Parse `<category>` from Feeds (RSS, Atom and JSON)

2 files changed, 34 insertions, 3 deletions

diff --git a/storage/entry.go b/storage/entry.go
index 5f062f7f..f99d9ccc 100644
--- a/storage/entry.go
+++ b/storage/entry.go
@@ -119,7 +119,8 @@ func (s *Storage) createEntry(tx *sql.Tx, entry *model.Entry) error {
 				feed_id,
 				reading_time,
 				changed_at,
-				document_vectors
+				document_vectors,
+				tags
 			)
 		VALUES
 			(
@@ -134,7 +135,8 @@ func (s *Storage) createEntry(tx *sql.Tx, entry *model.Entry) error {
 				$9,
 				$10,
 				now(),
-				setweight(to_tsvector(left(coalesce($1, ''), 500000)), 'A') || setweight(to_tsvector(left(coalesce($6, ''), 500000)), 'B')
+				setweight(to_tsvector(left(coalesce($1, ''), 500000)), 'A') || setweight(to_tsvector(left(coalesce($6, ''), 500000)), 'B'),
+				$11
 			)
 		RETURNING
 			id, status
@@ -151,6 +153,7 @@ func (s *Storage) createEntry(tx *sql.Tx, entry *model.Entry) error {
 		entry.UserID,
 		entry.FeedID,
 		entry.ReadingTime,
+		pq.Array(removeDuplicates(entry.Tags)),
 	).Scan(&entry.ID, &entry.Status)
 
 	if err != nil {
@@ -183,7 +186,8 @@ func (s *Storage) updateEntry(tx *sql.Tx, entry *model.Entry) error {
 			content=$4,
 			author=$5,
 			reading_time=$6,
-			document_vectors = setweight(to_tsvector(left(coalesce($1, ''), 500000)), 'A') || setweight(to_tsvector(left(coalesce($4, ''), 500000)), 'B')
+			document_vectors = setweight(to_tsvector(left(coalesce($1, ''), 500000)), 'A') || setweight(to_tsvector(left(coalesce($4, ''), 500000)), 'B'),
+			tags=$10
 		WHERE
 			user_id=$7 AND feed_id=$8 AND hash=$9
 		RETURNING
@@ -200,6 +204,7 @@ func (s *Storage) updateEntry(tx *sql.Tx, entry *model.Entry) error {
 		entry.UserID,
 		entry.FeedID,
 		entry.Hash,
+		pq.Array(removeDuplicates(entry.Tags)),
 	).Scan(&entry.ID)
 
 	if err != nil {
@@ -535,3 +540,16 @@ func (s *Storage) UnshareEntry(userID int64, entryID int64) (err error) {
 	}
 	return
 }
+
+// removeDuplicate removes duplicate entries from a slice
+func removeDuplicates[T string | int](sliceList []T) []T {
+	allKeys := make(map[T]bool)
+	list := []T{}
+	for _, item := range sliceList {
+		if _, value := allKeys[item]; !value {
+			allKeys[item] = true
+			list = append(list, item)
+		}
+	}
+	return list
+}
diff --git a/storage/entry_query_builder.go b/storage/entry_query_builder.go
index 437e769c..dae7bc1f 100644
--- a/storage/entry_query_builder.go
+++ b/storage/entry_query_builder.go
@@ -135,6 +135,17 @@ func (e *EntryQueryBuilder) WithStatuses(statuses []string) *EntryQueryBuilder {
 	return e
 }
 
+// WithTags filter by a list of entry tags.
+func (e *EntryQueryBuilder) WithTags(tags []string) *EntryQueryBuilder {
+	if len(tags) > 0 {
+		for _, cat := range tags {
+			e.conditions = append(e.conditions, fmt.Sprintf("$%d = ANY(e.tags)", len(e.args)+1))
+			e.args = append(e.args, cat)
+		}
+	}
+	return e
+}
+
 // WithoutStatus set the entry status that should not be returned.
 func (e *EntryQueryBuilder) WithoutStatus(status string) *EntryQueryBuilder {
 	if status != "" {
@@ -250,6 +261,7 @@ func (e *EntryQueryBuilder) GetEntries() (model.Entries, error) {
 			e.reading_time,
 			e.created_at,
 			e.changed_at,
+			e.tags,
 			f.title as feed_title,
 			f.feed_url,
 			f.site_url,
@@ -312,6 +324,7 @@ func (e *EntryQueryBuilder) GetEntries() (model.Entries, error) {
 			&entry.ReadingTime,
 			&entry.CreatedAt,
 			&entry.ChangedAt,
+			pq.Array(&entry.Tags),
 			&entry.Feed.Title,
 			&entry.Feed.FeedURL,
 			&entry.Feed.SiteURL,