quickstart.rst

Quick Start Guide

Basic Usage

Indexing Documents

from src.indexer.inverted_index import InvertedIndex

# Créer un index
index = InvertedIndex()

# Ajouter des documents
index.add_document("doc1", ["python", "search", "engine"])
index.add_document("doc2", ["information", "retrieval", "system"])

Searching

# Recherche simple
results = index.search(["python", "engine"])
print(results)  # {'doc1': 0.857}

# Recherche de terme unique
docs = index.search_term("python")
print(docs)  # {'doc1'}

Using the CLI

# Indexer un dossier
python src/cli.py index --input data/sample_documents/ --output index.pkl

# Rechercher
python src/cli.py search --index index.pkl --query "python search"

# Statistiques
python src/cli.py stats --index index.pkl

Web API

# Lancer le serveur
python src/app.py

# Rechercher via API
curl "http://localhost:5000/search?q=python+search"

Compression

# Compresser les posting lists
posting_list = index.get_postings("python")
compressed = posting_list.get_doc_ids_compressed()

# Décompresser
doc_ids = posting_list.decompress_doc_ids(compressed)

Benchmarking

# Tester la vitesse d'indexation
python benchmarks/benchmark_indexing.py

# Tester la latence de requête
python benchmarks/benchmark_query.py

# Tester la compression
python benchmarks/benchmark_compression.py