summaryrefslogtreecommitdiff
path: root/ia-upload-crawl.sh
blob: 76105453330df975605d71de3ff4d74cc059ca01 (plain)
1
2
3
4
5
6
7
8
9
10
11
#!/bin/sh

title="$(echo $1 | sed 's/\(-[0-9][0-9][0-9][0-9]-[0-9]\+-[0-9]\+.*\)//g')"
date="$(echo $1 | sed -n 's/.*\([0-9][0-9][0-9][0-9]-[0-9]\+-[0-9]\+\).*/\1/p')"
id=$title-$date
seeds="$(sed 'N;s/\n/, /' $1/seed_urls)"

ia upload $id $1/*.gz --metadata="mediatype:web" \
--metadata="date:$date" --metadata="title:$title" \
--metadata="description:recursive crawl of $seeds using https://git.jordan.im/crawl, \
taken on $date" --metadata="source:$title"