aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2021-05-18 11:38:49 -0700
committerBryan Newbold <bnewbold@archive.org>2021-05-18 11:38:49 -0700
commit0afb1e361fcc601bf2a0f0fe2ea69f6ae8bd14fd (patch)
tree09fa3740794bf4c8e0df88402e4ad23888d2fa79
parent98479486b5c55865f1f67637387010e72fe2ddb0 (diff)
downloadfatcat-scholar-0afb1e361fcc601bf2a0f0fe2ea69f6ae8bd14fd.tar.gz
fatcat-scholar-0afb1e361fcc601bf2a0f0fe2ea69f6ae8bd14fd.zip
sitemaps: pdfs -> access
-rw-r--r--extra/sitemap/README.md2
-rwxr-xr-xextra/sitemap/access_urls_query.sh (renamed from extra/sitemap/pdf_urls_query.sh)2
-rwxr-xr-xextra/sitemap/generate_sitemap_indices.py4
3 files changed, 4 insertions, 4 deletions
diff --git a/extra/sitemap/README.md b/extra/sitemap/README.md
index 1a24620..53f9518 100644
--- a/extra/sitemap/README.md
+++ b/extra/sitemap/README.md
@@ -6,7 +6,7 @@ installed. Run these commands on a production machine.
cd /srv/fatcat_scholar/sitemap
/srv/fatcat_scholar/src/extra/sitemap/work_urls_query.sh
- /srv/fatcat_scholar/src/extra/sitemap/pdf_urls_query.sh
+ /srv/fatcat_scholar/src/extra/sitemap/access_urls_query.sh
/srv/fatcat_scholar/src/extra/sitemap/generate_sitemap_indices.py
## Background
diff --git a/extra/sitemap/pdf_urls_query.sh b/extra/sitemap/access_urls_query.sh
index fb1a4b8..bc7035f 100755
--- a/extra/sitemap/pdf_urls_query.sh
+++ b/extra/sitemap/access_urls_query.sh
@@ -11,4 +11,4 @@ fatcat-cli search scholar 'doc_type:work (fulltext.access_type:ia_file OR fullte
| jq '[.fulltext.access_type, .fulltext.access_url] | @tsv' -r \
| rg -v '^null' \
| ./transform_access_url.py \
- | split --lines 20000 - sitemap-pdfs- -d -a 5 --additional-suffix .txt
+ | split --lines 20000 - sitemap-access- -d -a 5 --additional-suffix .txt
diff --git a/extra/sitemap/generate_sitemap_indices.py b/extra/sitemap/generate_sitemap_indices.py
index ed4b38b..bfdc486 100755
--- a/extra/sitemap/generate_sitemap_indices.py
+++ b/extra/sitemap/generate_sitemap_indices.py
@@ -21,8 +21,8 @@ def index_entity(entity_type, output):
def main():
with open('sitemap-index-works.xml', 'w') as output:
index_entity("works", output)
- with open('sitemap-index-pdfs.xml', 'w') as output:
- index_entity("pdfs", output)
+ with open('sitemap-index-access.xml', 'w') as output:
+ index_entity("access", output)
if __name__=="__main__":
main()