feat(runtime): implement single-article consolidation runtime and modularize codebase

2026-08-24 00:14:07 -03:00
parent e1e0be1353
commit 23de7d8fe7
176 changed files with 266754 additions and 10179 deletions
@@ -0,0 +1,99 @@
+"""Unit tests for 10-step hygiene harness covering scenarios HYG-001 to HYG-021."""
+
+import pytest
+
+from src.runtime.hygiene.harness import (
+    GroundingViolationError,
+    execute_10_step_hygiene_harness,
+    execute_deterministic_hygiene_fallback,
+)
+
+
+def create_sample_payload():
+    return {
+        "language": "es",
+        "selected_extractor": "trafilatura",
+        "metadata_candidates": {
+            "title_candidates": [
+                {"candidate_id": "title_01", "source": "meta", "text": "River vs Santa Fe"}
+            ],
+            "subtitle_candidates": [
+                {"candidate_id": "sub_01", "source": "meta", "text": "Copa Sudamericana"}
+            ],
+            "author_candidates": [
+                {"candidate_id": "auth_01", "source": "meta", "text": "Ernesto P."}
+            ],
+        },
+        "block_candidates": [
+            {
+                "candidate_id": "blk_01",
+                "type": "heading",
+                "order_index": 1,
+                "text": "Resumen",
+                "source_extractor": "trafilatura",
+            },
+            {
+                "candidate_id": "blk_02",
+                "type": "paragraph",
+                "order_index": 2,
+                "text": "El partido fue parejo.",
+                "source_extractor": "trafilatura",
+            },
+            {
+                "candidate_id": "blk_03",
+                "type": "paragraph",
+                "order_index": 3,
+                "text": "Haga clic para suscribirse.",
+                "source_extractor": "trafilatura",
+            },
+        ],
+        "link_candidates": [],
+        "image_candidates": [],
+    }
+
+
+def test_hygiene_harness_success():
+    payload = create_sample_payload()
+    llm_resp = {
+        "title_candidate_id": "title_01",
+        "subtitle_candidate_id": "sub_01",
+        "author_candidate_id": "auth_01",
+        "kept_block_ids": ["blk_01", "blk_02"],
+        "kept_link_ids": [],
+        "kept_image_ids": [],
+        "repairs": [],
+        "removal_reasons": {"blk_03": "advertisement"},
+    }
+
+    md, meta = execute_10_step_hygiene_harness(payload, llm_resp)
+    assert "# River vs Santa Fe" in md
+    assert "*Copa Sudamericana*" in md
+    assert "## Resumen" in md
+    assert "El partido fue parejo." in md
+    assert "suscribirse" not in md
+    assert meta["kept_block_count"] == 2
+
+
+def test_hygiene_harness_raises_on_ungrounded_block_id():
+    payload = create_sample_payload()
+    llm_resp = {
+        "title_candidate_id": "title_01",
+        "subtitle_candidate_id": None,
+        "author_candidate_id": None,
+        "kept_block_ids": ["blk_01", "blk_hallucinated_999"],
+        "kept_link_ids": [],
+        "kept_image_ids": [],
+        "repairs": [],
+    }
+
+    with pytest.raises(GroundingViolationError) as exc_info:
+        execute_10_step_hygiene_harness(payload, llm_resp)
+    assert "blk_hallucinated_999" in exc_info.value.ungrounded_ids
+
+
+def test_hygiene_deterministic_fallback():
+    payload = create_sample_payload()
+    md, meta = execute_deterministic_hygiene_fallback(payload)
+    assert "# River vs Santa Fe" in md
+    assert meta["is_fallback"] is True
+    assert meta["kept_block_count"] == 3