Compare commits

...

5 Commits

Author SHA1 Message Date
1ac58222b8 fix(editor): follow live terminal size in external editor
All checks were successful
CI Code / Check spelling (pull_request) Successful in 14s
CI Code / Check coding style (pull_request) Successful in 23s
CI Code / Code Coverage (pull_request) Successful in 3m26s
CI Code / Linux (debian) (pull_request) Successful in 5m19s
CI Code / Linux (ubuntu) (pull_request) Successful in 5m23s
CI Code / Linux (arch) (pull_request) Successful in 7m17s
The compose editor is spawned via fork+execvp and inherits profanity's
LINES/COLUMNS, which hold the size captured at startup and are never
refreshed on resize. A curses editor (nano, vim, ...) honors them over
the real window, so it renders at the launch-time size after a resize.

Drop LINES/COLUMNS from the child's environment before forking so its
curses falls back to ioctl(TIOCGWINSZ). Assembling the env in the parent
lets the child only reassign environ instead of calling unsetenv()
between fork and exec, keeping it clear of unsetenv()'s allocator work in
the multithreaded fork->exec window. profanity itself is unaffected.
2026-07-13 15:34:23 +03:00
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
9913344bbf refactor(ai): align AI client with OpenAI chat completions API
Some checks failed
CI Code / Check spelling (push) Successful in 16s
CI Code / Check coding style (push) Successful in 29s
CI Code / Linux (debian) (push) Failing after 3m38s
CI Code / Code Coverage (push) Failing after 5m19s
CI Code / Linux (ubuntu) (push) Failing after 6m15s
CI Code / Linux (arch) (push) Failing after 9m31s
Update request endpoint to /v1/chat/completions and switch payload key
from input to messages. Remove store flag and legacy Perplexity response
parsing to standardize on OpenAI's content extraction.
2026-07-04 09:58:17 +00:00
91631aa91a feat(ai): add suport for /ai set custom parameters
Modify _build_json_payload_from_list to accept an AIProvider parameter
and dynamically merge its custom settings into the JSON payload. The
settings are serialized as additional key-value pairs alongside the
standard model and input fields, enabling per-provider configuration
options without hardcoding them.
2026-07-04 09:47:28 +00:00
8 changed files with 168 additions and 47 deletions

