sabeel
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

[stats wf] added extra logging in the context data retrieval phase

This commit is contained in:
Claudio Atzori 2021-10-13 17:31:53 +02:00
parent 5d9cc2452d
commit b292e4a700
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/contexts.sh
View File

@ -11,9 +11,13 @@ TARGET_DB=$2
TMP=/tmp/stats-update-`tr -dc A-Za-z0-9 </dev/urandom | head -c 6`
echo "Downloading context data"
echo "Downloading context ids"
curl -L ${CONTEXT_API}/contexts/?type=ri,community -H "accept: application/json" | /usr/local/sbin/jq -r '.[] | "\(.id),\(.label)"' > contexts.csv
echo "Downloading categories data"
cat contexts.csv | cut -d , -f1 | xargs -I {} curl -L ${CONTEXT_API}/context/{}/?all=true | /usr/local/sbin/jq -r '.[]|"\(.id|split(":")[0]),\(.id),\(.label)"' > categories.csv
echo "Downloading concepts data"
cat categories.csv | cut -d , -f2 | sed 's/:/%3A/g'| xargs -I {} curl -L ${CONTEXT_API}/context/category/{}/?all=true | /usr/local/sbin/jq -r '.[]|"\(.id|split("::")[0])::\(.id|split("::")[1]),\(.id),\(.label)"' > concepts.csv
cat contexts.csv | sed 's/^\(.*\),\(.*\)/\1,\1::other,\2/' >> categories.csv
cat categories.csv | sed 's/^.*,\(.*\),\(.*\)/\1,\1::other,\2/' >> concepts.csv