The fallback memory extractor (used by routes/memory_routes.py when the LLM
extractor fails) matched list items with `r'^[-*•]|\d+\.\s*(.*)'`. Operator
precedence makes that `(^[-*•]) | (\d+\.\s*(.*))`, so the capture group only
exists on the numbered-list branch.
A bullet line ("- foo") matches the first branch, so `group(1)` is None and
`text_match.group(1).strip()` raises AttributeError — crashing extraction for
any assistant message that contains a bullet list (i.e. most of them). Numbered
lists happened to work.
Group both markers — `r'^(?:[-*•]|\d+\.)\s*(.*)'` — so the capture applies to
bullets and numbers alike.
Adds tests/test_memory_bullet_extraction.py (red before, green after).
27 lines
1.2 KiB
Python
27 lines
1.2 KiB
Python
"""Regression test: extract_memory_from_chat must not crash on bullet lines.
|
|
|
|
The fallback memory extractor (invoked by routes/memory_routes.py when the LLM
|
|
extractor fails) matched list items with ``r'^[-*•]|\\d+\\.\\s*(.*)'``. Because
|
|
of alternation precedence that pattern is ``(^[-*•]) | (\\d+\\.\\s*(.*))`` — the
|
|
capture group lives only in the numbered-list branch. A bullet line ("- ...")
|
|
matches the first branch, so ``group(1)`` is ``None`` and ``.strip()`` raised
|
|
``AttributeError``, crashing extraction for any assistant message that contains
|
|
a bullet list (the dominant case).
|
|
"""
|
|
from src.memory import MemoryManager
|
|
|
|
|
|
def test_extract_memory_from_chat_handles_bullets(tmp_path):
|
|
mgr = MemoryManager(str(tmp_path))
|
|
chat = [{
|
|
"role": "assistant",
|
|
"content": "- User likes coffee\n* Prefers tea in winter\n1. Wakes at 6am",
|
|
}]
|
|
|
|
out = mgr.extract_memory_from_chat(chat)
|
|
texts = [m["text"] for m in out]
|
|
|
|
assert "User likes coffee" in texts # '-' bullet (used to crash)
|
|
assert "Prefers tea in winter" in texts # '*' bullet (used to crash)
|
|
assert "Wakes at 6am" in texts # numbered list (already worked)
|