From 0d2d8ae4a020c6d34c65bbc5de8513f04b5d722a Mon Sep 17 00:00:00 2001 From: Bryan Newbold Date: Tue, 22 Jun 2021 15:37:46 -0700 Subject: sitemaps: change filters; only primary release fulltext (via jq); scp to replica --- extra/sitemap/README.md | 4 ++++ extra/sitemap/access_urls_query.sh | 7 ++++--- extra/sitemap/work_urls_query.sh | 6 ++++-- 3 files changed, 12 insertions(+), 5 deletions(-) (limited to 'extra') diff --git a/extra/sitemap/README.md b/extra/sitemap/README.md index 53f9518..61ade0f 100644 --- a/extra/sitemap/README.md +++ b/extra/sitemap/README.md @@ -9,6 +9,10 @@ installed. Run these commands on a production machine. /srv/fatcat_scholar/src/extra/sitemap/access_urls_query.sh /srv/fatcat_scholar/src/extra/sitemap/generate_sitemap_indices.py +Then copy to alternate/replica machine: + + scp *.txt *.xml $SCHOLARREPLICAHOST:/srv/fatcat_scholar/sitemap + ## Background Google has a limit of 50k lines / 10 MByte for text sitemap files, and 50K diff --git a/extra/sitemap/access_urls_query.sh b/extra/sitemap/access_urls_query.sh index 7f1c3ef..6238a9d 100755 --- a/extra/sitemap/access_urls_query.sh +++ b/extra/sitemap/access_urls_query.sh @@ -8,10 +8,11 @@ SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )" # query for specific works; about 8.6 million circa 2021-04-29 # NOTE: filter out 12-digit access URLs (vs. 14-digit) -fatcat-cli search scholar 'doc_type:work (fulltext.access_type:ia_file OR fulltext.access_type:wayback) (NOT biblio.arxiv_id:*) (NOT biblio.pmcid:*) (NOT biblio.publisher_type:big5) (year:<1926 OR tags:*)' --index-json --limit 0 \ - | pv -l \ - | jq '[.work_ident, .fulltext.access_type, .fulltext.access_url] | @tsv' -r \ +fatcat-cli search scholar 'doc_type:work (fulltext.access_type:ia_file OR fulltext.access_type:wayback) (NOT biblio.arxiv_id:*) (NOT biblio.pmcid:*) ((NOT biblio.publisher_type:big5) OR year:<1926 OR tags:oa)' --index-json --limit 0 \ + | jq -c 'select(.biblio.release_ident == .fulltext.release_ident)' \ + | jq -r '[.work_ident, .fulltext.access_type, .fulltext.access_url] | @tsv' \ | rg -v '^null' \ | rg -v 'web.archive.org/web/\d{12}/' \ | $SCRIPT_DIR/transform_access_url.py \ + | pv -l \ | split --lines 20000 - sitemap-access- -d -a 5 --additional-suffix .txt diff --git a/extra/sitemap/work_urls_query.sh b/extra/sitemap/work_urls_query.sh index c369c35..813596b 100755 --- a/extra/sitemap/work_urls_query.sh +++ b/extra/sitemap/work_urls_query.sh @@ -7,10 +7,12 @@ set -o pipefail # fail if part of a '|' command fails # query for specific works; about 8.6 million circa 2021-04-29 # NOTE: filter out 12-digit access URLs (vs. 14-digit) -fatcat-cli search scholar 'doc_type:work (fulltext.access_type:ia_file OR fulltext.access_type:wayback) (NOT biblio.arxiv_id:*) (NOT biblio.pmcid:*) (NOT biblio.publisher_type:big5) (year:<1926 OR tags:*)' --index-json --limit 0 \ - | pv -l \ +fatcat-cli search scholar 'doc_type:work (fulltext.access_type:ia_file OR fulltext.access_type:wayback) (NOT biblio.arxiv_id:*) (NOT biblio.pmcid:*) ((NOT biblio.publisher_type:big5) OR year:<1926 OR tags:oa)' --index-json --limit 0 \ + | jq -c 'select(.biblio.release_ident == .fulltext.release_ident)' \ + | rg -v '^null' \ | rg -v 'web.archive.org/web/\d{12}/' \ | jq .key -r \ | tr '_' '/' \ | awk '{print "https://scholar.archive.org/" $1}' \ + | pv -l \ | split --lines 20000 - sitemap-works- -d -a 5 --additional-suffix .txt -- cgit v1.2.3