aboutsummaryrefslogtreecommitdiffstats
path: root/python/sandcrawler_worker.py
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2020-11-03 22:40:57 -0800
committerBryan Newbold <bnewbold@archive.org>2020-11-03 22:42:05 -0800
commit8f964b9b48572ac71f27ba64207816dfd3a6dc36 (patch)
tree68a0146e71a0eef4e43f7b7d686b806eb210cda5 /python/sandcrawler_worker.py
parent653fac9632c6ae9dd036ad844454cf419cd5320b (diff)
downloadsandcrawler-8f964b9b48572ac71f27ba64207816dfd3a6dc36.tar.gz
sandcrawler-8f964b9b48572ac71f27ba64207816dfd3a6dc36.zip
small fixes from local testing for XML ingest
Diffstat (limited to 'python/sandcrawler_worker.py')
-rwxr-xr-xpython/sandcrawler_worker.py2
1 files changed, 0 insertions, 2 deletions
diff --git a/python/sandcrawler_worker.py b/python/sandcrawler_worker.py
index 24dbdd0..3681d7f 100755
--- a/python/sandcrawler_worker.py
+++ b/python/sandcrawler_worker.py
@@ -163,7 +163,6 @@ def run_persist_xml_doc(args: argparse.Namespace) -> None:
consume_topic=consume_topic,
group="persist-xml-doc",
push_batches=False,
- raw_records=True,
batch_size=25,
)
pusher.run()
@@ -182,7 +181,6 @@ def run_persist_html_teixml(args: argparse.Namespace) -> None:
consume_topic=consume_topic,
group="persist-html-teixml",
push_batches=False,
- raw_records=True,
batch_size=25,
)
pusher.run()