diff options
Diffstat (limited to 'sql/dump_ungrobid_pdf.sql')
-rw-r--r-- | sql/dump_ungrobid_pdf.sql | 13 |
1 files changed, 8 insertions, 5 deletions
diff --git a/sql/dump_ungrobid_pdf.sql b/sql/dump_ungrobid_pdf.sql index 3e6d782..e65edd5 100644 --- a/sql/dump_ungrobid_pdf.sql +++ b/sql/dump_ungrobid_pdf.sql @@ -1,15 +1,18 @@ -- Run like: --- psql sandcrawler < dump_ungrobid_pdf.sql | sort -S 4G | uniq -w 40 | cut -f2 > dump_ungrobid_pdf.2019-09-23.json +-- psql sandcrawler < dump_ungrobid_pdf.sql BEGIN TRANSACTION ISOLATION LEVEL SERIALIZABLE READ ONLY DEFERRABLE; COPY ( - SELECT cdx.sha1hex, row_to_json(cdx) FROM cdx - WHERE cdx.mimetype = 'application/pdf' - AND NOT EXISTS (SELECT grobid.sha1hex FROM grobid WHERE cdx.sha1hex = grobid.sha1hex) + SELECT DISTINCT ON (cdx.sha1hex) row_to_json(cdx) + FROM cdx + WHERE cdx.mimetype = 'application/pdf' + AND NOT EXISTS (SELECT grobid.sha1hex FROM grobid WHERE cdx.sha1hex = grobid.sha1hex AND grobid.status IS NOT NULL) + -- uncomment/comment this to control whether only fatcat files are included + --AND EXISTS (SELECT fatcat_file.sha1hex FROM fatcat_file WHERE cdx.sha1hex = fatcat_file.sha1hex) ) -TO STDOUT +TO '/grande/snapshots/dump_ungrobided_pdf.fatcat.2020-08-04.json' WITH NULL ''; ROLLBACK; |