Buckets:

download
raw
3.39 kB
#!/usr/bin/env python3
"""Demonstrate smart context functionality."""
from chunker import (
HybridContextStrategy,
RelevanceContextStrategy,
TreeSitterSmartContextProvider,
chunk_file,
)
def main():
"""Run smart context demo."""
# Get chunks from a Python file
chunks = chunk_file("examples/example.py", "python")
if not chunks:
print("No chunks found!")
return
# Create smart context provider
provider = TreeSitterSmartContextProvider()
# Create context strategies
relevance_strategy = RelevanceContextStrategy()
hybrid_strategy = HybridContextStrategy()
# Demo 1: Get semantic context for a chunk
print("=== SEMANTIC CONTEXT DEMO ===")
target_chunk = chunks[0]
print(f"\nTarget chunk: {target_chunk.node_type} at line {target_chunk.start_line}")
print(f"Content preview: {target_chunk.content[:100]}...")
context_str, metadata = provider.get_semantic_context(target_chunk, max_tokens=500)
print(f"\nSemantic context (relevance: {metadata.relevance_score:.2f}):")
print(context_str[:300] + "..." if len(context_str) > 300 else context_str)
# Demo 2: Get dependencies
print("\n\n=== DEPENDENCY CONTEXT DEMO ===")
dependencies = provider.get_dependency_context(target_chunk, chunks)
print(f"\nFound {len(dependencies)} dependencies:")
for dep_chunk, dep_meta in dependencies[:3]: # Show top 3
print(
f" - {dep_chunk.node_type} at line {dep_chunk.start_line} "
f"(relevance: {dep_meta.relevance_score:.2f})",
)
# Demo 3: Get usages
print("\n\n=== USAGE CONTEXT DEMO ===")
usages = provider.get_usage_context(target_chunk, chunks)
print(f"\nFound {len(usages)} usages:")
for usage_chunk, usage_meta in usages[:3]: # Show top 3
print(
f" - {usage_chunk.node_type} at line {usage_chunk.start_line} "
f"(relevance: {usage_meta.relevance_score:.2f})",
)
# Demo 4: Get structural context
print("\n\n=== STRUCTURAL CONTEXT DEMO ===")
structural = provider.get_structural_context(target_chunk, chunks)
print(f"\nFound {len(structural)} structural relations:")
for struct_chunk, struct_meta in structural[:3]: # Show top 3
print(
f" - {struct_chunk.node_type} at line {struct_chunk.start_line} "
f"(distance: {struct_meta.distance} lines)",
)
# Demo 5: Use context strategy to select most relevant chunks
print("\n\n=== CONTEXT SELECTION DEMO ===")
all_candidates = dependencies + usages + structural
# Use relevance strategy
selected_by_relevance = relevance_strategy.select_context(
target_chunk,
all_candidates,
max_tokens=1000,
)
print(f"\nRelevance strategy selected {len(selected_by_relevance)} chunks")
# Use hybrid strategy
selected_by_hybrid = hybrid_strategy.select_context(
target_chunk,
all_candidates,
max_tokens=1000,
)
print(f"Hybrid strategy selected {len(selected_by_hybrid)} chunks")
# Show ranking
ranked = relevance_strategy.rank_candidates(target_chunk, all_candidates)
print("\nTop 5 ranked chunks by relevance:")
for chunk, score in ranked[:5]:
print(f" - {chunk.node_type} at line {chunk.start_line} (score: {score:.3f})")
if __name__ == "__main__":
main()

Xet Storage Details

Size:
3.39 kB
·
Xet hash:
7ea78d9faac49b8a08b57b90e15a8cb19c9a882891d1ce720d232450d27d3a26

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.