File size: 7,188 Bytes
914e970 0db9f89 914e970 0db9f89 914e970 | 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 | """Document ingestion and retrieval for the DRIFT companion."""
import re
import uuid
from pathlib import Path
from typing import List, Optional
import chromadb
from infj_bot.core.config import PROJECT_ROOT, DATA_DIR
from infj_bot.core.embeddings import (
get_default_embedding_function,
LocalEmbeddingFunction,
)
SUPPORTED_TEXT = {
".txt",
".md",
".py",
".js",
".ts",
".jsx",
".tsx",
".json",
".yaml",
".yml",
".csv",
".sh",
".html",
".css",
".rs",
".go",
".java",
".c",
".cpp",
".h",
}
MAX_INGEST_FILE_BYTES = 2_000_000
MAX_DIRECTORY_FILES = 300
def _is_relative_to(child: Path, parent: Path) -> bool:
try:
child.relative_to(parent)
return True
except ValueError:
return False
def _resolve_ingest_path(path: str) -> Path:
target = Path(path).expanduser()
if not target.is_absolute():
target = PROJECT_ROOT / target
target = target.resolve()
allowed_roots = [PROJECT_ROOT.resolve(), Path.home().resolve()]
if not any(_is_relative_to(target, root) for root in allowed_roots):
raise PermissionError(f"Path {path} is outside the allowed ingestion roots.")
return target
def _chunk_text(text: str, chunk_size: int = 800, overlap: int = 100) -> List[str]:
"""Split text into overlapping chunks by paragraphs."""
paragraphs = [p.strip() for p in re.split(r"\n\s*\n", text) if p.strip()]
chunks = []
current: List[str] = []
current_len = 0
for para in paragraphs:
para_len = len(para)
if current_len + para_len > chunk_size and current:
chunks.append("\n\n".join(current))
# Keep overlap
overlap_text: List[str] = []
overlap_len = 0
for p in reversed(current):
if overlap_len + len(p) > overlap:
break
overlap_text.insert(0, p)
overlap_len += len(p)
current = overlap_text
current_len = overlap_len
current.append(para)
current_len += para_len
if current:
chunks.append("\n\n".join(current))
return chunks
def _read_pdf(path: Path) -> str:
try:
from pypdf import PdfReader
reader = PdfReader(str(path))
parts = []
for page in reader.pages:
text = page.extract_text()
if text:
parts.append(text)
return "\n\n".join(parts)
except Exception as exc:
raise RuntimeError(f"PDF read failed: {exc}")
def _read_file(path: Path) -> str:
if path.stat().st_size > MAX_INGEST_FILE_BYTES:
raise ValueError(
f"File too large for ingestion: {path} ({path.stat().st_size} bytes)"
)
suffix = path.suffix.lower()
if suffix == ".pdf":
return _read_pdf(path)
return path.read_text(encoding="utf-8", errors="replace")
class DocumentStore:
def __init__(
self, persist_directory=None, embedding_function=None, use_semantic=True
):
if persist_directory is None:
persist_directory = str(DATA_DIR / "chroma_db")
if embedding_function is None:
if use_semantic:
embedding_function = get_default_embedding_function()
else:
embedding_function = LocalEmbeddingFunction()
self.embedding_function = embedding_function
self.client = chromadb.PersistentClient(path=persist_directory)
self.collection = self.client.get_or_create_collection(
name="infj_documents",
embedding_function=embedding_function,
)
def ingest(self, file_path: str, tags: Optional[List[str]] = None) -> int:
path = _resolve_ingest_path(file_path)
if not path.exists():
raise FileNotFoundError(f"Not found: {path}")
if path.is_dir():
return self.ingest_directory(path, tags=tags)
text = _read_file(path)
if not text.strip():
return 0
chunks = _chunk_text(text)
if not chunks:
return 0
ids = [f"doc-{uuid.uuid4().hex[:12]}" for _ in chunks]
metadatas = []
for i, _chunk in enumerate(chunks):
meta = {
"source": str(path),
"filename": path.name,
"chunk_index": i,
"total_chunks": len(chunks),
"tags": ",".join(tags or []),
}
metadatas.append(meta)
self.collection.add(
documents=chunks,
ids=ids,
metadatas=metadatas,
)
return len(chunks)
def ingest_directory(
self, dir_path: Path, tags: Optional[List[str]] = None, recursive: bool = True
) -> int:
total = 0
scanned = 0
pattern = "**/*" if recursive else "*"
for child in Path(dir_path).glob(pattern):
if child.is_file() and child.suffix.lower() in SUPPORTED_TEXT | {".pdf"}:
scanned += 1
if scanned > MAX_DIRECTORY_FILES:
raise ValueError(
f"Directory ingestion stopped after {MAX_DIRECTORY_FILES} supported files."
)
try:
n = self.ingest(str(child), tags=tags)
total += n
except Exception as exc:
print(f"[ingest skip] {child}: {exc}")
return total
def search(self, query: str, n_results: int = 5) -> List[dict]:
results = self.collection.query(
query_texts=[query],
n_results=n_results,
)
out = []
for i, doc in enumerate(results["documents"][0]):
meta = results["metadatas"][0][i]
out.append(
{
"document": doc,
"source": meta.get("source", "?"),
"filename": meta.get("filename", "?"),
"chunk_index": meta.get("chunk_index", 0),
}
)
return out
def list_sources(self) -> List[str]:
results = self.collection.get(include=["metadatas"])
sources = set()
for meta in results.get("metadatas", []):
if meta:
sources.add(meta.get("source", "?"))
return sorted(sources)
def delete_source(self, source_path: str) -> int:
results = self.collection.get(
where={"source": source_path},
include=[],
)
ids = results.get("ids", [])
if ids:
self.collection.delete(ids=ids)
return len(ids)
def count(self) -> int:
return self.collection.count()
def format_doc_results(results: List[dict]) -> str:
if not results:
return "No matching documents found."
lines = []
for r in results:
lines.append(
f"[{r['filename']} chunk {r['chunk_index']}]\n{r['document'][:600]}"
)
return "\n---\n".join(lines)
if __name__ == "__main__":
store = DocumentStore()
print(f"Document store initialized. Documents: {store.count()}")
|