aria-code 0.111.0__tar.gz → 0.112.0__tar.gz

{aria_code-0.111.0/src/aria_code.egg-info → aria_code-0.112.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: aria-code
-Version: 0.111.0
+Version: 0.112.0
 Summary: Open-source AI agent for coding, financial analysis and logistics operations: numbers you can check, client data kept apart, actions a person approves
 License: Apache-2.0
 Project-URL: Homepage, https://github.com/artheras/aria-code
@@ -171,7 +171,7 @@ Numbers you can check, client data kept apart, actions a person approves.</p>
 <p align="center">
   <a href="https://www.npmjs.com/package/@artheras/aria-code"><img src="https://img.shields.io/npm/v/@artheras/aria-code/latest?style=flat-square&logo=npm&label=npm" alt="npm version"></a>
   <!-- Release automation pins PyPI: the default endpoint still selects legacy 4.x. -->
-  <a href="https://pypi.org/project/aria-code/0.111.0/"><img src="https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fpypi.org%2Fpypi%2Faria-code%2F0.111.0%2Fjson&amp;query=%24.info.version&amp;prefix=v&amp;style=flat-square&amp;logo=pypi&amp;label=PyPI&amp;color=blue&amp;cacheSeconds=300" alt="PyPI version"></a>
+  <a href="https://pypi.org/project/aria-code/0.112.0/"><img src="https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fpypi.org%2Fpypi%2Faria-code%2F0.112.0%2Fjson&amp;query=%24.info.version&amp;prefix=v&amp;style=flat-square&amp;logo=pypi&amp;label=PyPI&amp;color=blue&amp;cacheSeconds=300" alt="PyPI version"></a>
   <a href="https://github.com/artheras/aria-code/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/artheras/aria-code/ci.yml?branch=main&style=flat-square&logo=githubactions&label=CI" alt="CI status"></a>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-Apache%202.0-64748b?style=flat-square" alt="Apache License 2.0"></a>
 </p>

{aria_code-0.111.0 → aria_code-0.112.0}/README.md

@@ -15,7 +15,7 @@ Numbers you can check, client data kept apart, actions a person approves.</p>
 <p align="center">
   <a href="https://www.npmjs.com/package/@artheras/aria-code"><img src="https://img.shields.io/npm/v/@artheras/aria-code/latest?style=flat-square&logo=npm&label=npm" alt="npm version"></a>
   <!-- Release automation pins PyPI: the default endpoint still selects legacy 4.x. -->
-  <a href="https://pypi.org/project/aria-code/0.111.0/"><img src="https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fpypi.org%2Fpypi%2Faria-code%2F0.111.0%2Fjson&amp;query=%24.info.version&amp;prefix=v&amp;style=flat-square&amp;logo=pypi&amp;label=PyPI&amp;color=blue&amp;cacheSeconds=300" alt="PyPI version"></a>
+  <a href="https://pypi.org/project/aria-code/0.112.0/"><img src="https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fpypi.org%2Fpypi%2Faria-code%2F0.112.0%2Fjson&amp;query=%24.info.version&amp;prefix=v&amp;style=flat-square&amp;logo=pypi&amp;label=PyPI&amp;color=blue&amp;cacheSeconds=300" alt="PyPI version"></a>
   <a href="https://github.com/artheras/aria-code/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/artheras/aria-code/ci.yml?branch=main&style=flat-square&logo=githubactions&label=CI" alt="CI status"></a>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-Apache%202.0-64748b?style=flat-square" alt="Apache License 2.0"></a>
 </p>

{aria_code-0.111.0 → aria_code-0.112.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "aria-code"
-version = "0.111.0"
+version = "0.112.0"
 description = "Open-source AI agent for coding, financial analysis and logistics operations: numbers you can check, client data kept apart, actions a person approves"
 readme = "README.md"
 license = { text = "Apache-2.0" }

{aria_code-0.111.0 → aria_code-0.112.0}/src/aria_code/_version.py

@@ -17,4 +17,4 @@ from __future__ import annotations
 
 __all__ = ["__version__"]
 
-__version__ = "0.111.0"
+__version__ = "0.112.0"

{aria_code-0.111.0 → aria_code-0.112.0}/src/aria_code/apps/cli/providers/vertexai_stream.py

@@ -33,6 +33,37 @@ logger = logging.getLogger(__name__)
 
