refactor(tests): reorganize tests into unittest/ and e2e/ directories

- Move all unit tests from tests/ to tests/unittest/ - Add tests/e2e/ directory for end-to-end tests - Update conftest.py for new test structure - Add new tests for auth dependencies, policies, route injection - Add new tests for run callbacks, create store, execution artifacts - Remove obsolete tests for deleted persistence layer Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2026-05-21 07:26:50 +00:00 · 2026-04-22 11:24:53 +08:00
parent 38a6ec496f
commit 2fe0856e33
149 changed files with 3450 additions and 4664 deletions
@@ -0,0 +1,55 @@
+"""Tests for readability extraction fallback behavior."""
+
+import subprocess
+
+import pytest
+
+from deerflow.utils.readability import ReadabilityExtractor
+
+
+def test_extract_article_falls_back_when_readability_js_fails(monkeypatch):
+    """When Node-based readability fails, extraction should fall back to Python mode."""
+
+    calls: list[bool] = []
+
+    def _fake_simple_json_from_html_string(html: str, use_readability: bool = False):
+        calls.append(use_readability)
+        if use_readability:
+            raise subprocess.CalledProcessError(
+                returncode=1,
+                cmd=["node", "ExtractArticle.js"],
+                stderr="boom",
+            )
+        return {"title": "Fallback Title", "content": "<p>Fallback Content</p>"}
+
+    monkeypatch.setattr(
+        "deerflow.utils.readability.simple_json_from_html_string",
+        _fake_simple_json_from_html_string,
+    )
+
+    article = ReadabilityExtractor().extract_article("<html><body>test</body></html>")
+
+    assert calls == [True, False]
+    assert article.title == "Fallback Title"
+    assert article.html_content == "<p>Fallback Content</p>"
+
+
+def test_extract_article_re_raises_unexpected_exception(monkeypatch):
+    """Unexpected errors should be surfaced instead of silently falling back."""
+
+    calls: list[bool] = []
+
+    def _fake_simple_json_from_html_string(html: str, use_readability: bool = False):
+        calls.append(use_readability)
+        if use_readability:
+            raise RuntimeError("unexpected parser failure")
+        return {"title": "Should Not Reach Fallback", "content": "<p>Fallback</p>"}
+
+    monkeypatch.setattr(
+        "deerflow.utils.readability.simple_json_from_html_string",
+        _fake_simple_json_from_html_string,
+    )
+
+    with pytest.raises(RuntimeError, match="unexpected parser failure"):
+        ReadabilityExtractor().extract_article("<html><body>test</body></html>")
+    assert calls == [True]