github-actions[bot] commented on code in PR #67671:
URL: https://github.com/apache/doris/pull/67671#discussion_r3967491087
##########
be/src/exprs/function/ai/ai_adapter.h:
##########
@@ -455,8 +464,8 @@ class LocalAdapter : public AIAdapter {
return Status::OK();
}
- Status parse_response(const std::string& response_body,
- std::vector<std::string>& results) const override {
+ Status parse_response(const std::string& response_body,
std::vector<std::string>& results,
+ bool /* expand_batch */ = true) const override {
rapidjson::Document doc;
doc.Parse(response_body.c_str());
Review Comment:
[P1] Propagate opaque mode through every text adapter
`ai_agg` now calls this virtual as `parse_response(..., false)` and rejects
any result count other than one, but this override drops the flag, as do the
Gemini, Anthropic, and Mock overrides. Each still calls
`append_parsed_text_result` with its default `true`, so a valid aggregate
result such as `["north","south"]` is expanded into two results and
deterministically trips the new exception instead of being returned intact.
This is distinct from the existing OpenAI Responses thread because that path
now forwards the flag; please propagate `expand_batch` through every
text-generating override and add a non-OpenAI opaque-mode test.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
