summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorJordan <me@jordan.im>2021-12-28 14:02:29 -0700
committerJordan <me@jordan.im>2021-12-28 14:02:29 -0700
commit51aa333b4f5b6b2ba3929a53cac8b38aea58c5c8 (patch)
tree96614996f9161994e076c73bdc0e48f66525d6ea
parentf5d402fdc34f8a6a06a7100aa3260e0193e5e49d (diff)
downloaddotfiles-51aa333b4f5b6b2ba3929a53cac8b38aea58c5c8.tar.gz
dotfiles-51aa333b4f5b6b2ba3929a53cac8b38aea58c5c8.zip
add ia-upload-warc
-rwxr-xr-xscripts/ia-upload-warc8
1 files changed, 8 insertions, 0 deletions
diff --git a/scripts/ia-upload-warc b/scripts/ia-upload-warc
new file mode 100755
index 0000000..789d1a8
--- /dev/null
+++ b/scripts/ia-upload-warc
@@ -0,0 +1,8 @@
+#!/usr/bin/env bash
+
+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
+start_url="$(cat $1/start_url)"
+
+ia upload $id $1/*.gz $1/*.cdx --metadata="mediatype:web" --metadata="date:$date" --metadata="title:$title" --metadata="description:recursive crawl of $start_url using grab-site/wpull, taken on $date" --metadata="source:$title"