, htm, html, pptx, docx, csv, json, yaml, yml)
          max: null
          min: null
          name: file
          options: []
          placeholder: null
          precision: null
          required: true
          scope: null
          template: null
          type: file
        params:
          file: ''
        plugin_id: langgenius/dify_extractor
        plugin_unique_identifier: langgenius/dify_extractor:0.0.5@ba7e2fd9165eda73bfcc68e31a108855197e88706e5556c058e0777ab08409b3
        provider_icon: /console/api/workspaces/current/plugin/icon?tenant_id=a8eb4f94-7bef-4513-84ac-b9be6c571d95&filename=59c233d363aeda0c93b8d682d458987c5b6a20f0ea6a85802cccf670cd2ddf5a.svg
        provider_id: langgenius/dify_extractor/dify_extractor
        provider_name: langgenius/dify_extractor/dify_extractor
        provider_type: builtin
        selected: false
        title: Dify文本提取器
        tool_configurations: {}
        tool_description: Dify Extractor
        tool_label: Dify文本提取器
        tool_name: dify_extractor
        tool_node_version: '2'
        tool_parameters:
          file:
            type: variable
            value:
            - '1764134193043'
            - file
        type: tool
      height: 52
      id: '1764211657552'
      position:
        x: 871.4439367138168
        y: 626.6267135908679
      positionAbsolute:
        x: 871.4439367138168
        y: 626.6267135908679
      selected: false
      sourcePosition: right
      targetPosition: left
      type: custom
      width: 242
    - data:
        output_type: string
        selected: false
        title: 变量聚合器
        type: variable-aggregator
        variables:
        - - '1764134206101'
          - text
        - - '1764211657552'
          - text
      height: 134
      id: '1764211916124'
      position:
        x: 1349.4992500768303
        y: 300.47311075082047
      positionAbsolute:
        x: 1349.4992500768303
        y: 300.47311075082047
      selected: false
      sourcePosition: right
      targetPosition: left
      type: custom
      width: 242
    viewport:
      x: -38.16221498601499
      y: 274.03799831030483
      zoom: 0.8705505632961267
  rag_pipeline_variables:
  - allow_file_extension: null
    allow_file_upload_methods: null
    allowed_file_types: null
    belong_to_node_id: '1764134193043'
    default_value: 1024
    label: 分隔长度
    max_length: 48
    options: []
    placeholder: null
    required: true
    tooltips: null
    type: number
    unit: null
    variable: Maximum_Parent_Length
  - allow_file_extension: null
    allow_file_upload_methods: null
    allowed_file_types: null
    belong_to_node_id: '1764134193043'
    default_value: '\n\n# '
    label: 分割符
    max_length: 48
    options: []
    placeholder: null
    required: true
    tooltips: null
    type: text-input
    unit: null
    variable: Delimiter
  - allow_file_extension: null
    allow_file_upload_methods: null
    allowed_file_types: null
    belong_to_node_id: '1764134193043'
    default_value: 50
    label: Overlap
    max_length: 48
    options: []
    placeholder: null
    required: true
    tooltips: null
    type: number
    unit: null
    variable: Overlap

[dosubot[bot]]

The minerU plugin fails to process large files that take longer than ~250–300 seconds because its polling timeout is hardcoded in the plugin code: it tries up to 50 times with a 5-second interval, then raises a timeout error if the result isn't ready. These timeout and retry values are not configurable via environment variables or pipeline parameters, which is why you can't find them in your .env file or pipeline config. See the implementation here.

Task IDs (batch_id) are generated by the backend parsing server for each run and always change, so cache reuse is not possible—the plugin doesn't control or reuse task IDs, nor does it implement any caching mechanism source.

Workarounds:

• To process larger files, you can modify the plugin code to increase MAX_RETRIES or reduce the sleep interval, but this requires a custom build.
• If your infrastructure (reverse proxy, Docker, etc.) enforces its own timeout (often at 300s), you may need to adjust those settings as well.
• Splitting large files into smaller chunks before processing can help avoid hitting the timeout.

There are currently no official configuration options or plugin updates that address these limitations. If this is a blocker, consider requesting a feature or discussing it with the maintainers. For reference, see the minerU plugin's README and recent PRs.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}