pubmed

from metapub import PubMedFetcher
import pickle as pk
from tqdm import tqdm
import uuid
fetch = PubMedFetcher()

import os
os.environ["NCBI_API_KEYY"] = "abcd"

final_data = {}

def get_keywords(keyword, max_n):
    # get the first 1000 pmids matching "breast neoplasm" keyword search
    pmids = fetch.pmids_for_query(keyword, retmax=max_n)
    # get abstract for each article:
    abstracts = {}
    for pmid in pmids:
        abstracts[pmid] = fetch.article_by_pmid(pmid).abstract
    return abstracts


def get_abstracts(abs_limit):
    
    with open('topics.pk','rb') as f:
        ks = pk.load(f)
        
    all_keywords = list(set(ks))
    result = {}
    for i in tqdm(all_keywords):
        try:
            temp_result    = {}
            keywrs         = get_keywords(i, abs_limit)
            temp_result[i] = keywrs
            result[i]      = keywrs
            with open('./data/' + str(uuid.uuid4()) + '.pk', 'wb') as f:
                pk.dump(temp_result,f)
        except Exception as e:
            print(e)
    
    with open('./data/all_data_in_one_' + str(i) + '.pk', 'wb') as f:
        pk.dump(result,f)
    return result