View File

@@ -186,6 +186,22 @@ ai_json_escape(const gchar* str)
return g_string_free(result, FALSE);
}
/* Keys emitted as fixed payload fields; custom settings must not duplicate them */
static gboolean
_is_reserved_payload_key(const gchar* key)
{
return g_strcmp0(key, "model") == 0 || g_strcmp0(key, "messages") == 0 || g_strcmp0(key, "stream") == 0;
}
/* RFC 8259 scalars (number/true/false/null) may be emitted bare in a JSON payload */
static gboolean
_is_json_bare_token(const gchar* value)
{
if (g_strcmp0(value, "true") == 0 || g_strcmp0(value, "false") == 0 || g_strcmp0(value, "null") == 0)
return TRUE;
return g_regex_match_simple("^-?(0|[1-9][0-9]*)(\\.[0-9]+)?([eE][+-]?[0-9]+)?$", value, 0, 0);
}
/**
* Convert a 4-hex-digit sequence at @p to a 16-bit code point.
* Returns -1 on invalid hex.
@@ -392,6 +408,7 @@ ai_provider_new(const gchar* name, const gchar* api_url)
provider->models = NULL;
provider->models_fresh = FALSE;
provider->ref_count = 1;
pthread_mutex_init(&provider->settings_lock, NULL);
return provider;
}
@@ -419,6 +436,7 @@ ai_provider_unref(AIProvider* provider)
g_free(provider->project_id);
g_free(provider->default_model);
g_hash_table_destroy(provider->settings);
pthread_mutex_destroy(&provider->settings_lock);
GList* curr = provider->models;
while (curr) {
@@ -729,31 +747,44 @@ ai_get_provider_default_model(const gchar* provider_name)
return provider->default_model;
}
void
gboolean
ai_set_provider_setting(const gchar* provider_name, const gchar* setting, const gchar* value)
{
if (!provider_name || !setting)
return;
return FALSE;
if (_is_reserved_payload_key(setting)) {
log_warning("Setting '%s' for provider '%s' rejected: reserved payload key", setting, provider_name);
return FALSE;
}
AIProvider* provider = ai_get_provider(provider_name);
if (!provider) {
log_warning("Provider '%s' not found for setting '%s'", provider_name, setting);
return;
return FALSE;
}
/* settings is iterated by the request thread; mutate under settings_lock */
pthread_mutex_lock(&provider->settings_lock);
if (!provider->settings) {
provider->settings = g_hash_table_new_full(g_str_hash, g_str_equal, g_free, g_free);
}
if (value) {
g_hash_table_insert(provider->settings, g_strdup(setting), g_strdup(value));
} else {
g_hash_table_remove(provider->settings, setting);
}
pthread_mutex_unlock(&provider->settings_lock);
if (value) {
prefs_ai_set_setting(provider_name, setting, value);
log_info("Setting '%s' for provider '%s' set to: %s", setting, provider_name, value);
} else {
g_hash_table_remove(provider->settings, setting);
prefs_ai_remove_setting(provider_name, setting);
log_info("Setting '%s' removed for provider '%s'", setting, provider_name);
}
return TRUE;
}
gchar*
@@ -766,8 +797,10 @@ ai_get_provider_setting(const gchar* provider_name, const gchar* setting)
if (!provider || !provider->settings)
return NULL;
gchar* value = g_hash_table_lookup(provider->settings, setting);
return g_strdup(value);
pthread_mutex_lock(&provider->settings_lock);
gchar* value = g_strdup(g_hash_table_lookup(provider->settings, setting));
pthread_mutex_unlock(&provider->settings_lock);
return value;
}
/* ========================================================================
@@ -1435,13 +1468,19 @@ ai_session_switch(AISession* session, const gchar* provider_name,
* This is the thread-safe variant used by _ai_request_thread() which holds
* the session lock while calling it.
*
* Custom provider settings are merged into the payload as additional JSON
* key-value pairs alongside "model" and "messages": scalar values
* (number/true/false/null) are emitted bare, anything else as a JSON string;
* reserved keys (model/messages/stream) are skipped.
*
* @param provider The AI provider (for custom settings)
* @param model The model identifier
* @param history GList of AIMessage* (must be held under session lock)
* @param prompt The new user prompt to append
* @return Newly allocated JSON string (caller must free)
*/
static gchar*
_build_json_payload_from_list(const gchar* model, GList* history, const gchar* prompt)
_build_json_payload_from_list(AIProvider* provider, const gchar* model, GList* history, const gchar* prompt)
{
GString* messages_json = g_string_new("");
GList* curr = history;
@@ -1468,11 +1507,50 @@ _build_json_payload_from_list(const gchar* model, GList* history, const gchar* p
g_string_append_printf(messages_json, "{\"role\":\"user\",\"content\":\"%s\"}", escaped_prompt);
auto_gchar gchar* escaped_model = ai_json_escape(model);
gchar* json_payload = g_strdup_printf(
"{\"model\":\"%s\",\"input\":[%s],\"stream\":false,\"store\":false}",
escaped_model, messages_json->str);
/* Build custom settings JSON fragment; settings is mutated on the main
* thread (/ai set custom), so iterate under settings_lock */
GString* custom_json = g_string_new("");
if (provider) {
pthread_mutex_lock(&provider->settings_lock);
if (provider->settings) {
GHashTableIter iter;
gpointer key, value;
g_hash_table_iter_init(&iter, provider->settings);
while (g_hash_table_iter_next(&iter, &key, &value)) {
if (_is_reserved_payload_key((gchar*)key)) {
continue; /* fixed payload fields stay authoritative */
}
auto_gchar gchar* escaped_key = ai_json_escape((gchar*)key);
if (custom_json->len > 0) {
g_string_append_c(custom_json, ',');
}
if (_is_json_bare_token((gchar*)value)) {
g_string_append_printf(custom_json, "\"%s\": %s", escaped_key, (gchar*)value);
} else {
/* non-scalar values must be quoted, else the payload is invalid JSON */
auto_gchar gchar* escaped_val = ai_json_escape((gchar*)value);
g_string_append_printf(custom_json, "\"%s\": \"%s\"", escaped_key, escaped_val);
}
}
}
pthread_mutex_unlock(&provider->settings_lock);
}
/* Assemble final payload: model, messages, custom settings, then fixed keys */
gchar* json_payload;
if (custom_json->len > 0) {
json_payload = g_strdup_printf(
"{\"model\":\"%s\",\"messages\":[%s],%s,\"stream\":false}",
escaped_model, messages_json->str, custom_json->str);
} else {
json_payload = g_strdup_printf(
"{\"model\":\"%s\",\"messages\":[%s],\"stream\":false}",
escaped_model, messages_json->str);
}
g_string_free(messages_json, TRUE);
g_string_free(custom_json, TRUE);
return json_payload;
}
@@ -1482,12 +1560,7 @@ ai_parse_response(const gchar* response_json)
if (!response_json || strlen(response_json) == 0)
return NULL;
/* Perplexity /v1/agent: {"output":[{"content":[{"text":"...","type":"output_text"}]}]}
* Legacy OpenAI: {"choices":[{"message":{"content":"..."}}]} */
gchar* text = _extract_json_string(response_json, "text");
if (text)
return text;
/* OpenAI /v1/chat/completions: {"choices":[{"message":{"content":"..."}}]} */
return _extract_json_string(response_json, "content");
}
@@ -1566,7 +1639,7 @@ _ai_request_thread(gpointer data)
/* Build JSON payload from history + prompt (under lock), then add user message to history */
log_debug("[AI-THREAD] Building JSON payload...");
pthread_mutex_lock(&session->lock);
auto_gchar gchar* json_payload = _build_json_payload_from_list(local_model, session->history, prompt);
auto_gchar gchar* json_payload = _build_json_payload_from_list(local_provider, local_model, session->history, prompt);
/* Add user message to history now (it's in JSON but not yet in history) */
AIMessage* user_msg = g_new0(AIMessage, 1);
@@ -1587,7 +1660,7 @@ _ai_request_thread(gpointer data)
/* Build request URL */
const gchar* api_url = local_provider->api_url;
auto_gchar gchar* request_url = g_strdup_printf("%s%sv1/responses", api_url, g_str_has_suffix(api_url, "/") ? "" : "/");
auto_gchar gchar* request_url = g_strdup_printf("%s%sv1/chat/completions", api_url, g_str_has_suffix(api_url, "/") ? "" : "/");
log_debug("[AI-THREAD] API URL: %s", api_url);
log_debug("[AI-THREAD] API Request URL: %s", request_url);
log_debug("[AI-THREAD] Model: %s", local_model);

