Open Router support #7
Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Summary of Changes
Hello @meithon, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!
This pull request significantly expands the plugin's capabilities by adding support for OpenRouter as an alternative AI provider. It introduces a new configuration mechanism, integrates the OpenRouter API for chat completions, and updates the documentation to guide users on how to leverage these new features. Additionally, it includes improvements to code clarity through type hinting and a new sample file.
Highlights
- OpenRouter API Integration: Implemented the core logic in
lua/nes/api.luato support making API calls to OpenRouter. This includes handling API key retrieval, constructing the request payload, and processing streaming responses from the OpenRouter chat completions endpoint. - Flexible Configuration: Introduced new configuration options via
nes.setup()inlua/nes/init.luaand documented inREADME.md. Users can now specifyprovider(either 'copilot' or 'openrouter'),api_key,base_url, andmodelto customize their AI backend and model choice. - Improved Type Hinting: Added comprehensive LuaDoc type definitions in
lua/nes/context.luafor key data structures likeCurrentVersion,Context,Message, andPayload. This enhances code readability, maintainability, and tooling support. - New Sample File: A new Go language sample file,
samplecode.go, has been added, likely for testing or demonstration purposes within the project's context.
Using Gemini Code Assist
The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.
Invoking Gemini
You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.
| Feature | Command | Description |
|---|---|---|
| Code Review | /gemini review |
Performs a code review for the current pull request in its current state. |
| Pull Request Summary | /gemini summary |
Provides a summary of the current pull request in its current state. |
| Comment | @gemini-code-assist | Responds in comments when explicitly tagged, both in issue comments and review comments. |
| Help | /gemini help |
Displays a list of available commands. |
Customization
To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.
Limitations & Feedback
Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.
You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.
Footnotes
-
Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩
There was a problem hiding this comment.
Code Review
This pull request adds support for OpenRouter as a provider. The stream processing logic for OpenRouter responses is not robust and could fail if multiple events are received in one chunk. There are also some leftover debug print statements that should be removed. Finally, the README documentation for the new base_url configuration is inconsistent with the implementation, which could confuse users.
| stream = function(_, chunk) | ||
| if not chunk then | ||
| return | ||
| end | ||
| if vim.startswith(chunk, "data: ") then | ||
| chunk = chunk:sub(6) | ||
| end | ||
| if chunk == "[DONE]" then | ||
| return | ||
| end | ||
| local ok, event = pcall(vim.json.decode, chunk) | ||
| if not ok then | ||
| return | ||
| end | ||
| if event and event.choices and event.choices[1] then | ||
| local choice = event.choices[1] | ||
| if choice.delta and choice.delta.content then | ||
| output = output .. choice.delta.content | ||
| end | ||
| end | ||
| end, |
There was a problem hiding this comment.
The current stream processing logic may not be robust enough to handle cases where multiple Server-Sent Events (SSE) arrive in a single data chunk. The implementation assumes one data: line per chunk, which could lead to JSON decoding errors and incomplete responses if the provider sends multiple events together.
I recommend iterating over each line in the chunk to process each event separately. This will make the streaming logic more resilient.
stream = function(_, chunk)
if not chunk then
return
end
for line in chunk:gmatch("[^\r\n]+") do
if vim.startswith(line, "data: ") then
local data = line:sub(7)
if data == "[DONE]" then
return
end
local ok, event = pcall(vim.json.decode, data)
if ok and event and event.choices and event.choices[1] then
local choice = event.choices[1]
if choice.delta and choice.delta.content then
output = output .. choice.delta.content
end
end
end
end
end
| base_url = 'https://api.openrouter.ai/v1', -- optional, default "https://api.openrouter.ai/v1" | ||
| model = 'gpt-3.5-turbo', -- optional, default "gpt-3.5-turbo" for openrouter, "copilot-nes-v" for copilot |
There was a problem hiding this comment.
The documentation for base_url is inconsistent with the implementation. The example value and the default mentioned in the comment (https://api.openrouter.ai/v1) point to a base path, but the code expects the full endpoint URL for chat completions. The actual default in the code is https://openrouter.ai/api/v1/chat/completions.
This discrepancy will likely confuse users and lead to incorrect configurations. Please update the documentation to reflect the correct default value and clarify that the full endpoint URL is expected.
| base_url = 'https://api.openrouter.ai/v1', -- optional, default "https://api.openrouter.ai/v1" | |
| model = 'gpt-3.5-turbo', -- optional, default "gpt-3.5-turbo" for openrouter, "copilot-nes-v" for copilot | |
| base_url = 'https://openrouter.ai/api/v1/chat/completions', -- optional, default "https://openrouter.ai/api/v1/chat/completions" | |
| model = 'gpt-3.5-turbo', -- optional, default "gpt-3.5-turbo" for openrouter, "copilot-nes-v" for copilot |
| -- __AUTO_GENERATED_PRINT_VAR_START__ | ||
| print([==[M.call#if#callback output:]==], vim.inspect(output)) -- __AUTO_GENERATED_PRINT_VAR_END__ |
There was a problem hiding this comment.
It appears a debug print statement was left in the code. This should be removed before merging to avoid polluting the user's output.
| -- __AUTO_GENERATED_PRINT_VAR_START__ | ||
| print([==[M.get_suggestion#(anon) stdout:]==], vim.inspect(stdout)) -- __AUTO_GENERATED_PRINT_VAR_END__ |
There was a problem hiding this comment.
This appears to be a leftover debug print statement. It should be removed before this branch is merged.
|
Have you tried pr #6 ? |
|
I haven't tried it. Should I give it a try? |
|
Apparently if I use that I won't need this PR. |
2 participants
No description provided.