ChemCrow中paperscraper正确下载（避坑版）

1. paperscraper下载

（1）错误下载方式

我们如果直接使用pip install paperscraper命令下载，会发现paperscraper中并没有search_papers模块，此时我们下载的是pyliaorachel/paperscraper工程：

pyliaorachel/paperscraper: A python module to scrape arxiv.org for specific date range and categorieshttps://github.com/pyliaorachel/paperscraper

（2）正确下载方式

适合ChemCrow的paperscraper模块来自blackadad/paper-scraper: A proof of concept to scrape papers from journalshttps://github.com/blackadad/paper-scraper我们应当直接下载https://github.com/blackadad/paper-scraper

pip install git+https://github.com/blackadad/paper-scraper.git

2. paper_scraper函数修改

新版的paper-scraper模块将search_papers修改为异步的a_search_papers，我们需要对chemcrow工程做出对应的修改。

我的虚拟环境目录为：

~/miniconda3/envs/chemcrowpy310

那么chemcrow中paper_scraper函数路径为：

~/miniconda3/envs/chemcrowpy310/lib/python3.10/site-packages/chemcrow/tools/search.py

在该路径下将paper_scraper函数做出修改，以下是原函数

def paper_scraper(search: str, pdir: str = "query", semantic_scholar_api_key: str = None) -> dict:
    try:
        return paperscraper.search_papers(
            search,
            pdir=pdir,
            semantic_scholar_api_key=semantic_scholar_api_key,
        )
    except KeyError:
        return {}

将paperscraper.search_papers修改为paperscraper.a_search_papers：

def paper_scraper(search: str, pdir: str = "query", semantic_scholar_api_key: str = None) -> dict:
    try:
        return paperscraper.a_search_papers(
            search,
            pdir=pdir,
            semantic_scholar_api_key=semantic_scholar_api_key,
        )
    except KeyError:
        return {}