View File

@@ -17,14 +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 */
GHashTable* settings; /* Extensible per-provider settings (e.g., tools=enabled, search=disabled) */
GList* models; /* Cached models (gchar*) */
gboolean models_fresh; /* Whether models cache is current */
gint32 ref_count; /* Reference count (atomic) */
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) */
} AIProvider;
/**
@@ -142,11 +143,13 @@ const gchar* ai_get_provider_default_model(const gchar* provider_name);
/**
* Set a custom setting for a provider.
* Reserved payload keys (model, messages, stream) are rejected.
* @param provider_name The provider name
* @param setting The setting key (e.g., "tools", "search")
* @param value The setting value
* @param setting The setting key (e.g., "temperature")
* @param value The setting value, or NULL to remove
* @return TRUE on success, FALSE on unknown provider or reserved key
*/
void ai_set_provider_setting(const gchar* provider_name, const gchar* setting, const gchar* value);
gboolean ai_set_provider_setting(const gchar* provider_name, const gchar* setting, const gchar* value);
/**
* Get a custom setting for a provider.
@@ -184,9 +187,9 @@ gboolean ai_models_are_fresh(const gchar* provider_name);
void ai_parse_models_from_json(AIProvider* provider, const gchar* json);
/**
* Extract assistant content from an LLM response body. Tries Perplexity
* /v1/responses "text" first, falls back to OpenAI "content". Decodes
* the \" and \n escape sequences only.
* Extract assistant content from an OpenAI chat-completions response body
* ({"choices":[{"message":{"content":"..."}}]}). Decodes the \" and \n
* escape sequences only.
* @param response_json The JSON body from the provider
* @return Newly allocated content string, or NULL if not found (caller frees)
*/

