OpenParable
Skip to content
← Back to skills

Extract Content With Image

Documents
extract_content_with_image

Extract Content With Image is an Agent Skill for documents published in the landingbj/LinkMind repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

将本地 PDF、TXT、Word、PPT 文件分割为文本和图片chunk。

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/landingbj/LinkMind/blob/HEAD/lagi-web/src/main/resources/skills/extract_content_with_image/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/extract-content-with-image, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

extract_content_with_image

使用方式

执行脚本前请先激活环境

  • 脚本入口：scripts/extract_content_with_image.py
  • 参数：argv[1] 为待处理文件的本地绝对路径
  • 支持输入：.pdf、.txt、.doc、.docx、.ppt、.pptx

输出格式（stdout）

  • 仅输出一个 JSON 对象，便于 Java 侧直接 json.loads / parseJsonObject
  • 成功：{"status":"success","filepath":"...pdf","data":[...]}
  • 失败：{"status":"failed","msg":"<具体异常信息>"}
  • data 中每个元素形如：{"text":"...", "image":"<图片列表的 JSON 字符串或空串>"}

运行依赖

  • 必需：PyMuPDF（fitz）和 Pillow
  • 可选：soffice，用于 .doc/.docx/.ppt/.pptx/.txt 转 PDF；可用环境变量 SOFFICE_PATH 指定
  • 无 soffice 时：
    • .txt 会直接用 fitz 生成 PDF，并基于原始文本做分块
    • .doc/.docx/.ppt/.pptx 会返回失败 JSON
  • 可选：transformers + TOKENIZER_DIR 或 MODEL_DIR
    • 配置后按 tokenizer 的 token 数分块，和 VicunaIndex 更接近
    • 未配置时按字符数分块，默认 CHUNK_SIZE=512

行为说明

  • 脚本会先把输入文件复制到 SKILL_OUTPUT_DIR/extract_content_with_image/<run_id>/files/
  • 若发生格式转换，filepath 返回转换后的 PDF 路径
  • 图片会裁剪到同一个运行目录下，并在 image 字段中以绝对路径返回
  • page_dir 会在结束时清理，保留裁剪后的图片文件