Compare commits

..

2 Commits

Author SHA1 Message Date
5d7b7bb23d test(ai): align parse_response tests with chat completions API
All checks were successful
CI Code / Check coding style (pull_request) Successful in 29s
CI Code / Check spelling (pull_request) Successful in 15s
CI Code / Code Coverage (pull_request) Successful in 3m37s
CI Code / Linux (debian) (pull_request) Successful in 7m36s
CI Code / Linux (ubuntu) (pull_request) Successful in 7m52s
CI Code / Linux (arch) (pull_request) Successful in 11m26s
CI Code / Check spelling (push) Successful in 15s
CI Code / Check coding style (push) Successful in 28s
CI Code / Code Coverage (push) Successful in 3m30s
CI Code / Linux (debian) (push) Successful in 5m19s
CI Code / Linux (ubuntu) (push) Successful in 5m25s
CI Code / Linux (arch) (push) Successful in 7m11s
The chat-completions refactor removed the legacy Perplexity "text"
extraction but left two tests asserting the old behavior, breaking
make check on every CI flavor:

- test_ai_parse_response_perplexity_text expected the legacy /v1/agent
  nested format to parse; it now yields NULL — renamed to
  test_ai_parse_response_legacy_text_format_unsupported.
- test_ai_parse_response_text_preferred_over_content expected "text"
  to win; "content" is authoritative now — renamed to
  test_ai_parse_response_content_preferred_over_text.

Add test_ai_set_provider_setting_reserved_key covering the new
reserved-key rejection.
2026-07-04 14:07:22 +03:00
fa6857f4c8 fix(ai): harden custom settings payload against invalid JSON and races
Custom setting values were JSON-escaped but emitted unquoted, so any
non-numeric value (e.g. "high") produced an invalid JSON payload and
failed the whole request; quoting the value manually could not work
either, since the escaper turns quotes into \". Emit RFC 8259 scalars
(number/true/false/null) bare and everything else as a quoted JSON
string.

Reserved payload keys (model, messages, stream) would duplicate the
fixed fields with parser-dependent precedence; reject them in
ai_set_provider_setting (surfaced as an error by /ai set custom) and
skip them at payload-build time for settings loaded from hand-edited
prefs.

provider->settings was mutated on the main thread while the request
thread iterates it when building the payload; guard both sides with a
new per-provider settings_lock.

Also refresh stale docs: ai_parse_response no longer mentions the
dropped Perplexity "text" path, the payload docstring says "messages"
instead of "input", and the /ai help example uses a realistic scalar
setting.
2026-07-04 14:07:22 +03:00

View File

@@ -17,15 +17,15 @@ typedef struct ai_message_t
*/
typedef struct ai_provider_t
{
gchar* name; /* Provider name (e.g., "openai", "perplexity") */
gchar* api_url; /* API endpoint URL */
gchar* project_id; /* Optional project ID (for some providers) */
gchar* default_model; /* Default model for this provider */
gchar* name; /* Provider name (e.g., "openai", "perplexity") */
gchar* api_url; /* API endpoint URL */
gchar* project_id; /* Optional project ID (for some providers) */
gchar* default_model; /* Default model for this provider */
pthread_mutex_t settings_lock; /* Protects settings (iterated by the request thread) */
GHashTable* settings; /* Extensible per-provider settings (e.g., temperature=0.7) */
GList* models; /* Cached models (gchar*) */
gboolean models_fresh; /* Whether models cache is current */
gint32 ref_count; /* Reference count (atomic) */
GHashTable* settings; /* Extensible per-provider settings (e.g., temperature=0.7) */
GList* models; /* Cached models (gchar*) */
gboolean models_fresh; /* Whether models cache is current */
gint32 ref_count; /* Reference count (atomic) */
} AIProvider;
/**