aboutsummaryrefslogtreecommitdiffstats
path: root/python/fatcat_import.py
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@robocracy.org>2019-05-22 12:20:34 -0700
committerBryan Newbold <bnewbold@robocracy.org>2019-05-22 12:20:34 -0700
commitd33c8cf05e3c9732b04f56cf356180b9d76e04e0 (patch)
tree7c69bd1becae9d760b30c91012abcb159a62b73a /python/fatcat_import.py
parent4a3112f9f8de73511f354e7f1ceff3f8e2b7036d (diff)
downloadfatcat-d33c8cf05e3c9732b04f56cf356180b9d76e04e0.tar.gz
fatcat-d33c8cf05e3c9732b04f56cf356180b9d76e04e0.zip
creative importer for bulk JSTOR imports
Diffstat (limited to 'python/fatcat_import.py')
-rwxr-xr-xpython/fatcat_import.py18
1 files changed, 18 insertions, 0 deletions
diff --git a/python/fatcat_import.py b/python/fatcat_import.py
index 91fa2279..02a3441f 100755
--- a/python/fatcat_import.py
+++ b/python/fatcat_import.py
@@ -41,6 +41,12 @@ def run_pubmed(args):
else:
Bs4XmlFilePusher(pi, args.xml_file, "PubmedArticle").run()
+def run_jstor(args):
+ ji = JstorImporter(args.api,
+ args.issn_map_file,
+ edit_batch_size=args.batch_size)
+ Bs4XmlFileListPusher(ji, args.list_file, "article").run()
+
def run_orcid(args):
foi = OrcidImporter(args.api,
edit_batch_size=args.batch_size)
@@ -210,6 +216,18 @@ def main():
action='store_true',
help="consume from kafka topic (not stdin)")
+ sub_jstor = subparsers.add_parser('jstor')
+ sub_jstor.set_defaults(
+ func=run_jstor,
+ auth_var="FATCAT_AUTH_WORKER_JSTOR",
+ )
+ sub_jstor.add_argument('list_file',
+ help="List of JSTOR XML file paths to import from",
+ default=sys.stdin, type=argparse.FileType('r'))
+ sub_jstor.add_argument('issn_map_file',
+ help="ISSN to ISSN-L mapping file",
+ default=None, type=argparse.FileType('r'))
+
sub_orcid = subparsers.add_parser('orcid')
sub_orcid.set_defaults(
func=run_orcid,