Upload a dataset release to GitHub

dataset.sh

#! /bin/bash
#
# Upload release datasets to github.
#
# Usage:
#  ~/dataset.sh ./path/to/dataset
#
# Note that the path should go to the internal file inside the git repo
# in question, and not that by convention this folder name MUST MATCH
# the name of the repo.
#
TOKEN=`cat ~/.mbtoken`
ORG=menpo
P=$1

NAME=`basename $P`
echo tarring up dataset called $NAME at $P
TMP_PATH=/tmp/$NAME.tar.gz
tar -zcf $TMP_PATH $P
SHASUM=`shasum $TMP_PATH`
echo "SHA: $SHASUM"

RELEASES=`curl -sS -H "Authorization: token $TOKEN" https://api.github.com/repos/$ORG/$NAME/releases`
UPLOAD_TEMPLATE=`echo $RELEASES | python -c 'import json,sys;obj=json.load(sys.stdin);print(obj[0]["upload_url"])'`
UPLOAD_URL=`echo $UPLOAD_TEMPLATE | sed "s/{?name}/?name=$NAME.tar.gz/g"`
ASSET_ID=`echo $RELEASES | python -c 'import json,sys;obj=json.load(sys.stdin);print(obj[0]["assets"][0]["id"])'`
DELETE_URL=https://api.github.com/repos/$ORG/$NAME/releases/assets/$ASSET_ID
echo "Clearing any existing asset on latest release for $NAME"
curl -X DELETE -H "Authorization: token $TOKEN" $DELETE_URL
echo "Uploading $TMP_PATH to latest release for $NAME"
curl -X POST -H "Authorization: token $TOKEN" -H "Content-Type: application/x-gzip" $UPLOAD_URL --data-binary "@$TMP_PATH" > /dev/null
echo clearing up $TMP_PATH
rm $TMP_PATH

Warning to self - this just looks for the latest release and uploads. In the future you might want to add a second argument (version) to prevent any unexpected issues (uploading before a new tag is made)