aboutsummaryrefslogtreecommitdiffstats
path: root/python/sandcrawler/persist.py
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2020-06-17 18:06:12 -0700
committerBryan Newbold <bnewbold@archive.org>2020-06-17 18:06:12 -0700
commit386cb8335d4d1a66b75301a244f7baed49658588 (patch)
treea837ded7f4579ca7d9adcbd93f711347c7455b86 /python/sandcrawler/persist.py
parent815c2d115bbc2a64595a682bd15b95beac497c82 (diff)
downloadsandcrawler-386cb8335d4d1a66b75301a244f7baed49658588.tar.gz
sandcrawler-386cb8335d4d1a66b75301a244f7baed49658588.zip
tweak kafka topic names and seaweedfs layout
Diffstat (limited to 'python/sandcrawler/persist.py')
-rw-r--r--python/sandcrawler/persist.py3
1 files changed, 2 insertions, 1 deletions
diff --git a/python/sandcrawler/persist.py b/python/sandcrawler/persist.py
index 196c4b9..8d421ad 100644
--- a/python/sandcrawler/persist.py
+++ b/python/sandcrawler/persist.py
@@ -452,6 +452,7 @@ class PersistThumbnailWorker(SandcrawlerWorker):
default_bucket=kwargs['s3_bucket'],
)
self.s3_extension = kwargs.get('s3_extension', ".jpg")
+ self.s3_folder = kwargs.get('s3_folder', "pdf")
def process(self, blob, key=None):
"""
@@ -463,7 +464,7 @@ class PersistThumbnailWorker(SandcrawlerWorker):
assert len(blob) >= 50
resp = self.s3.put_blob(
- folder="thumbnail",
+ folder=self.s3_folder,
blob=blob,
sha1hex=key,
extension=self.s3_extension,