Celes Renata
30 lines
1.0 KiB
Python
30 lines
1.0 KiB
Python
import asyncio, asyncpg, json, os, redis
|
|
|
|
async def main():
|
|
pool = await asyncpg.create_pool(
|
|
host=os.environ["POSTGRES_HOST"],
|
|
port=int(os.environ["POSTGRES_PORT"]),
|
|
database=os.environ["POSTGRES_DB"],
|
|
user=os.environ["POSTGRES_USER"],
|
|
password=os.environ["POSTGRES_PASSWORD"],
|
|
)
|
|
r = redis.from_url(f"redis://:{os.environ.get('REDIS_PASSWORD', '')}@{os.environ['REDIS_HOST']}:{os.environ['REDIS_PORT']}/0")
|
|
|
|
rows = await pool.fetch(
|
|
"SELECT d.id, dcm.ticker FROM documents d "
|
|
"LEFT JOIN document_company_mentions dcm ON d.id = dcm.document_id "
|
|
"WHERE d.source_type = 'news_api' AND d.parse_quality_score > 0.7 "
|
|
"ORDER BY d.parse_quality_score DESC LIMIT 5"
|
|
)
|
|
|
|
for row in rows:
|
|
r.rpush("stonks:queue:extraction", json.dumps({
|
|
"document_id": str(row["id"]),
|
|
"ticker": row["ticker"] or "",
|
|
}))
|
|
|
|
print(f"Enqueued {len(rows)} high-quality docs for re-extraction")
|
|
await pool.close()
|
|
|
|
asyncio.run(main())
|