"""PDF reader — uses ``pypdf`` (lazy import).
Install with: pip install ractogateway[rag-pdf]
"""
from __future__ import annotations
import io
from pathlib import Path
from typing import Any
def _require_pypdf() -> Any:
try:
import pypdf
except ImportError as exc:
raise ImportError(
"Reading PDF files requires the 'pypdf' package. "
"Install it with: pip install ractogateway[rag-pdf]"
) from exc
return pypdf
from ractogateway.rag._models.document import Document
from ractogateway.rag.readers.base import BaseReader
[docs]
class PdfReader(BaseReader):
"""Extract text from PDF files using ``pypdf``.
Accepts a file path (``str`` / ``Path``), raw ``bytes``, or any binary
file-like object with a ``.read()`` method.
Parameters
----------
extract_images:
Reserved for future use — image extraction is not yet supported.
"""
def __init__(self, extract_images: bool = False) -> None:
self._extract_images = extract_images
@property
def supported_extensions(self) -> frozenset[str]:
return frozenset({".pdf"})
def _read_path(self, path: Path) -> Document:
pypdf = _require_pypdf()
pages_text: list[str] = []
page_map: list[tuple[int, str]] = []
with pypdf.PdfReader(str(path)) as reader:
total_pages = len(reader.pages)
for i, page in enumerate(reader.pages):
text = page.extract_text() or ""
pages_text.append(text)
page_map.append((i + 1, text))
full_text = "\n\n".join(t for t in pages_text if t.strip())
return Document(
content=full_text,
source=str(path.resolve()),
metadata={
"extension": ".pdf",
"filename": path.name,
"size_bytes": path.stat().st_size,
"total_pages": total_pages,
"page_map": page_map,
},
)
def _read_bytes(self, data: bytes, *, source_label: str = "<bytes>") -> Document:
pypdf = _require_pypdf()
pages_text: list[str] = []
page_map: list[tuple[int, str]] = []
with pypdf.PdfReader(io.BytesIO(data)) as reader:
total_pages = len(reader.pages)
for i, page in enumerate(reader.pages):
text = page.extract_text() or ""
pages_text.append(text)
page_map.append((i + 1, text))
full_text = "\n\n".join(t for t in pages_text if t.strip())
return Document(
content=full_text,
source=source_label,
metadata={
"size_bytes": len(data),
"total_pages": total_pages,
"page_map": page_map,
},
)