Improve tailoring and job fetch resilience

2026-08-03 00:57:19 +03:30
parent 81d816ae54
commit 50cc879e1e
6 changed files with 311 additions and 19 deletions
@@ -0,0 +1,108 @@
+from __future__ import annotations
+
+from typing import Any
+
+import httpx
+import pytest
+
+from resume_agent import web
+
+
+def _mock_client(
+    monkeypatch: pytest.MonkeyPatch,
+    handler: Any,
+    captured: dict[str, Any] | None = None,
+) -> None:
+    original_client = httpx.Client
+    transport = httpx.MockTransport(handler)
+
+    def client_factory(**kwargs: Any) -> httpx.Client:
+        if captured is not None:
+            captured.update(kwargs)
+        return original_client(transport=transport, **kwargs)
+
+    monkeypatch.setattr(web.httpx, "Client", client_factory)
+    monkeypatch.setattr(web, "_validate_public_url", lambda _url: None)
+
+
+def test_job_fetch_retries_transient_connection_timeout(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    attempts = 0
+
+    def handler(request: httpx.Request) -> httpx.Response:
+        nonlocal attempts
+        attempts += 1
+        if attempts == 1:
+            raise httpx.ConnectTimeout("timed out", request=request)
+        return httpx.Response(
+            200,
+            headers={"content-type": "text/html; charset=utf-8"},
+            text="<html><body><h1>Backend Engineer</h1></body></html>",
+        )
+
+    _mock_client(monkeypatch, handler)
+
+    result = web.fetch_job_page(
+        "https://jobs.example/role",
+        timeout_seconds=120,
+        retries=1,
+        retry_backoff_seconds=0,
+    )
+
+    assert result == "Backend Engineer"
+    assert attempts == 2
+
+
+def test_job_fetch_reports_timeout_as_job_page_error(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    attempts = 0
+
+    def handler(request: httpx.Request) -> httpx.Response:
+        nonlocal attempts
+        attempts += 1
+        raise httpx.ConnectTimeout("timed out", request=request)
+
+    _mock_client(monkeypatch, handler)
+
+    with pytest.raises(web.JobPageError, match="after 3 attempts"):
+        web.fetch_job_page(
+            "https://jobs.example/role",
+            timeout_seconds=120,
+            retries=2,
+            retry_backoff_seconds=0,
+        )
+
+    assert attempts == 3
+
+
+def test_job_fetch_timeout_is_configurable(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    captured: dict[str, Any] = {}
+
+    def handler(_request: httpx.Request) -> httpx.Response:
+        return httpx.Response(
+            200,
+            headers={"content-type": "text/plain"},
+            text="Complete job description",
+        )
+
+    _mock_client(monkeypatch, handler, captured)
+    monkeypatch.setenv("RESUME_AGENT_JOB_FETCH_TIMEOUT_SECONDS", "240")
+    monkeypatch.setenv("RESUME_AGENT_JOB_FETCH_RETRIES", "0")
+
+    assert web.fetch_job_page("https://jobs.example/role") == "Complete job description"
+    timeout = captured["timeout"]
+    assert timeout.connect == 60
+    assert timeout.read == 240
+
+
+def test_job_fetch_rejects_invalid_timeout_configuration(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    monkeypatch.setenv("RESUME_AGENT_JOB_FETCH_TIMEOUT_SECONDS", "not-a-number")
+
+    with pytest.raises(web.JobPageError, match="must be a number"):
+        web.fetch_job_page("https://jobs.example/role")