Skip to content

Develop document preprocessing pipeline #136

Description

@vasyafomiuk

Description

Create a text preprocessing workflow that cleans and normalizes retrieved Confluence documents (remove markup, tokenize, strip code blocks where appropriate) prior to vectorization.

Technical Requirements

  • Handle Confluence storage format cleanup (HTML/XML)
  • Strip macros and code blocks
  • Tokenize text and normalize whitespace
  • Output documents ready for embedding generation
  • Unit tests for text edge cases

Background Context

Meeting Source: Coding Agent discussion on July 17, 2025
Project Context: coding-agent - Mentor Mode knowledge base

Acceptance Criteria

  • Function that transforms raw Confluence storage format to clean plain text
  • Coverage of at least 90% test cases with various Confluence macros
  • Integration with ingestion pipeline

Dependencies & Blockers

  • Confluence API integration completed

Additional Notes

Automatically generated from meeting discussion on 2025-07-17

Metadata

Metadata

Assignees

No one assigned

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions