perplexityai 0.12.1__py3-none-any.whl → 0.14.0__py3-none-any.whl

perplexity/_client.py

@@ -105,6 +105,8 @@ class Perplexity(SyncAPIClient):
             _strict_response_validation=_strict_response_validation,
         )
 
+        self._default_stream_cls = Stream
+
         self.chat = chat.ChatResource(self)
         self.async_ = async_.AsyncResource(self)
         self.search = search.SearchResource(self)
@@ -277,6 +279,8 @@ class AsyncPerplexity(AsyncAPIClient):
             _strict_response_validation=_strict_response_validation,
         )
 
+        self._default_stream_cls = AsyncStream
+
         self.chat = chat.AsyncChatResource(self)
         self.async_ = async_.AsyncAsyncResource(self)
         self.search = search.AsyncSearchResource(self)

perplexity/_streaming.py

@@ -55,7 +55,26 @@ class Stream(Generic[_T]):
         iterator = self._iter_events()
 
         for sse in iterator:
-            yield process_data(data=sse.json(), cast_to=cast_to, response=response)
+            if sse.data.startswith("[DONE]"):
+                break
+
+            if sse.event == "error":
+                body = sse.data
+
+                try:
+                    body = sse.json()
+                    err_msg = f"{body}"
+                except Exception:
+                    err_msg = sse.data or f"Error code: {response.status_code}"
+
+                raise self._client._make_status_error(
+                    err_msg,
+                    body=body,
+                    response=self.response,
+                )
+
+            if sse.event is None:
+                yield process_data(data=sse.json(), cast_to=cast_to, response=response)
 
         # Ensure the entire stream is consumed
         for _sse in iterator:
@@ -119,7 +138,26 @@ class AsyncStream(Generic[_T]):
         iterator = self._iter_events()
 
         async for sse in iterator:
-            yield process_data(data=sse.json(), cast_to=cast_to, response=response)
+            if sse.data.startswith("[DONE]"):
+                break
+
+            if sse.event == "error":
+                body = sse.data
+
+                try:
+                    body = sse.json()
+                    err_msg = f"{body}"
+                except Exception:
+                    err_msg = sse.data or f"Error code: {response.status_code}"
+
+                raise self._client._make_status_error(
+                    err_msg,
+                    body=body,
+                    response=self.response,
+                )
+
+            if sse.event is None:
+                yield process_data(data=sse.json(), cast_to=cast_to, response=response)
 
         # Ensure the entire stream is consumed
         async for _sse in iterator:

perplexity/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
 
 __title__ = "perplexity"
-__version__ = "0.12.1"  # x-release-please-version
+__version__ = "0.14.0"  # x-release-please-version

perplexity/resources/async_/chat/completions.py

@@ -58,10 +58,7 @@ class CompletionsResource(SyncAPIResource):
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> CompletionCreateResponse:
         """
-        FastAPI wrapper around async chat completions
-
-        This endpoint creates an asynchronous chat completion job and returns a job ID
-        that can be used to poll for results.
+        Submit an asynchronous chat completion request.
 
         Args:
           extra_headers: Send extra headers
@@ -97,7 +94,7 @@ class CompletionsResource(SyncAPIResource):
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> CompletionListResponse:
-        """list all async chat completion requests for a given user."""
+        """Retrieve a list of all asynchronous chat completion requests for a given user."""
         return self._get(
             "/async/chat/completions",
             options=make_request_options(
@@ -113,6 +110,7 @@ class CompletionsResource(SyncAPIResource):
         local_mode: bool | Omit = omit,
         x_client_env: str | Omit = omit,
         x_client_name: str | Omit = omit,
+        x_created_at_epoch_seconds: str | Omit = omit,
         x_request_time: str | Omit = omit,
         x_usage_tier: str | Omit = omit,
         x_user_id: str | Omit = omit,
@@ -124,7 +122,7 @@ class CompletionsResource(SyncAPIResource):
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> CompletionGetResponse:
         """
-        get the response for a given async chat completion request.
+        Retrieve the response for a given asynchronous chat completion request.
 
         Args:
           extra_headers: Send extra headers
@@ -142,6 +140,7 @@ class CompletionsResource(SyncAPIResource):
                 {
                     "x-client-env": x_client_env,
                     "x-client-name": x_client_name,
+                    "x-created-at-epoch-seconds": x_created_at_epoch_seconds,
                     "x-request-time": x_request_time,
                     "x-usage-tier": x_usage_tier,
                     "x-user-id": x_user_id,
@@ -195,10 +194,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> CompletionCreateResponse:
         """
-        FastAPI wrapper around async chat completions
-
-        This endpoint creates an asynchronous chat completion job and returns a job ID
-        that can be used to poll for results.
+        Submit an asynchronous chat completion request.
 
         Args:
           extra_headers: Send extra headers
@@ -234,7 +230,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         extra_body: Body | None = None,
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> CompletionListResponse:
-        """list all async chat completion requests for a given user."""
+        """Retrieve a list of all asynchronous chat completion requests for a given user."""
         return await self._get(
             "/async/chat/completions",
             options=make_request_options(
@@ -250,6 +246,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         local_mode: bool | Omit = omit,
         x_client_env: str | Omit = omit,
         x_client_name: str | Omit = omit,
+        x_created_at_epoch_seconds: str | Omit = omit,
         x_request_time: str | Omit = omit,
         x_usage_tier: str | Omit = omit,
         x_user_id: str | Omit = omit,
@@ -261,7 +258,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
         timeout: float | httpx.Timeout | None | NotGiven = not_given,
     ) -> CompletionGetResponse:
         """
-        get the response for a given async chat completion request.
+        Retrieve the response for a given asynchronous chat completion request.
 
         Args:
           extra_headers: Send extra headers
@@ -279,6 +276,7 @@ class AsyncCompletionsResource(AsyncAPIResource):
                 {
                     "x-client-env": x_client_env,
                     "x-client-name": x_client_name,
+                    "x-created-at-epoch-seconds": x_created_at_epoch_seconds,
                     "x-request-time": x_request_time,
                     "x-usage-tier": x_usage_tier,
                     "x-user-id": x_user_id,