Spaces:
Running
Running
File size: 7,524 Bytes
553537a 5c3b4a6 a11d742 628d40e a9d5552 a1c1173 876b12f a9d5552 a1c1173 876b12f 553537a a11d742 a9d5552 628d40e a9d5552 628d40e 5c3b4a6 a1c1173 5c3b4a6 a1c1173 5c3b4a6 a1c1173 553537a a1c1173 5c3b4a6 a1c1173 876b12f 5c3b4a6 628d40e 5c3b4a6 628d40e a1c1173 a11d742 a1c1173 5c3b4a6 a1c1173 9a2420b a11d742 876b12f a1c1173 628d40e a11d742 5c3b4a6 a11d742 55cdb25 a11d742 55cdb25 a11d742 55cdb25 a11d742 a1c1173 5c3b4a6 a1c1173 628d40e a1c1173 876b12f a1c1173 a11d742 a1c1173 a11d742 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 |
from fastapi import APIRouter, HTTPException
from pydantic import BaseModel, HttpUrl
from typing import Dict, Any, List, Literal
import logging
import os
from supabase import AsyncClient
from dotenv import load_dotenv
from mediaunmasked.scrapers.article_scraper import ArticleScraper
from mediaunmasked.analyzers.scoring import MediaScorer
from mediaunmasked.utils.logging_config import setup_logging
# Load environment variables
load_dotenv()
# Initialize logging
setup_logging()
logger = logging.getLogger(__name__)
# Initialize router and dependencies
router = APIRouter(tags=["analysis"])
scraper = ArticleScraper()
# Get Supabase credentials
SUPABASE_URL = os.getenv("SUPABASE_URL")
SUPABASE_KEY = os.getenv("SUPABASE_KEY")
# Initialize Supabase client
if not SUPABASE_URL or not SUPABASE_KEY:
raise Exception("Supabase credentials not found in environment variables")
supabase = AsyncClient(SUPABASE_URL, SUPABASE_KEY)
# Define analysis mode type
AnalysisMode = Literal['ai', 'traditional']
class ArticleRequest(BaseModel):
url: HttpUrl
use_ai: bool = True # Default to AI-powered analysis
class MediaScoreDetails(BaseModel):
headline_analysis: Dict[str, Any]
sentiment_analysis: Dict[str, Any]
bias_analysis: Dict[str, Any]
evidence_analysis: Dict[str, Any]
class MediaScore(BaseModel):
media_unmasked_score: float
rating: str
details: MediaScoreDetails
class AnalysisResponse(BaseModel):
headline: str
content: str
sentiment: str
bias: str
bias_score: float
bias_percentage: float
media_score: MediaScore
analysis_mode: AnalysisMode
@router.post("/analyze", response_model=AnalysisResponse)
async def analyze_article(request: ArticleRequest) -> AnalysisResponse:
"""
Analyze an article for bias, sentiment, and credibility.
Args:
request: ArticleRequest containing the URL to analyze and analysis preferences
Returns:
AnalysisResponse with complete analysis results
Raises:
HTTPException: If scraping or analysis fails
"""
try:
# Determine analysis mode
analysis_mode: AnalysisMode = 'ai' if request.use_ai else 'traditional'
logger.info(f"Analyzing article: {request.url} (Analysis Mode: {analysis_mode})")
# Check cache with both URL and analysis mode
try:
cached_result = await supabase.table('article_analysis') \
.select('*') \
.eq('url', str(request.url)) \
.eq('analysis_mode', analysis_mode) \
.limit(1) \
.single() \
.execute()
if cached_result and cached_result.data:
logger.info(f"Found cached analysis for URL with {analysis_mode} mode")
return AnalysisResponse.parse_obj(cached_result.data)
except Exception as cache_error:
logger.warning(f"Cache lookup failed: {str(cache_error)}")
# Continue with analysis if cache lookup fails
# Scrape article
article = scraper.scrape_article(str(request.url))
if not article:
raise HTTPException(
status_code=400,
detail="Failed to scrape article content"
)
# Initialize scorer with specified analysis preference
scorer = MediaScorer(use_ai=request.use_ai)
# Analyze content
analysis = scorer.calculate_media_score(
article["headline"],
article["content"]
)
# Log raw values for debugging
logger.info("Raw values:")
logger.info(f"media_unmasked_score type: {type(analysis['media_unmasked_score'])}")
logger.info(f"media_unmasked_score value: {analysis['media_unmasked_score']}")
# Prepare response data
response_dict = {
"headline": str(article['headline']),
"content": str(article['content']),
"sentiment": str(analysis['details']['sentiment_analysis']['sentiment']),
"bias": str(analysis['details']['bias_analysis']['bias']),
"bias_score": float(analysis['details']['bias_analysis']['bias_score']),
"bias_percentage": float(analysis['details']['bias_analysis']['bias_percentage']),
"analysis_mode": analysis_mode,
"media_score": {
"media_unmasked_score": float(analysis['media_unmasked_score']),
"rating": str(analysis['rating']),
"details": {
"headline_analysis": {
"headline_vs_content_score": float(analysis['details']['headline_analysis']['headline_vs_content_score']),
"flagged_phrases": analysis['details']['headline_analysis'].get('flagged_phrases', [])
},
"sentiment_analysis": {
"sentiment": str(analysis['details']['sentiment_analysis']['sentiment']),
"manipulation_score": float(analysis['details']['sentiment_analysis']['manipulation_score']),
"flagged_phrases": list(analysis['details']['sentiment_analysis']['flagged_phrases'])
},
"bias_analysis": {
"bias": str(analysis['details']['bias_analysis']['bias']),
"bias_score": float(analysis['details']['bias_analysis']['bias_score']),
"bias_percentage": float(analysis['details']['bias_analysis']['bias_percentage']),
"flagged_phrases": list(analysis['details']['bias_analysis']['flagged_phrases'])
},
"evidence_analysis": {
"evidence_based_score": float(analysis['details']['evidence_analysis']['evidence_based_score']),
"flagged_phrases": list(analysis['details']['evidence_analysis']['flagged_phrases'])
}
}
}
}
# Save to Supabase with analysis mode
try:
await supabase.table('article_analysis').upsert({
'url': str(request.url),
'headline': response_dict['headline'],
'content': response_dict['content'],
'sentiment': response_dict['sentiment'],
'bias': response_dict['bias'],
'bias_score': response_dict['bias_score'],
'bias_percentage': response_dict['bias_percentage'],
'media_score': response_dict['media_score'],
'analysis_mode': analysis_mode, # Store the analysis mode
'created_at': 'now()' # Use server timestamp
}, on_conflict='url,analysis_mode').execute() # Specify composite unique constraint
logger.info(f"Saved analysis to database with mode: {analysis_mode}")
except Exception as db_error:
logger.error(f"Failed to save to database: {str(db_error)}")
# Continue since we can still return the analysis even if saving fails
# Return the response
return AnalysisResponse.parse_obj(response_dict)
except Exception as e:
logger.error(f"Analysis failed: {str(e)}", exc_info=True)
raise HTTPException(
status_code=500,
detail=f"Analysis failed: {str(e)}"
)
|