 _EMPTY_ROUND_RETRIES = 2
 
+# Waits before re-sending a request Vertex turned away as busy: 429
+# RESOURCE_EXHAUSTED (quota or shared capacity) and 500/503/504. The 40-task
+# eval run lost inventory-reorder this way, mid-task, to a single 429 that
+# ended the turn. Retried only while the round has produced nothing, so no
+# text or tool call is ever delivered twice.
+_BUSY_BACKOFF = (2, 5, 10, 20, 40)
+_BUSY_CODES = frozenset({429, 500, 503, 504})
+_BUSY_STATUSES = frozenset({"RESOURCE_EXHAUSTED", "UNAVAILABLE", "INTERNAL", "DEADLINE_EXCEEDED"})
+_sleep = asyncio.sleep
+
+
+def _busy(error) -> bool:
+    """True when Vertex refused the request for load, not for anything in it."""
+    return (getattr(error, "code", None) in _BUSY_CODES
+            or str(getattr(error, "status", "") or "").upper() in _BUSY_STATUSES)
+
+
+async def _wait(seconds: float, cancel_event: Optional[asyncio.Event]) -> bool:
+    """Sleep up to *seconds*; True if the user cancelled meanwhile."""
+    if cancel_event is None:
+        await _sleep(seconds)
+        return False
+    waited = 0.0
+    while waited < seconds:
+        if cancel_event.is_set():
+            return True
+        step = min(0.5, seconds - waited)
+        await _sleep(step)
+        waited += step
+    return cancel_event.is_set()
+
 
 def _chunk_text(chunk) -> str:
     """The text parts of a chunk, read without chunk.text.
@@ -311,38 +342,58 @@ class VertexAIProvider(LLMProvider):
             usage = {}
             tool_calls = []
             finish_reasons: list = []
+            busy_tries = 0
             for attempt in range(1 + _EMPTY_ROUND_RETRIES):
-                response_stream = await client.aio.models.generate_content_stream(
-                    model=self.model,
-                    contents=contents,
-                    config=config,
-                )
-                async for chunk in response_stream:
-                    if cancel_event and cancel_event.is_set():
-                        yield LLMDone(response=full_response, provider="vertexai", success=True, cancelled=True)
-                        return
-
-                    text = _chunk_text(chunk)
-                    if text:
-                        full_response += text
-                        yield LLMToken(text=text)
-
-                    if chunk.function_calls:
-                        for fc in chunk.function_calls:
-                            args = {k: v for k, v in fc.args.items()} if fc.args else {}
-                            yield LLMToolCall(tool=fc.name, params=args)
-                            tool_calls.append({"tool": fc.name, "params": args})
-
-                    for candidate in (getattr(chunk, "candidates", None) or []):
-                        reason = getattr(candidate, "finish_reason", None)
-                        if reason:
-                            finish_reasons.append(str(getattr(reason, "name", reason)))
-
-                    if chunk.usage_metadata:
-                        usage = {
-                            "prompt_tokens": chunk.usage_metadata.prompt_token_count,
-                            "completion_tokens": chunk.usage_metadata.candidates_token_count,
-                        }
+                # The SDK sends the request when the stream is first read, so
+                # a 429 surfaces inside the loop below, not at the call. A busy
+                # refusal is not an empty round: it has its own budget and does
+                # not use up one of those retries.
+                while True:
+                    try:
+                        response_stream = await client.aio.models.generate_content_stream(
+                            model=self.model,
+                            contents=contents,
+                            config=config,
+                        )
+                        async for chunk in response_stream:
+                            if cancel_event and cancel_event.is_set():
+                                yield LLMDone(response=full_response, provider="vertexai", success=True, cancelled=True)
+                                return
+
+                            text = _chunk_text(chunk)
+                            if text:
+                                full_response += text
+                                yield LLMToken(text=text)
+
+                            if chunk.function_calls:
+                                for fc in chunk.function_calls:
+                                    args = {k: v for k, v in fc.args.items()} if fc.args else {}
+                                    yield LLMToolCall(tool=fc.name, params=args)
+                                    tool_calls.append({"tool": fc.name, "params": args})
+
+                            for candidate in (getattr(chunk, "candidates", None) or []):
+                                reason = getattr(candidate, "finish_reason", None)
+                                if reason:
+                                    finish_reasons.append(str(getattr(reason, "name", reason)))
+
+                            if chunk.usage_metadata:
+                                usage = {
+                                    "prompt_tokens": chunk.usage_metadata.prompt_token_count,
+                                    "completion_tokens": chunk.usage_metadata.candidates_token_count,
+                                }
+                        break
+                    except APIError as e:
+                        if full_response or tool_calls or not _busy(e) or busy_tries >= len(_BUSY_BACKOFF):
+                            raise
+                        delay = _BUSY_BACKOFF[busy_tries]
+                        busy_tries += 1
+                        finish_reasons = []
+                        logger.info("Vertex is busy (%s %s); retrying in %ss (%d/%d)",
+                                    getattr(e, "code", ""), getattr(e, "status", ""),
+                                    delay, busy_tries, len(_BUSY_BACKOFF))
+                        if await _wait(delay, cancel_event):
+                            yield LLMDone(response="", provider="vertexai", success=True, cancelled=True)
+                            return
                 if full_response.strip() or tool_calls:
                     break
                 logger.info("Vertex returned an empty round (finish_reason %s), attempt %d",

