An open API service indexing awesome lists of open source software.

https://github.com/dsdanielpark/arxiv2text

Converting PDF files to text, mainly with a focus on arXiv papers.
https://github.com/dsdanielpark/arxiv2text

crawling data datamining translation

Last synced: 10 months ago
JSON representation

Converting PDF files to text, mainly with a focus on arXiv papers.

Awesome Lists containing this project

README

          

Development Status :: 3 - Alpha

# arxiv2text


PyPI package
Code style: black

A Python package that converts arXiv documents into structured text using arXiv PDF URLs. If it's a PDF link other than arXiv, it's mostly feasible. If extraction to markdown fails, try extracting to text or HTML.

However, this package may be extracted as incomplete text, particularly in relation to expressions, as it is used as a sub-package for archival translations into other languages (e.g., Korean). Some parts of the text may be incomplete, and further post-processing will be required. The current plan is to perform self SFT to correctly align these expressions into LaTeX and aim to translate and summarize them using LLM.

We are considering transformations into the [LaTeXML](https://github.com/brucemiller/LaTeXML) format in the future. Therefore, it is currently challenging to expect a fully complete extraction in a general sense.


## Install
```
$ pip install arxiv2text
```
```
$ pip install git+https://github.com/dsdanielpark/arxiv2text.git
```


## Usage

*arxiv_to_text*

```python
from arxiv2text import arxiv_to_text

pdf_url = "https://arxiv.org/pdf/1706.03762"

extracted_text = arxiv_to_text(pdf_url, output_dir) # output_dir is optional argument
```

*arxiv_to_html*

```python
from arxiv2text import arxiv_to_html

pdf_url = "https://arxiv.org/pdf/1706.03762"

output_dir = "output_folder"
arxiv_to_html(pdf_url, output_dir)
```

*arxiv_to_md*
```python
from arxiv2text import arxiv_to_md

pdf_url = "https://arxiv.org/pdf/1706.03762"

output_dir = "output_folder"
arxiv_to_md(pdf_url, output_dir)
```


## [arXiv API](https://info.arxiv.org/help/api/user-manual.html)
```python
import urllib.request as libreq
with libreq.urlopen('http://export.arxiv.org/api/query?search_query=all:electron&start=0&max_results=1') as url:
r = url.read()
print(r)

```


## Project Plan
I welcome any contributions related to the project plan. Also, I'm waiting for collaborative contributors to enrich the package together. Thank you.
- Image extraction in paper
- Extracting annotations and images from the paper and automatically matching them.
- Conversion with LeTeXML
- LaTeX syntax error correction
- Summary and translation based on LLM


## License
[MIT](https://opensource.org/license/mit/)

```
The MIT License (MIT)

Copyright (c) 2023 Minwoo Park

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
```

## Bugs and Issues
Sincerely grateful for any reports on new features or bugs. Your valuable feedback on the code is highly appreciated.

## Contacts
- Core maintainer: [Daniel Park, South Korea](https://github.com/DSDanielPark)

- E-mail: parkminwoo1991@gmail.com


*Copyright (c) 2023 MinWoo Park, South Korea*