fix: parse built-in repl prompts instead of faulty custom markers - #55
fix: parse built-in repl prompts instead of faulty custom markers#55shadowusr wants to merge 1 commit into
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: e575ef281b
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| this._buffer = ""; | ||
| this._pending = { | ||
| startMarker, | ||
| endMarker, | ||
| kind: "prompt", |
There was a problem hiding this comment.
When sendRaw() is the first operation on a fresh ReplConnection, the REPL's initial > prompt can arrive after this prompt waiter is installed but before the line has been evaluated. That startup prompt then resolves the fallback command immediately, so the next raw line can be completed by the fallback's later undefined\n> output and return the wrong result while the user's actual output remains on the socket; drain or explicitly ignore the connection's initial prompt before using prompt-only completion.
Useful? React with 👍 / 👎.
1 participant
No description provided.