aboutsummaryrefslogtreecommitdiffstats
path: root/sql/dump_reingest_quarterly.sql
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2022-10-03 10:16:26 -0700
committerBryan Newbold <bnewbold@archive.org>2022-10-03 10:16:26 -0700
commit54e14814080d9a706ff6f15694b3b54918200169 (patch)
treec2ce152acb64a365408a57728c18d960b954c1f5 /sql/dump_reingest_quarterly.sql
parenta04468041cd81ad90aa76ec15788a5ffacb6eec2 (diff)
downloadsandcrawler-54e14814080d9a706ff6f15694b3b54918200169.tar.gz
sandcrawler-54e14814080d9a706ff6f15694b3b54918200169.zip
reingests: update scripts and SQL
Diffstat (limited to 'sql/dump_reingest_quarterly.sql')
-rw-r--r--sql/dump_reingest_quarterly.sql14
1 files changed, 11 insertions, 3 deletions
diff --git a/sql/dump_reingest_quarterly.sql b/sql/dump_reingest_quarterly.sql
index c377bf0..dbeb199 100644
--- a/sql/dump_reingest_quarterly.sql
+++ b/sql/dump_reingest_quarterly.sql
@@ -8,17 +8,25 @@ 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() - '8 hour'::INTERVAL
AND ingest_request.created > NOW() - '91 day'::INTERVAL
AND (ingest_request.ingest_request_source = 'fatcat-changelog'
- OR ingest_request.ingest_request_source = 'fatcat-ingest')
+ OR ingest_request.ingest_request_source = 'fatcat-ingest'
+ OR ingest_request.ingest_request_source = 'fatcat-ingest-container'
+ OR ingest_request.ingest_request_source = 'unpaywall'
+ OR ingest_request.ingest_request_source = 'arxiv'
+ OR ingest_request.ingest_request_source = 'pmc'
+ OR ingest_request.ingest_request_source = 'doaj'
+ OR ingest_request.ingest_request_source = 'dblp')
AND (
ingest_file_result.status like 'spn2-%'
OR ingest_file_result.status = 'cdx-error'
OR ingest_file_result.status = 'wayback-error'
- OR ingest_file_result.status = 'wayback-content-error'
+ -- OR ingest_file_result.status = 'wayback-content-error'
OR ingest_file_result.status = 'petabox-error'
OR ingest_file_result.status = 'gateway-timeout'
OR ingest_file_result.status = 'no-capture'