View File

@@ -2833,7 +2833,7 @@ static const struct cmd_t command_defs[] = {
{ "set provider <name> <url>", "Add or update a provider with custom API endpoint" },
{ "set token <provider> <token>", "Set API token for a provider (e.g., openai, perplexity)" },
{ "set default-model <provider> <model>", "Set default model for a provider" },
{ "set custom <provider> <setting> <value>", "Set provider-level setting (e.g., tools, search)" },
{ "set custom <provider> <setting> <value>", "Set provider-level setting (e.g., temperature, max_tokens)" },
{ "remove provider <name>", "Remove a custom provider" },
{ "providers", "List configured providers with full details" },
{ "start <provider> [<model>]", "Start new AI chat (space-separated, uses defaults if omitted)" },
@@ -2846,7 +2846,7 @@ static const struct cmd_t command_defs[] = {
"/ai set token perplexity pplx-xxx",
"/ai set provider custom https://my-api.com/v1/chat/completions",
"/ai set default-model perplexity sonar",
"/ai set custom perplexity tools enabled",
"/ai set custom perplexity temperature 0.7",
"/ai remove provider custom",
"/ai start",
"/ai start perplexity",

View File

@@ -10787,8 +10787,11 @@ cmd_ai_set(ProfWin* window, const char* const command, gchar** args)
cons_bad_cmd_usage(command);
return TRUE;
}
ai_set_provider_setting(args[2], args[3], args[4]);
cons_show("Setting '%s' for provider '%s' set to: %s", args[3], args[2], args[4]);
if (ai_set_provider_setting(args[2], args[3], args[4])) {
cons_show("Setting '%s' for provider '%s' set to: %s", args[3], args[2], args[4]);
} else {
cons_show_error("Cannot set '%s' for provider '%s': unknown provider or reserved key (model, messages, stream).", args[3], args[2]);
}
cons_show("");
return TRUE;
}

View File

@@ -25,6 +25,8 @@
#include "ui/ui.h"
#include "xmpp/xmpp.h"
extern char** environ;
typedef struct EditorContext
{
gchar* filename;
@@ -140,6 +142,22 @@ launch_editor(gchar* initial_content, void (*callback)(gchar* content, void* dat
GSource* sigchld_warmup = g_child_watch_source_new(getpid());
g_source_unref(sigchld_warmup);
// Build the editor's env without LINES/COLUMNS pre-fork, so the child only
// reassigns environ instead of calling unsetenv() between fork and exec.
// The editor's (n)curses then reads the live window via ioctl(TIOCGWINSZ).
gsize env_len = 0;
while (environ[env_len]) {
env_len++;
}
gchar** editor_env = g_new0(gchar*, env_len + 1);
gsize env_kept = 0;
for (gsize i = 0; i < env_len; i++) {
if (g_str_has_prefix(environ[i], "LINES=") || g_str_has_prefix(environ[i], "COLUMNS=")) {
continue;
}
editor_env[env_kept++] = environ[i];
}
pid_t pid = fork();
if (pid == -1) {
log_error("[Editor] Failed to fork: %s", strerror(errno));
@@ -148,12 +166,15 @@ launch_editor(gchar* initial_content, void (*callback)(gchar* content, void* dat
ui_resize();
cons_show_error("Failed to start editor: %s", strerror(errno));
g_strfreev(editor_argv);
g_free(editor_env); // array only; strings are borrowed from environ
g_free(ctx->filename);
g_free(ctx);
return TRUE;
} else if (pid == 0) {
// Child process: Inherits TTY from parent
environ = editor_env; // live TIOCGWINSZ size, not the inherited LINES/COLUMNS
// SIGTSTP=SIG_DFL lets vim's :stop / Ctrl-Z work; profanity catches
// the STOPPED state via editor_check_stopped() and drops to the shell.
signal(SIGINT, SIG_DFL);
@@ -170,6 +191,7 @@ launch_editor(gchar* initial_content, void (*callback)(gchar* content, void* dat
editor_pid = pid;
g_child_watch_add((GPid)pid, _editor_exit_cb, ctx);
g_strfreev(editor_argv);
g_free(editor_env); // frees the parent's array only; the child keeps its own COW copy until execvp
return FALSE;
}

