fix(ai): close parser and AUTO fallback gaps

Parser:
- Never return reasoning text as the reply: the no-output_text path
  anchors strictly on the message "content" array; a reasoning-only body
  (truncated by max_output_tokens, carries only "summary") now fails the
  parse instead of leaking the chain-of-thought. Regression test added.
- _find_json_field skips field-name occurrences that are string values
  (no ':' after), so {"type":"text","text":"hi"} parses instead of
  failing on the value of "type".
- _same_json_object tracks brace depth, so a nested object between the
  "text" key and the "output_text" tag (e.g. "annotations") no longer
  rejects the part's own text.
- Chat-completions extraction is bounded to the choices array (new
  _json_array_end helper), so "message"/"content" in a sibling object
  (e.g. a top-level "warning") is not returned as the reply.

AUTO fallback:
- Model errors are classified only from a structured JSON error
  envelope; a route-missing 404 page merely mentioning models (e.g.
  listing /v1/models) no longer suppresses the fallback. Trade-off: a
  model error in a non-JSON body costs one extra fallback attempt.
- Fall back on a curl-level failure: an endpoint that accepts the
  connection but never responds no longer burns the 60s timeout without
  trying the flavour that works.
- Fall back on an unparseable 2xx: a gateway answering unknown paths
  with a 200 error page previously reset the hint and re-probed the same
  dead flavour on every request. The body is parsed inside the attempt
  loop and no longer re-parsed after it.
- The second attempt's 401/403/429/5xx now wins over the stashed 400/422
  payload rejection, so an invalid key or rate limit is not hidden
  behind a payload-shape error; a second 404 still defers to the first.
- When the first flavour's endpoint is missing but the other answers
  (even unsuccessfully, e.g. 401), the resolved hint moves to the
  surviving flavour, so the dead endpoint is not probed first forever.

Settings:
- Reserved payload keys are rejected only when setting a value; removal
  is allowed again, so a setting persisted before its key became
  reserved (e.g. "input") is no longer stuck in the config.
This commit is contained in:
2026-07-08 20:31:19 +03:00
parent b3a269342d
commit 3a96d0dba1
4 changed files with 151 additions and 49 deletions

View File

@@ -1074,6 +1074,19 @@ test_ai_parse_response_responses_reasoning_summary_skipped(void** state)
assert_string_equal("Actual answer", out);
}
void
test_ai_parse_response_responses_reasoning_only_returns_null(void** state)
{
/* An incomplete/truncated body with only a reasoning item (no message
* "content" array at all) must yield a parse failure, not leak the
* reasoning summary's "text" as the assistant reply. */
const gchar* json = "{\"status\":\"incomplete\",\"output\":["
"{\"type\":\"reasoning\",\"summary\":[{\"type\":\"summary_text\",\"text\":\"secret reasoning\"}]}"
"]}";
gchar* out = ai_parse_response(json);
assert_null(out);
}
void
test_ai_parse_response_choices_in_string_value(void** state)
{

View File

@@ -78,6 +78,7 @@ void test_ai_parse_response_content_preferred_over_text(void** state);
void test_ai_parse_response_responses_brace_in_text(void** state);
void test_ai_parse_response_choices_logprobs_before_content(void** state);
void test_ai_parse_response_reasoning_not_leaked_without_text(void** state);
void test_ai_parse_response_responses_reasoning_only_returns_null(void** state);
void test_ai_parse_response_responses_text_value_text(void** state);
void test_ai_parse_response_escaped_quote(void** state);
void test_ai_parse_response_newline_escape(void** state);

View File

@@ -788,6 +788,7 @@ main(int argc, char* argv[])
cmocka_unit_test(test_ai_parse_response_responses_brace_in_text),
cmocka_unit_test(test_ai_parse_response_choices_logprobs_before_content),
cmocka_unit_test(test_ai_parse_response_reasoning_not_leaked_without_text),
cmocka_unit_test(test_ai_parse_response_responses_reasoning_only_returns_null),
cmocka_unit_test(test_ai_parse_response_responses_text_value_text),
cmocka_unit_test(test_ai_parse_response_escaped_quote),
cmocka_unit_test(test_ai_parse_response_newline_escape),