diff options
| author | Jules Laplace <julescarbon@gmail.com> | 2018-11-25 22:19:15 +0100 |
|---|---|---|
| committer | Jules Laplace <julescarbon@gmail.com> | 2018-11-25 22:19:15 +0100 |
| commit | ee3d0d98e19f1d8177d85af1866fd0ee431fe9ea (patch) | |
| tree | 41372528e78d4328bc2a47bbbabac7e809c58894 /scraper/s2-raw-papers.py | |
| parent | 255b8178af1e25a71fd23703d30c0d1f74911f47 (diff) | |
moving stuff
Diffstat (limited to 'scraper/s2-raw-papers.py')
| -rw-r--r-- | scraper/s2-raw-papers.py | 44 |
1 files changed, 44 insertions, 0 deletions
diff --git a/scraper/s2-raw-papers.py b/scraper/s2-raw-papers.py new file mode 100644 index 00000000..2d61f450 --- /dev/null +++ b/scraper/s2-raw-papers.py @@ -0,0 +1,44 @@ +import os +import sys +import csv +import subprocess +import time +import random +import re +import json +import click +from s2 import SemanticScholarAPI +from util import * + +s2 = SemanticScholarAPI() + +@click.command() +@click.option('--fn', '-i', default='missing.csv', help='Filename of CSV (id,)') +def fetch_raw_papers(fn): + lines = read_csv(fn, keys=False) + for line in lines: + paper_id = line[0] + fetch_raw_paper(paper_id) + +def fetch_raw_paper(paper_id): + os.makedirs(make_raw_paper_path(paper_id), exist_ok=True) + paper_fn = make_raw_paper_fn(paper_id) + if os.path.exists(paper_fn): + return read_json(paper_fn) + print(paper_id) + paper = s2.raw_paper(paper_id) + if paper is None: + print("Got empty paper?? {}".format(paper_id)) + time.sleep(random.randint(5, 10)) + return None + write_json(paper_fn, paper) + time.sleep(random.randint(5, 10)) + return paper + +def make_raw_paper_path(paper_id): + return './datasets/s2/raw_papers/{}/{}'.format(paper_id[0:2], paper_id) +def make_raw_paper_fn(paper_id): + return './datasets/s2/raw_papers/{}/{}/paper.json'.format(paper_id[0:2], paper_id) + +if __name__ == '__main__': + fetch_raw_papers() |