{aria_code-0.111.0 → aria_code-0.112.0/src/aria_code.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: aria-code
-Version: 0.111.0
+Version: 0.112.0
 Summary: Open-source AI agent for coding, financial analysis and logistics operations: numbers you can check, client data kept apart, actions a person approves
 License: Apache-2.0
 Project-URL: Homepage, https://github.com/artheras/aria-code
@@ -171,7 +171,7 @@ Numbers you can check, client data kept apart, actions a person approves.</p>
 <p align="center">
   <a href="https://www.npmjs.com/package/@artheras/aria-code"><img src="https://img.shields.io/npm/v/@artheras/aria-code/latest?style=flat-square&logo=npm&label=npm" alt="npm version"></a>
   <!-- Release automation pins PyPI: the default endpoint still selects legacy 4.x. -->
-  <a href="https://pypi.org/project/aria-code/0.111.0/"><img src="https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fpypi.org%2Fpypi%2Faria-code%2F0.111.0%2Fjson&amp;query=%24.info.version&amp;prefix=v&amp;style=flat-square&amp;logo=pypi&amp;label=PyPI&amp;color=blue&amp;cacheSeconds=300" alt="PyPI version"></a>
+  <a href="https://pypi.org/project/aria-code/0.112.0/"><img src="https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fpypi.org%2Fpypi%2Faria-code%2F0.112.0%2Fjson&amp;query=%24.info.version&amp;prefix=v&amp;style=flat-square&amp;logo=pypi&amp;label=PyPI&amp;color=blue&amp;cacheSeconds=300" alt="PyPI version"></a>
   <a href="https://github.com/artheras/aria-code/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/artheras/aria-code/ci.yml?branch=main&style=flat-square&logo=githubactions&label=CI" alt="CI status"></a>
   <a href="LICENSE"><img src="https://img.shields.io/badge/license-Apache%202.0-64748b?style=flat-square" alt="Apache License 2.0"></a>
 </p>

{aria_code-0.111.0 → aria_code-0.112.0}/src/aria_code.egg-info/SOURCES.txt

@@ -773,6 +773,7 @@ tests/test_verification_hint.py
 tests/test_verification_policy.py
 tests/test_version_consistency.py
 tests/test_version_counter.py
+tests/test_vertex_busy_retry.py
 tests/test_vertex_empty_rounds.py
 tests/test_vertex_models.py
 tests/test_video_analysis.py

