diff options
author | Bryan Newbold <bnewbold@archive.org> | 2022-10-03 10:16:26 -0700 |
---|---|---|
committer | Bryan Newbold <bnewbold@archive.org> | 2022-10-03 10:16:26 -0700 |
commit | 54e14814080d9a706ff6f15694b3b54918200169 (patch) | |
tree | c2ce152acb64a365408a57728c18d960b954c1f5 /sql/dump_reingest_spn.sql | |
parent | a04468041cd81ad90aa76ec15788a5ffacb6eec2 (diff) | |
download | sandcrawler-54e14814080d9a706ff6f15694b3b54918200169.tar.gz sandcrawler-54e14814080d9a706ff6f15694b3b54918200169.zip |
reingests: update scripts and SQL
Diffstat (limited to 'sql/dump_reingest_spn.sql')
-rw-r--r-- | sql/dump_reingest_spn.sql | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/sql/dump_reingest_spn.sql b/sql/dump_reingest_spn.sql index 65a8796..a83125c 100644 --- a/sql/dump_reingest_spn.sql +++ b/sql/dump_reingest_spn.sql @@ -8,7 +8,9 @@ COPY ( AND ingest_file_result.ingest_type = ingest_request.ingest_type WHERE (ingest_request.ingest_type = 'pdf' - OR ingest_request.ingest_type = 'html') + OR ingest_request.ingest_type = 'html' + OR ingest_request.ingest_type = 'xml' + OR ingest_request.ingest_type = 'component') AND ingest_file_result.hit = false AND ingest_request.created < NOW() - '6 hour'::INTERVAL AND ingest_request.created > NOW() - '180 day'::INTERVAL |