retrieval: add unit tests for pure components

surfsense_backend/tests/unit/agents/multi_agent_chat/shared/retrieval/test_adapter.py

@@ -0,0 +1,50 @@
+"""Tests for mapping a DocumentHit to a renderable RetrievedDocument."""
+
+from __future__ import annotations
+
+import pytest
+
+from app.agents.chat.multi_agent_chat.shared.retrieval.adapter import (
+    to_retrieved_document,
+)
+from app.agents.chat.multi_agent_chat.shared.retrieval.models import (
+    ChunkHit,
+    DocumentHit,
+)
+
+pytestmark = pytest.mark.unit
+
+
+def test_maps_identity_source_label_and_passages() -> None:
+    hit = DocumentHit(
+        document_id=42,
+        title="Q3 Launch Notes",
+        document_type="SLACK_CONNECTOR",
+        metadata={},
+        score=0.9,
+        chunks=[
+            ChunkHit(chunk_id=880, content="a", position=4, score=0.9),
+            ChunkHit(chunk_id=881, content="b", position=7, score=0.5),
+        ],
+    )
+
+    document = to_retrieved_document(hit)
+
+    assert document.document_id == 42
+    assert document.title == "Q3 Launch Notes"
+    assert document.source_label == "Slack"
+    assert [(p.chunk_id, p.content) for p in document.passages] == [(880, "a"), (881, "b")]
+    assert all(p.document_id == 42 for p in document.passages)
+
+
+def test_document_with_no_chunks_maps_to_no_passages() -> None:
+    hit = DocumentHit(
+        document_id=1,
+        title="Empty",
+        document_type=None,
+        metadata={},
+        score=0.0,
+        chunks=[],
+    )
+
+    assert to_retrieved_document(hit).passages == []

surfsense_backend/tests/unit/agents/multi_agent_chat/shared/retrieval/test_service.py

@@ -0,0 +1,65 @@
+"""Tests for the build_context pipeline (rerank → adapt → render)."""
+
+from __future__ import annotations
+
+from typing import Any
+
+import pytest
+
+from app.agents.chat.multi_agent_chat.shared.citations import CitationRegistry
+from app.agents.chat.multi_agent_chat.shared.retrieval.models import (
+    ChunkHit,
+    DocumentHit,
+)
+from app.agents.chat.multi_agent_chat.shared.retrieval.service import build_context
+
+pytestmark = pytest.mark.unit
+
+
+def _hit(document_id: int, chunk_id: int) -> DocumentHit:
+    return DocumentHit(
+        document_id=document_id,
+        title=f"Doc {document_id}",
+        document_type="FILE",
+        metadata={},
+        score=1.0 / document_id,
+        chunks=[ChunkHit(chunk_id=chunk_id, content=f"text {chunk_id}", position=0, score=1.0)],
+    )
+
+
+def test_no_hits_renders_nothing() -> None:
+    assert build_context("q", [], CitationRegistry()) is None
+
+
+def test_renders_block_and_registers_labels_in_order() -> None:
+    registry = CitationRegistry()
+
+    block = build_context("q", [_hit(1, 880), _hit(2, 12)], registry)
+
+    assert block is not None
+    assert "[1] text 880" in block
+    assert "[2] text 12" in block
+    assert registry.resolve(1).locator == {"document_id": 1, "chunk_id": 880}
+    assert registry.resolve(2).locator == {"document_id": 2, "chunk_id": 12}
+
+
+class _ReverseReranker:
+    """Stand-in reranker that simply reverses document order."""
+
+    def rerank_documents(
+        self, query_text: str, documents: list[dict[str, Any]]
+    ) -> list[dict[str, Any]]:
+        return list(reversed(documents))
+
+
+def test_reranker_reorders_documents_before_labeling() -> None:
+    registry = CitationRegistry()
+
+    block = build_context(
+        "q", [_hit(1, 880), _hit(2, 12)], registry, reranker=_ReverseReranker()
+    )
+
+    assert block is not None
+    # Reversed: doc 2 now renders first and gets [1].
+    assert registry.resolve(1).locator == {"document_id": 2, "chunk_id": 12}
+    assert registry.resolve(2).locator == {"document_id": 1, "chunk_id": 880}

surfsense_backend/tests/unit/agents/multi_agent_chat/shared/retrieval/test_source_label.py

@@ -0,0 +1,35 @@
+"""Tests for building a document's source label."""
+
+from __future__ import annotations
+
+import pytest
+
+from app.agents.chat.multi_agent_chat.shared.retrieval.source_label import source_label
+
+pytestmark = pytest.mark.unit
+
+
+def test_known_type_uses_friendly_name() -> None:
+    assert source_label("SLACK_CONNECTOR", {}) == "Slack"
+
+
+def test_unmapped_type_is_prettified() -> None:
+    assert source_label("GOOGLE_DRIVE_FILE", {}) == "Google Drive"
+
+
+def test_url_host_is_appended_and_www_stripped() -> None:
+    label = source_label("CRAWLED_URL", {"url": "https://www.docs.python.org/3/"})
+
+    assert label == "Web · docs.python.org"
+
+
+def test_host_only_when_type_unknown() -> None:
+    assert source_label(None, {"url": "https://example.com/a"}) == "example.com"
+
+
+def test_returns_none_when_nothing_known() -> None:
+    assert source_label(None, {}) is None
+
+
+def test_non_http_url_is_ignored() -> None:
+    assert source_label("FILE", {"url": "/local/path"}) == "File"