View File

@@ -614,6 +614,24 @@ test_ai_set_provider_setting(void** state)
assert_null(ai_get_provider_setting("p_set_settings", "nonexistent_setting"));
}
void
test_ai_set_provider_setting_reserved_key(void** state)
{
ai_add_provider("p_reserved_keys", "https://example.test/rk/");
/* Reserved payload keys are rejected (would duplicate fixed fields) */
assert_false(ai_set_provider_setting("p_reserved_keys", "model", "x"));
assert_false(ai_set_provider_setting("p_reserved_keys", "messages", "[]"));
assert_false(ai_set_provider_setting("p_reserved_keys", "stream", "true"));
assert_null(ai_get_provider_setting("p_reserved_keys", "model"));
/* Ordinary keys still work */
assert_true(ai_set_provider_setting("p_reserved_keys", "temperature", "0.7"));
auto_gchar gchar* temp = ai_get_provider_setting("p_reserved_keys", "temperature");
assert_non_null(temp);
assert_string_equal("0.7", temp);
}
void
test_ai_get_provider_setting(void** state)
{
@@ -919,22 +937,22 @@ test_ai_parse_response_openai_content(void** state)
}
void
test_ai_parse_response_perplexity_text(void** state)
test_ai_parse_response_legacy_text_format_unsupported(void** state)
{
/* Legacy Perplexity /v1/agent format: "content" is an array, not a string.
* Dropped with the chat-completions alignment, so parsing yields NULL. */
const gchar* json = "{\"output\":[{\"content\":[{\"text\":\"Search result\",\"type\":\"output_text\"}]}]}";
auto_gchar gchar* out = ai_parse_response(json);
assert_non_null(out);
assert_string_equal("Search result", out);
assert_null(ai_parse_response(json));
}
void
test_ai_parse_response_text_preferred_over_content(void** state)
test_ai_parse_response_content_preferred_over_text(void** state)
{
/* When both formats are present, "text" wins (Perplexity path tried first). */
/* Chat-completions "content" is extracted; a stray "text" field is ignored. */
const gchar* json = "{\"text\":\"from text field\",\"content\":\"from content field\"}";
auto_gchar gchar* out = ai_parse_response(json);
assert_non_null(out);
assert_string_equal("from text field", out);
assert_string_equal("from content field", out);
}
void

View File

@@ -43,6 +43,7 @@ void test_ai_providers_find_case_insensitive(void** state);
void test_ai_set_provider_default_model(void** state);
void test_ai_get_provider_default_model(void** state);
void test_ai_set_provider_setting(void** state);
void test_ai_set_provider_setting_reserved_key(void** state);
void test_ai_get_provider_setting(void** state);
/* Model caching tests */
void test_ai_models_are_fresh_initial(void** state);
@@ -64,8 +65,8 @@ int ai_client_teardown_with_prefs(void** state);
/* Chat response parser tests (ai_parse_response) */
void test_ai_parse_response_openai_content(void** state);
void test_ai_parse_response_perplexity_text(void** state);
void test_ai_parse_response_text_preferred_over_content(void** state);
void test_ai_parse_response_legacy_text_format_unsupported(void** state);
void test_ai_parse_response_content_preferred_over_text(void** state);
void test_ai_parse_response_escaped_quote(void** state);
void test_ai_parse_response_newline_escape(void** state);
void test_ai_parse_response_empty_content(void** state);

View File

@@ -752,6 +752,7 @@ main(int argc, char* argv[])
cmocka_unit_test_setup_teardown(test_ai_set_provider_default_model, ai_client_setup, ai_client_teardown),
cmocka_unit_test_setup_teardown(test_ai_get_provider_default_model, ai_client_setup, ai_client_teardown),
cmocka_unit_test_setup_teardown(test_ai_set_provider_setting, ai_client_setup, ai_client_teardown),
cmocka_unit_test_setup_teardown(test_ai_set_provider_setting_reserved_key, ai_client_setup, ai_client_teardown),
cmocka_unit_test_setup_teardown(test_ai_get_provider_setting, ai_client_setup, ai_client_teardown),
/* Model caching tests */
cmocka_unit_test_setup_teardown(test_ai_models_are_fresh_initial, ai_client_setup, ai_client_teardown),
@@ -774,8 +775,8 @@ main(int argc, char* argv[])
cmocka_unit_test(test_ai_json_escape_backslash_quote),
/* Chat response parser */
cmocka_unit_test(test_ai_parse_response_openai_content),
cmocka_unit_test(test_ai_parse_response_perplexity_text),
cmocka_unit_test(test_ai_parse_response_text_preferred_over_content),
cmocka_unit_test(test_ai_parse_response_legacy_text_format_unsupported),
cmocka_unit_test(test_ai_parse_response_content_preferred_over_text),
cmocka_unit_test(test_ai_parse_response_escaped_quote),
cmocka_unit_test(test_ai_parse_response_newline_escape),
cmocka_unit_test(test_ai_parse_response_empty_content),