Spaces:
Running on Zero
Running on Zero
File size: 8,570 Bytes
0828c2c | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 | """Script to build indexes for retrieval strategies."""
import logging
import sys
from pathlib import Path
from .config_loader import get_config
from .document_processor import process_documents
# Import retrieval components (this registers strategies with the factory)
from .retrieval import RetrieverFactory
from .retrieval.strategies import BM25Strategy, BM25VectorStrategy, VectorStrategy # noqa: F401
logging.basicConfig(
level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
)
logger = logging.getLogger(__name__)
def build_index(
documents_dir: str | None = None,
strategy: str | None = None,
force_rebuild: bool = False,
):
"""Build index for the specified retrieval strategy.
Args:
documents_dir: Path to documents directory (uses config default if None)
strategy: Retrieval strategy to use (uses config default if None)
force_rebuild: If True, delete existing indexes before building
"""
logger.info("=" * 60)
logger.info("Building Index for Profile Chatbot")
logger.info("=" * 60)
# Load configuration
config = get_config()
# Get strategy from config if not provided
if strategy is None:
strategy = config.get("retrieval.strategy", "vector")
logger.info(f"Retrieval Strategy: {strategy}")
# Get documents directory
if documents_dir is None:
documents_dir = config.get_env("DOCUMENTS_DIR", "./data/documents")
# Log main document info
main_doc_enabled = config.get("main_document.enabled", False)
main_doc_path = config.get("main_document.path", "")
exclude_from_index = config.get("main_document.exclude_from_index", True)
if main_doc_enabled and main_doc_path:
logger.info("Main Document Feature: ENABLED")
if exclude_from_index:
logger.info(
f" Main document ({Path(main_doc_path).name}) will be excluded from index"
)
logger.info(" (It's loaded directly into prompts, not retrieved)")
else:
logger.info(
f" Main document ({Path(main_doc_path).name}) will also be indexed "
"(exclude_from_index=false)"
)
logger.info("")
documents_path = Path(documents_dir)
if not documents_path.exists():
logger.error(f"Documents directory not found: {documents_dir}")
logger.error("Please create the directory and add your profile documents.")
sys.exit(1)
# Check if directory has files
files = list(documents_path.rglob("*"))
if not files or all(f.is_dir() for f in files):
logger.warning(f"No files found in {documents_dir}")
logger.warning("Please add your PDF, Word, HTML, or text documents to this directory.")
sys.exit(1)
# Process documents
logger.info(f"Processing documents from: {documents_dir}")
documents = process_documents(documents_dir)
if not documents:
logger.error("No documents were processed. Check your files and try again.")
sys.exit(1)
logger.info(f"Successfully processed {len(documents)} document chunks")
# Handle force rebuild for BM25 index
if force_rebuild and strategy in ["bm25", "bm25_vector"]:
bm25_path = Path(config.get("retrieval.bm25.persist_path", "./bm25_index"))
if bm25_path.exists():
import shutil
logger.warning(f"Force rebuild - deleting BM25 index at {bm25_path}")
shutil.rmtree(bm25_path)
# Handle force rebuild for vector store
if force_rebuild and strategy in ["vector", "bm25_vector"]:
try:
from .vectorstore import get_vectorstore_manager
vectorstore_manager = get_vectorstore_manager()
logger.warning("Force rebuild - deleting vector store...")
vectorstore_manager.delete_collection()
except Exception as e:
logger.info(f"No existing vector collection to delete: {e}")
# Create retrieval strategy
logger.info(f"Creating retrieval strategy: {strategy}")
try:
# Get the full config as dict for the strategy
config_dict = {
"retrieval": {
"strategy": strategy,
"final_k": config.get("retrieval.final_k", 4),
"vector": {
"search_type": config.get("retrieval.vector.search_type", "similarity"),
"k": config.get("retrieval.vector.k", 10),
"search_kwargs": config.get("retrieval.vector.search_kwargs", {}),
},
"bm25": {
"k": config.get("retrieval.bm25.k", 10),
"persist_path": config.get("retrieval.bm25.persist_path", "./bm25_index"),
"tokenizer": config.get("retrieval.bm25.tokenizer", "simple"),
},
"fusion": {
"algorithm": config.get("retrieval.fusion.algorithm", "rrf"),
"rrf_k": config.get("retrieval.fusion.rrf_k", 60),
"weights": config.get("retrieval.fusion.weights", {"vector": 0.7, "bm25": 0.3}),
},
}
}
retrieval_strategy = RetrieverFactory.create(strategy, config_dict)
# Build the index
logger.info("Building index (this may take a few minutes)...")
retrieval_strategy.build_index(documents)
# Get stats
stats = retrieval_strategy.get_index_stats()
logger.info("=" * 60)
logger.info("✅ Index built successfully!")
logger.info(f" Strategy: {strategy}")
logger.info(f" Documents: {len(documents)} chunks")
if strategy == "vector":
logger.info(f" Vector Store: {stats.get('persist_directory', 'N/A')}")
elif strategy == "bm25":
logger.info(f" BM25 Index: {stats.get('persist_path', 'N/A')}")
elif strategy == "bm25_vector":
vector_stats = stats.get("vector", {})
bm25_stats = stats.get("bm25", {})
logger.info(f" Vector Store: {vector_stats.get('persist_directory', 'N/A')}")
logger.info(f" BM25 Index: {bm25_stats.get('persist_path', 'N/A')}")
fusion_stats = stats.get("fusion", {})
logger.info(
f" Fusion: {fusion_stats.get('algorithm', 'rrf')} "
f"(vector={fusion_stats.get('weights', {}).get('vector', 0.7)}, "
f"bm25={fusion_stats.get('weights', {}).get('bm25', 0.3)})"
)
# Log main document status
if main_doc_enabled and main_doc_path:
logger.info(f" Main Doc: {Path(main_doc_path).name} (loaded directly)")
logger.info("=" * 60)
logger.info("\n🚀 You can now run the chatbot with: streamlit run app.py\n")
except Exception as e:
logger.error(f"Error building index: {e}")
import traceback
traceback.print_exc()
sys.exit(1)
def main():
"""Main entry point for CLI."""
import argparse
parser = argparse.ArgumentParser(
description="Build retrieval index from documents",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
Examples:
# Build with default strategy from config
python -m src.build_vectorstore
# Build vector-only index
python -m src.build_vectorstore --strategy vector
# Build hybrid BM25 + Vector index
python -m src.build_vectorstore --strategy bm25_vector
# Force rebuild from scratch
python -m src.build_vectorstore --strategy bm25_vector --force-rebuild
Available strategies:
vector - Semantic similarity search using embeddings
bm25 - Lexical/keyword search using BM25 algorithm
bm25_vector - Hybrid combining BM25 and vector search (default)
""",
)
parser.add_argument(
"--documents-dir",
type=str,
default=None,
help="Path to documents directory (default: from config)",
)
parser.add_argument(
"--strategy",
type=str,
default=None,
choices=["vector", "bm25", "bm25_vector"],
help="Retrieval strategy to use (default: from config)",
)
parser.add_argument(
"--force-rebuild",
action="store_true",
help="Delete existing indexes before building",
)
args = parser.parse_args()
build_index(
documents_dir=args.documents_dir,
strategy=args.strategy,
force_rebuild=args.force_rebuild,
)
if __name__ == "__main__":
main()
|