refactor(rag): improve RAG manager implementation

[ErikBjare]

• Introduce shared RAG manager singleton via get_rag_manager()
• Add DEFAULT_COLLECTION constant for consistent collection naming
• Remove unused _clear_cache function
• Update gptme-rag dependency to 0.3.1

---

Important

Refactor RAG manager to use a shared singleton, add a default collection constant, remove an unused function, and update the gptme-rag dependency.

• RAG Manager:
  • Introduce shared singleton RAGManager via get_rag_manager() in _rag_context.py.
  • Replace rag_manager initialization in rag.py with get_rag_manager() calls.
• Constants:
  • Add DEFAULT_COLLECTION in _rag_context.py for consistent collection naming.
• Functions:
  • Remove unused _clear_cache() function in _rag_context.py.
• Dependencies:
  • Update gptme-rag dependency to version 0.3.1 in pyproject.toml.

^{This description was created by for edbf520. It will automatically update as commits are pushed.}

[ellipsis-dev]

❌ Changes requested. Reviewed everything up to 6d52f30 in 1 minute and 41 seconds

More details

• Looked at 137 lines of code in 3 files
• Skipped 1 files when reviewing.
• Skipped posting 0 drafted comments based on config settings.

Workflow ID: wflow_p2M0n9iq439lwJmQ

---

Want Ellipsis to fix these issues? Tag @ellipsis-dev in a comment. You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

Using a global variable for _rag_manager can lead to issues in multi-threaded environments or when the state needs to be reset. Consider using a class or context manager to handle the RAG manager instance.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 86.66667% with 2 lines in your changes missing coverage. Please review.

Project coverage is 72.52%. Comparing base (3f00be1) to head (edbf520).

✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
gptme/tools/rag.py	75.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #268      +/-   ##
==========================================
+ Coverage   72.36%   72.52%   +0.15%     
==========================================
  Files          64       64              
  Lines        4313     4316       +3     
==========================================
+ Hits         3121     3130       +9     
+ Misses       1192     1186       -6     

Flag	Coverage Δ
anthropic/claude-3-haiku-20240307	71.24% <86.66%> (+0.15%)	⬆️
openai/gpt-4o-mini	71.15% <86.66%> (+0.15%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

---

🚨 Try these New Features:

• JS Bundle Analysis - Avoid shipping oversized bundles

[ellipsis-dev]

❌ Changes requested. Incremental review on 46cf058 in 38 seconds

More details

• Looked at 46 lines of code in 2 files
• Skipped 0 files when reviewing.
• Skipped posting 0 drafted comments based on config settings.

Workflow ID: wflow_EcIhhVh9TNumyMlF

---

Want Ellipsis to fix these issues? Tag @ellipsis-dev in a comment. You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

Changing cloc tests/*.py to cloc tests might include non-Python files in the count, leading to inaccurate LoC statistics. Consider reverting to cloc tests/*.py.

Suggested change

	cloc tests --by-file
	cloc tests/*.py --by-file

[ellipsis-dev]

👍 Looks good to me! Incremental review on 0e9cb11 in 51 seconds

More details

• Looked at 137 lines of code in 3 files
• Skipped 1 files when reviewing.
• Skipped posting 0 drafted comments based on config settings.

Workflow ID: wflow_sJAu7mC5IPqELk8I

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

👍 Looks good to me! Incremental review on edbf520 in 24 seconds

More details

• Looked at 33 lines of code in 1 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. tests/test_tools_rag.py:10

• Draft comment:
  Combine multiple imports from the same module into a single line for better readability.

from gptme.tools.rag import _HAS_RAG, init as init_rag, rag_index, rag_search

• Reason this comment was not posted:
  Confidence changes required: 10%
  The import statements can be optimized for readability and maintainability.

Workflow ID: wflow_F4KAVc4T7bEKKLFC

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.