curllm

Orchestration Examples

Advanced extraction orchestration with multiple strategies.

Contents

File	Description
orchestrator_example.py	Multi-strategy extraction orchestrator
semantic_query_example.py	Semantic queries with LLM understanding

Orchestration Pipeline

1. Quick Page Check
   ↓ (price_count, product_links, page_type)
2. LLM-Guided Extraction
   ↓ (container selector, field detection)
3. Dynamic Detection (fallback)
   ↓ (statistical + LLM hybrid)
4. Iterative Extractor (fallback)
   ↓ (atomic DOM queries)
5. Multi-Criteria Filtering
   ↓ (price filters, keyword matching)
6. Result Validation

Usage

python orchestrator_example.py
python semantic_query_example.py

Example: Custom Orchestration

from curllm_core.extraction_orchestrator import ExtractionOrchestrator

async def extract_products(page, llm, instruction):
    orchestrator = ExtractionOrchestrator(page, llm, instruction, logger)
    
    # Try multiple strategies in order
    result = await orchestrator.extract()
    
    return result['products']

This site is open source. Improve this page.