Use metadata to determine Google Cloud project dynamically. Fixes: #601

services/clsi/bin/install_texlive_gce.sh

@@ -1,13 +1,19 @@
 #!/bin/sh
 METADATA=http://metadata.google.internal./computeMetadata/v1
 SVC_ACCT=$METADATA/instance/service-accounts/default
+PROJECT_URL=$METADATA/project/project-id
 ACCESS_TOKEN=$(curl -s -H 'Metadata-Flavor: Google' $SVC_ACCT/token | cut -d'"' -f 4)
-if [ -z "$ACCESS_TOKEN" ]; then
-	echo "No acccess token to download texlive-full images from google container, continuing without downloading. This is likely not a google cloud enviroment."
+#if [ -z "$ACCESS_TOKEN" ]; then
+#	echo "No acccess token to download texlive-full images from google container, continuing without downloading. This is likely not a google cloud enviroment."
+#	exit 0
+#fi
+PROJECT=$(curl -s -H 'Metadata-Flavor: Google' $PROJECT_URL)
+if [ -z "$PROJECT" ]; then
+	echo "No project name to download texlive-full images from google container, continuing without downloading. This is likely not a google cloud enviroment."
 	exit 0
 fi
 docker login -u '_token' -p $ACCESS_TOKEN https://gcr.io
-docker pull --all-tags gcr.io/henry-terraform-admin/texlive-full #TODO NEED TO MAKE THIS AN ENV VAR
+docker pull --all-tags gcr.io/$PROJECT/texlive-full
 cp /app/bin/synctex /app/bin/synctex-mount/synctex
 
 echo "Finished downloading texlive-full images"