sentiment-engine/sentiment_engine/tests/unit/test_output_sinks.py

"""Tests for output sinks"""

import pytest
from unittest.mock import AsyncMock, MagicMock, patch
from sentiment_engine.output.hazelcast_sink import HazelcastSink
from sentiment_engine.output.clickhouse_sink import ClickHouseSink
from sentiment_engine.output.manager import OutputManager
from sentiment_engine.schemas.output import SentimentOutput, AssetSentiment, MarketSentiment, IndustrySentiment, PumpDumpScore, VelocityMetrics, EventFlag


class TestHazelcastSink:
    """Tests for HazelcastSink"""

    @pytest.fixture
    def sink(self):
        return HazelcastSink()

    @pytest.mark.asyncio
    async def test_publish_scores(self, sink):
        """Test publishing scores to Hazelcast"""
        # This would require a running Hazelcast instance
        # For now, just verify the sink can be instantiated
        assert sink is not None

    def test_prepare_exf_data(self, sink):
        """Test preparing ExF data structure"""
        from sentiment_engine.schemas.output import SentimentOutput, MarketSentiment
        from datetime import datetime
        
        market = MarketSentiment(
            fear_state=25.0,
            greed_state=75.0,
            sentiment_index=50.0,
            hype_velocity=65.0,
            pub_velocity=55.0,
            aggregate_pump_risk=75.0,
            aggregate_dump_risk=20.0,
            top_pump_assets=["BTC", "ETH"],
            top_dump_assets=[],
            dominant_events=[],
            industry_breakdown={},
            last_update_ts=1234567890.0,
            total_sources=10,
            total_assets=50
        )
        output = SentimentOutput(
            timestamp=1234567890.0,
            market=market,
            assets={}
        )
        
        # Verify ACB signals extraction
        acb = output.get_acb_signals()
        assert "market_sentiment_state" in acb
        assert "aggregate_pump_risk" in acb
        assert "fear_state" in acb
        assert "greed_state" in acb
        assert "hype_velocity" in acb


class TestClickHouseSink:
    """Tests for ClickHouseSink"""

    @pytest.fixture
    def sink(self):
        return ClickHouseSink()

    def test_buffer_raw_item(self, sink):
        """Test buffering raw item for batch insert"""
        from sentiment_engine.schemas.payload import NormalizedPayload, SourceType, AssetMention, EngagementMetrics
        from datetime import datetime

        payload = NormalizedPayload(
            source_id="test",
            source_type=SourceType.NEWS,
            source_credibility_base=0.8,
            ingest_ts=1234567890.0,
            publish_ts=1234567880.0,
            asset_mentions=[AssetMention(asset_id="BTC", mention_span=(0, 3), confidence=0.9, source_text="BTC", mention_type="ticker")],
            raw_text="Test article",
            title="Test",
            url="https://test.com",
            author="Test",
            content_length=100,
            language="en",
            metadata={}
        )

        # This should not raise an error
        sink.buffer_raw_item(payload)
        assert len(sink._batch_buffer) == 1

    def test_buffer_processed_item(self, sink):
        """Test buffering processed item for batch insert"""
        from sentiment_engine.schemas.processed import ProcessedItem, EntityExtraction, SentimentScores, EmotionScores, EventClassification, EventType, TemporalAnchor, CredibilityScore
        from datetime import datetime

        item = ProcessedItem(
            payload_id="test_1",
            source_id="test",
            source_type="news",
            ingest_ts=1234567890.0,
            publish_ts=1234567880.0,
            entities=[EntityExtraction(asset_id="BTC", mention_span=(0, 3), confidence=0.9, entity_type="ticker", canonical_name="BTC")],
            sentiment_per_asset={"BTC": SentimentScores(polarity=0.7, confidence=0.85, positive_prob=0.8, negative_prob=0.1, neutral_prob=0.1)},
            emotions_per_asset={"BTC": EmotionScores(joy=0.8, fear=0.1, anger=0.05, greed=0.7, sadness=0.05, intensity=0.75)},
            events=[EventClassification(event_type=EventType.LISTING, confidence=0.7, assets_involved=["BTC"], key_details={}, severity=0.5)],
            temporal=TemporalAnchor(event_time=None, time_horizon="immediate", is_breaking=True, is_scheduled=False),
            credibility=CredibilityScore(source_base=0.8, content_quality=0.8, engagement_authenticity=0.7, cross_source_corroboration=0.6, historical_accuracy=0.8, composite=0.78),
            processed_ts=1234567895.0,
            processing_latency_ms=45.2,
            model_versions={}
        )

        sink.buffer_processed_item(item)
        assert len(sink._batch_buffer) == 1

    def test_buffer_score_output(self, sink):
        """Test buffering scored output"""
        from sentiment_engine.schemas.output import SentimentOutput, MarketSentiment, AssetSentiment, PumpDumpScore, VelocityMetrics
        import time

        market = MarketSentiment(
            fear_state=25.0, greed_state=75.0, sentiment_index=50.0,
            hype_velocity=65.0, pub_velocity=55.0,
            aggregate_pump_risk=75.0, aggregate_dump_risk=20.0,
            top_pump_assets=["BTC", "ETH"],
            top_dump_assets=[],
            dominant_events=[],
            industry_breakdown={},
            last_update_ts=time.time(),
            total_sources=10,
            total_assets=50
        )

        asset = AssetSentiment(
            asset_id="BTC",
            fear_state=20.0,
            greed_state=80.0,
            sentiment_polarity=60.0,
            pump_dump=PumpDumpScore(asset_id="BTC", pump_score=75.0, dump_score=15.0, pump_confidence=0.8, dump_confidence=0.7, last_update_ts=time.time()),
            last_update_ts=time.time()
        )

        output = SentimentOutput(
            timestamp=time.time(),
            market=market,
            assets={"BTC": asset}
        )

        sink.buffer_score_output(output)
        assert len(sink._batch_buffer) >= 2  # asset + market