mirror of https://github.com/hwchase17/langchain
community[minor]: add support for llmsherpa (#19741)
Thank you for contributing to LangChain! - [x] **PR title**: "community: added support for llmsherpa library" - [x] **Add tests and docs**: 1. Integration test: 'docs/docs/integrations/document_loaders/test_llmsherpa.py'. 2. an example notebook: `docs/docs/integrations/document_loaders/llmsherpa.ipynb`. - [x] **Lint and test**: Run `make format`, `make lint` and `make test` from the root of the package(s) you've modified. See contribution guidelines for more: https://python.langchain.com/docs/contributing/ Additional guidelines: - Make sure optional dependencies are imported within a function. - Please do not add dependencies to pyproject.toml files (even optional ones) unless they are required for unit tests. - Most PRs should not touch more than one package. - Changes should be backwards compatible. - If you are adding something to community, do not re-import it in langchain. If no one reviews your PR within a few days, please @-mention one of baskaryan, efriis, eyurtsev, hwchase17. --------- Co-authored-by: Bagatur <22008038+baskaryan@users.noreply.github.com> Co-authored-by: Bagatur <baskaryan@gmail.com>pull/14791/head^2
parent
a99bd098ac
commit
ba54f1577f
@ -0,0 +1,142 @@
|
||||
from pathlib import Path
|
||||
from typing import Iterator, Union
|
||||
from urllib.parse import urlparse
|
||||
|
||||
from langchain_core.documents import Document
|
||||
|
||||
from langchain_community.document_loaders.pdf import BaseLoader
|
||||
|
||||
DEFAULT_API = "https://readers.llmsherpa.com/api/document/developer/parseDocument?renderFormat=all"
|
||||
|
||||
|
||||
class LLMSherpaFileLoader(BaseLoader):
|
||||
"""Load Documents using `LLMSherpa`.
|
||||
|
||||
LLMSherpaFileLoader use LayoutPDFReader, which is part of the LLMSherpa library.
|
||||
This tool is designed to parse PDFs while preserving their layout information,
|
||||
which is often lost when using most PDF to text parsers.
|
||||
|
||||
Examples
|
||||
--------
|
||||
from langchain_community.document_loaders.llmsherpa import LLMSherpaFileLoader
|
||||
|
||||
loader = LLMSherpaFileLoader(
|
||||
"example.pdf",
|
||||
strategy="chunks",
|
||||
llmsherpa_api_url="http://localhost:5010/api/parseDocument?renderFormat=all",
|
||||
)
|
||||
docs = loader.load()
|
||||
"""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
file_path: Union[str, Path],
|
||||
new_indent_parser: bool = True,
|
||||
apply_ocr: bool = True,
|
||||
strategy: str = "chunks",
|
||||
llmsherpa_api_url: str = DEFAULT_API,
|
||||
):
|
||||
"""Initialize with a file path."""
|
||||
try:
|
||||
import llmsherpa # noqa:F401
|
||||
except ImportError:
|
||||
raise ImportError(
|
||||
"llmsherpa package not found, please install it with "
|
||||
"`pip install llmsherpa`"
|
||||
)
|
||||
_valid_strategies = ["sections", "chunks", "html", "text"]
|
||||
if strategy not in _valid_strategies:
|
||||
raise ValueError(
|
||||
f"Got {strategy} for `strategy`, "
|
||||
f"but should be one of `{_valid_strategies}`"
|
||||
)
|
||||
# validate llmsherpa url
|
||||
if not self._is_valid_url(llmsherpa_api_url):
|
||||
raise ValueError(f"Invalid URL: {llmsherpa_api_url}")
|
||||
self.url = self._validate_llmsherpa_url(
|
||||
url=llmsherpa_api_url,
|
||||
new_indent_parser=new_indent_parser,
|
||||
apply_ocr=apply_ocr,
|
||||
)
|
||||
|
||||
self.strategy = strategy
|
||||
self.file_path = str(file_path)
|
||||
|
||||
@staticmethod
|
||||
def _is_valid_url(url: str) -> bool:
|
||||
"""Check if the url is valid."""
|
||||
parsed = urlparse(url)
|
||||
return bool(parsed.netloc) and bool(parsed.scheme)
|
||||
|
||||
@staticmethod
|
||||
def _validate_llmsherpa_url(
|
||||
url: str, new_indent_parser: bool = True, apply_ocr: bool = True
|
||||
) -> str:
|
||||
"""Check if the llmsherpa url is valid."""
