feat: add preprocessor system (Step 1 — Local Agent V2)

- app/core/preprocessors/__init__.py: detect_content_type + preprocess dispatcher
- app/core/preprocessors/base.py: PreprocessResult dataclass
- app/core/preprocessors/email_html.py: BeautifulSoup HTML stripping, metadata extraction, thread splitting
- requirements.txt: add beautifulsoup4 and lxml
- tests/test_preprocessors.py: 10 tests with Langfuse scoring (preprocess.* scores)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Roberto Musso
2026-04-07 10:19:02 +02:00
parent aa8bcbf0d8
commit a2d6d689e4
5 changed files with 463 additions and 0 deletions

View File

@@ -0,0 +1,25 @@
"""Base types for the preprocessor system."""
from __future__ import annotations
from dataclasses import dataclass, field
@dataclass
class PreprocessResult:
"""Output of a preprocessor handler.
Attributes
----------
content_type:
The detected content type (e.g. ``"email_html"``, ``"plain_text"``).
clean_text:
Human-readable text stripped of markup/binary noise.
metadata:
Dict of extracted metadata (keys vary by handler).
Common keys: ``subject``, ``from``, ``to``, ``date``, ``filename``.
"""
content_type: str
clean_text: str
metadata: dict = field(default_factory=dict)