aria_code-0.112.0/tests/test_vertex_busy_retry.py

@@ -0,0 +1,136 @@
+"""A busy Vertex is waited out, not reported as the end of the task.
+
+The 40-task eval run on google/gemini-3.5-flash lost inventory-reorder to one
+"429 Resource exhausted" in the middle of the task: the turn ended, the agent
+never wrote reorder.json, and the task scored ERROR. 429 and 5xx are about
+load, not the request, and the same request usually succeeds a few seconds
+later. The SDK only sends the request when the stream is first read, so the
+error arrives from inside the read loop.
+"""
+
+from __future__ import annotations
+
+import asyncio
+from types import SimpleNamespace
+
+import pytest
+
+pytest.importorskip("google.genai")
+from google.genai.errors import APIError  # noqa: E402
+
+from aria_code.apps.cli.providers import vertexai_stream  # noqa: E402
+from aria_code.apps.cli.providers.vertexai_stream import VertexAIProvider  # noqa: E402
+
+
+def _error(code: int, status: str) -> APIError:
+    return APIError(code, {"error": {"code": code, "message": status.lower(), "status": status}})
+
+
+def _chunk(text=None, reason="STOP"):
+    parts = [SimpleNamespace(text=text, thought=False)] if text else []
+    candidate = SimpleNamespace(content=SimpleNamespace(parts=parts),
+                                finish_reason=SimpleNamespace(name=reason))
+    return SimpleNamespace(candidates=[candidate], function_calls=None, usage_metadata=None)
+
+
+class _Client:
+    """Each round is a list: chunks to stream, or an APIError raised where the SDK raises it."""
+
+    def __init__(self, rounds, *, raise_at_call=False):
+        self.rounds = list(rounds)
+        self.calls = 0
+        outer = self
+
+        class _Models:
+            async def generate_content_stream(self, **kwargs):
+                outer.calls += 1
+                round_ = outer.rounds.pop(0)
+                if isinstance(round_, APIError) and raise_at_call:
+                    raise round_
+
+                async def gen():
+                    if isinstance(round_, APIError):
+                        raise round_
+                    for item in round_:
+                        if isinstance(item, APIError):
+                            raise item
+                        yield item
+                return gen()
+
+        self.aio = SimpleNamespace(models=_Models())
+
+
+@pytest.fixture
+def waits(monkeypatch):
+    slept: list[float] = []
+
+    async def fake_sleep(seconds):
+        slept.append(seconds)
+
+    monkeypatch.setattr(vertexai_stream, "_sleep", fake_sleep)
+    return slept
+
+
+def _run(client, cancel_event=None):
+    provider = VertexAIProvider(model="gemini-3.5-flash", config={})
+    provider._client = client
+    provider._get_client = lambda: client
+
+    async def collect():
+        return [e async for e in provider.stream([{"role": "user", "content": "go"}], tools=[],
+                                                 cancel_event=cancel_event)]
+
+    return asyncio.run(collect())
+
+
+def test_a_429_on_first_read_is_waited_out(waits):
+    client = _Client([_error(429, "RESOURCE_EXHAUSTED"), [_chunk("done")]])
+    events = _run(client)
+    assert client.calls == 2
+    assert events[-1].success and events[-1].response == "done"
+    assert waits == [2]
+
+
+def test_a_429_from_the_call_itself_too(waits):
+    client = _Client([_error(429, "RESOURCE_EXHAUSTED"), _error(503, "UNAVAILABLE"), [_chunk("ok")]],
+                     raise_at_call=True)
+    events = _run(client)
+    assert client.calls == 3 and events[-1].success
+    assert waits == [2, 5]
+
+
+def test_busy_retries_do_not_use_up_the_empty_round_retries(waits):
+    rounds = [_error(429, "RESOURCE_EXHAUSTED")] * 4 + [[_chunk("ok")]]
+    events = _run(_Client(rounds))
+    assert events[-1].success and events[-1].response == "ok"
+
+
+def test_it_gives_up_after_the_last_wait_and_says_why(waits):
+    client = _Client([_error(429, "RESOURCE_EXHAUSTED")] * 6)
+    done = _run(client)[-1]
+    assert client.calls == 6
+    assert waits == [2, 5, 10, 20, 40]
+    assert not done.success and "Vertex AI API Error" in done.error
+
+
+def test_a_bad_request_is_not_retried(waits):
+    client = _Client([_error(400, "INVALID_ARGUMENT"), [_chunk("never")]])
+    done = _run(client)[-1]
+    assert client.calls == 1 and waits == []
+    assert not done.success
+
+
+def test_nothing_is_sent_twice_once_text_has_streamed(waits):
+    client = _Client([[_chunk("half of it "), _error(503, "UNAVAILABLE")], [_chunk("again")]])
+    events = _run(client)
+    assert client.calls == 1 and waits == []
+    assert [e.text for e in events if hasattr(e, "text") and not hasattr(e, "success")] == ["half of it "]
+    assert not events[-1].success
+
+
+def test_the_user_can_cancel_during_the_wait(waits):
+    cancel = asyncio.Event()
+    cancel.set()
+    client = _Client([_error(429, "RESOURCE_EXHAUSTED"), [_chunk("never")]])
+    done = _run(client, cancel_event=cancel)[-1]
+    assert done.cancelled and client.calls == 1