More reliable filtering of duplicate posts

Tweaks to filter_posts(), and now we use the
“Indempotency-Key” header, which should disallow
reposts for a short span of time (even if we lose
our history file).

manicito

@@ -51,7 +51,7 @@ fetch_hashtag_posts() {
 	local hashtag="$2"
 	local minimum_id="$3"
 
-	if test -n "$most_recent_post_id"; then
+	if test -n "$minimum_id"; then
 		curl --fail "$server/api/v1/timelines/tag/$hashtag?min_id=$minimum_id"
 	else
 		curl --fail "$server/api/v1/timelines/tag/$hashtag"
@@ -67,6 +67,7 @@ quote_post() {
 		 --request POST \
 		 --header "Authorization: Bearer $FEDI_AUTH" \
 		 --header 'Content-Type: application/json' \
+		 --header "Idempotency-Key: $(echo "$tagged_post_json" | jq -r .id)" \
 		 --data "$(post_json "$tagged_post_json")" "$FEDI_SERVER/api/v1/statuses"
 }
 
@@ -120,8 +121,7 @@ post_body() {
 filter_posts() {
 	local ids_to_filter="$1"
 	# If the list has more than give characters, use it to filter… (sanity check)
-	if test -n "$ids_to_filter" \
-			-a "$(echo "$ids_to_filter" | wc --bytes | awk '{print $1}')" -gt 5; then
+	if test -n "$ids_to_filter"; then
 		jq -cr .[] \
 			| grep --invert-match $(echo "$ids_to_filter" | sed 's%^%-e %')
 	else
@@ -133,8 +133,8 @@ filter_posts() {
 history_post_ids() {
 	local history_file="$1"
 	if test -f "$history_file"; then
-		grep -v '^[[:space:]]*$' "$history_file" \
-			| awk '{ print $2 }'
+		awk -F '\t' '{print $2}' "$history_file" \
+			| grep -v '^[[:space:]]*$'
 	fi
 }