fix(mem_os): split references on mixed comma styles

[simpleqt]

Description

Fixes #2417

split_continuous_references turned [1:aaa, 4:bbb] into [1:aaa][4:bbb] via two sequential str.replace passes — one for ", ", one for ",". The first matching pass rewrote the content substring, so the second pass could no longer find the original: a tag mixing both styles kept its earlier references merged, and the streaming chat path (api/handlers/chat_handler.py → process_streaming_references_complete) then rendered them as literal text instead of reference tags.

split_continuous_references("[1:92ff35fb,4:bfe6f044, 7:abcd1234]")
# before: '[1:92ff35fb,4:bfe6f044][7:abcd1234]'
# after:  '[1:92ff35fb][4:bfe6f044][7:abcd1234]'

Implementation: split the bracket content with re.sub(r",\\s*", "][", ...) in a single pass; the redundant inner import re in process_streaming_references_complete moved to module level. No dependency changes.

Type of change

• Bug fix (non-breaking change which fixes an issue)

How Has This Been Tested?

• Unit Test

New tests/mem_os/test_reference_utils.py covers spaced, tight, and mixed separators, surrounding-text preservation, and the non-reference early returns. Red/green verified: test_mixed_separators fails against unpatched main and passes with the fix; the other four pin unchanged behavior.

• Test Script Or Test Steps

poetry run pytest tests/mem_os/test_reference_utils.py -v

ruff check and ruff format --check pass on both touched files.

Checklist

• I have performed a self-review of my own code | 我已自行检查了自己的代码
• I have commented my code in hard-to-understand areas | 我已在难以理解的地方对代码进行了注释
• I have added tests that prove my fix is effective or that my feature works | 我已添加测试以证明我的修复有效或功能正常
• I have linked the issue to this PR (if applicable) | 我已将 issue 链接到此 PR（如果适用）

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[Memtensor-AI]

🤖 Open Code Review

Target: PR #2418
Task: 5d9195458007c54b
Base: main
Head: sq/ref-split-mixed-sep

✅ OpenCodeReview: Review complete: 0 finding(s) across 2 selected item(s).

Generated by cloud-assistant via Open Code Review.

[Memtensor-AI]

✅ Automated Test Results: PASSED

All tests passed (5/5 executed). memos_python_core/changed-repo-python: 5/5. Duration: 9s [advisory, non-gating] AI-generated tests on branch test/auto-gen-ff4aba06e30b05df-20260928112516: 96/96 passed — these do NOT affect the PR verdict; review the branch manually.

Branch: sq/ref-split-mixed-sep

[simpleqt]

Added a second commit (same theme): the split previously rewrote any bracketed comma list, so ordinary text like interval [x, y] was corrupted into interval [x][y] on the streaming chat path. The split now requires the reference-tag shape (a numeric id before a colon in every element), which is the documented input format. New test test_ordinary_bracketed_text_untouched is red/green verified.

[Memtensor-AI]

✅ Automated Test Results: PASSED

All tests passed (6/6 executed). memos_python_core/changed-repo-python: 6/6. Duration: 8s [advisory, non-gating] AI-generated tests on branch test/auto-gen-5d9195458007c54b-20260928145304: 84/84 passed — these do NOT affect the PR verdict; review the branch manually.

Branch: sq/ref-split-mixed-sep