Description

1. Refactor and improve the chunk decoding from llm backend into a separate sse.lua
2. Add lua_response_filter which runs lua_body_filter as defined by each plugin. We can't use body_filter directly because in case of ai response, the responses are managed directly by apisix, therefore this new phase is introduced.
3. Modifies ai-request-rewrite plugin to leverage lua_body_filter for writing response. This way is cleaner to read as all response modification for the ai plugins happens at one place lua_response_filter and each plugin just has a callback to take headers, body and return code,body.

• I have explained the need for this PR and the problem it solves
• I have explained the changes or the new features added to this PR
• I have added tests corresponding to this change
• I have updated the documentation to reflect this change
• I have verified that this change is backward compatible (If not, please discuss on the APISIX mailing list first)