refactor(parse): unify document parsing with AnyDoc

[qin-ctx]

Description

Unify the existing Office and EPUB parsing paths on AnyDoc, while keeping image extraction under OpenViking's ingestion policy. Replace the previous PDF/MinerU path with pdf-inspector-owned text and layout extraction plus pdfplumber-owned image extraction.

Human Involvement

• A human participated in the implementation or review loop
• This PR was generated entirely by AI agents without human participation in the loop

Related Issue

N/A

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Refactoring (no functional changes)
• Performance improvement
• Test update

Changes Made

• Route .doc, .docx, .pptx, .xls, .xlsx, .xlsm, and .epub through AnyDoc's public document model, preserving supported embedded-image positions and rendering PowerPoint speaker notes as ordered Speaker Notes sections in the same resource.
• Make pdf-inspector authoritative for PDF page text, headings, tables, reading order, and OCR diagnostics; retain pdfplumber only for per-page image extraction after page text.
• Remove the legacy format-specific parsers, MinerU integration, ExcelConfig, obsolete production dependencies, and outdated tests; share one OpenViking image validation policy across Markdown, AnyDoc, and PDF ingestion.

Testing

• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I have tested this on the following platforms:
  • Linux
  • macOS
  • Windows

Validated with:

• pytest -q -o addopts='' tests/parse tests/test_config_loader.py (494 passed, 31 skipped)
• Ruff, targeted mypy, compileall, uv lock --check, and git diff --check
• Existing-test audit confirming no new test files, test functions, or parameterized cases

Checklist

• My code follows the project's coding style
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• Any dependent changes have been merged and published

Screenshots (if applicable)

N/A

Additional Notes

• parsers.excel is removed and now rejected instead of being silently ignored.
• AnyDoc 0.1.8 does not expose embedded XLSX images through its document assets, so Excel currently guarantees worksheet Markdown only. No parallel openpyxl image extraction path is introduced.
• PDF pages requiring OCR produce explicit warnings; this change intentionally adds no OCR fallback.