|
||||
parsed = urlparse(url)
|
||||
valid_url = url
|
||||
if ("/api/parseDocument" not in parsed.path) and (
|
||||
"/api/document/developer/parseDocument" not in parsed.path
|
||||
):
|
||||
raise ValueError(f"Invalid LLMSherpa URL: {url}")
|
||||
|
||||
if "renderFormat=all" not in parsed.query:
|
||||
valid_url = valid_url + "?renderFormat=all"
|
||||
if new_indent_parser and "useNewIndentParser=true" not in parsed.query:
|
||||
valid_url = valid_url + "&useNewIndentParser=true"
|
||||
if apply_ocr and "applyOcr=yes" not in parsed.query:
|
||||
valid_url = valid_url + "&applyOcr=yes"
|
||||
|
||||
return valid_url
|
||||
|
||||
def lazy_load(
|
||||
self,
|
||||
) -> Iterator[Document]:
|
||||
"""Load file."""
|
||||
from llmsherpa.readers import LayoutPDFReader
|
||||
|
||||
docs_reader = LayoutPDFReader(self.url)
|
||||
doc = docs_reader.read_pdf(self.file_path)
|
||||
|
||||
if self.strategy == "sections":
|
||||
yield from [
|
||||
Document(
|
||||
page_content=section.to_text(include_children=True, recurse=True),
|
||||
metadata={
|
||||
"source": self.file_path,
|
||||
"section_number": section_num,
|
||||
"section_title": section.title,
|
||||
},
|
||||
)
|
||||
for section_num, section in enumerate(doc.sections())
|
||||
]
|
||||
if self.strategy == "chunks":
|
||||
yield from [
|
||||
Document(
|
||||
page_content=chunk.to_context_text(),
|
||||
metadata={
|
||||
"source": self.file_path,
|
||||
"chunk_number": chunk_num,
|
||||
"chunk_type": chunk.tag,
|
||||
},
|
||||
)
|
||||
for chunk_num, chunk in enumerate(doc.chunks())
|
||||
]
|
||||
if self.strategy == "html":
|
||||
yield from [
|
||||
Document(
|
||||
page_content=doc.to_html(),
|
||||
metadata={
|
||||
"source": self.file_path,
|
||||
},
|
||||
)
|
||||
]
|
||||
if self.strategy == "text":
|
||||
yield from [
|
||||
Document(
|
||||
page_content=doc.to_text(),
|
||||
metadata={
|
||||
"source": self.file_path,
|
||||
},
|
||||
)
|
||||
]
|
@ -0,0 +1,46 @@
|
||||
from langchain_community.document_loaders.llmsherpa import LLMSherpaFileLoader
|
||||
|
||||
file_path = "https://arxiv.org/pdf/2402.14207.pdf"
|
||||
|
||||
|
||||
def test_llmsherpa_file_loader_initialization() -> None:
|
||||
loader = LLMSherpaFileLoader(
|
||||
file_path=file_path,
|
||||
)
|
||||
docs = loader.load()
|
||||
assert isinstance(loader, LLMSherpaFileLoader)
|
||||
assert hasattr(docs, "__iter__")
|
||||
assert loader.strategy == "chunks"
|
||||
assert (
|
||||
loader.url
|
||||
== "https://readers.llmsherpa.com/api/document/developer/parseDocument?renderFormat=all&useNewIndentParser=true&applyOcr=yes"
|
||||
)
|
||||
assert len(docs) > 1
|
||||
|
||||
|
||||
def test_apply_ocr() -> None:
|
||||
loader = LLMSherpaFileLoader(
|
||||
file_path=file_path,
|
||||
apply_ocr=True,
|
||||
new_indent_parser=False,
|
||||
)
|
||||
docs = loader.load()
|
||||
assert (
|
||||
loader.url
|
||||
== "https://readers.llmsherpa.com/api/document/developer/parseDocument?renderFormat=all&applyOcr=yes"
|
||||
)
|
||||
assert len(docs) > 1
|
||||
|
||||
|
||||
def test_new_indent_parser() -> None:
|
||||
loader = LLMSherpaFileLoader(
|
||||
file_path=file_path,
|
||||
apply_ocr=False,
|
||||
new_indent_parser=True,
|
||||
)
|
||||
docs = loader.load()
|
||||
assert (
|
||||
loader.url
|
||||
== "https://readers.llmsherpa.com/api/document/developer/parseDocument?renderFormat=all&useNewIndentParser=true"
|
||||
)
|
||||
assert len(docs) > 1
|
Loading…
Reference in New Issue