hermes - 💡(How to fix) Fix [QQBot] 支持 PDF/Office 等附件内容提取 [1 participants]

TL;DR

Implement a document text extraction step using libraries like pymupdf, pdfminer, or python-docx after downloading attachments to enable LLM analysis.

Guidance

• Identify the file type of the downloaded attachment to determine the appropriate extraction library to use.
• Use existing functions like cache_document_from_bytes as a reference to integrate text extraction into the current workflow.
• Consider handling exceptions for unsupported file types or extraction failures to ensure the bot remains functional.
• Test the extraction process with various document types (PDF, Word, Excel) to ensure compatibility and accuracy.

Example

import pymupdf

def extract_text_from_pdf(file_path):
    pdf = pymupdf.PdfReader(file_path)
    text = ''
    for page in pdf.pages:
        text += page.extract_text()
    return text

Notes

The choice of extraction library may depend on the specific requirements and constraints of the project, such as performance, accuracy, and compatibility with different file formats.

Recommendation

Apply a workaround by integrating a text extraction library into the existing attachment handling code, as the issue implies a missing feature rather than a bug in an existing version.