1 files changed, 80 insertions, 0 deletions
diff --git a/notes/note.go b/notes/note.go
new file mode 100644
index 0000000..15f6f67
--- /dev/null
+++ b/notes/note.go
@@ -0,0 +1,80 @@
+package notes
+
+import (
+	"time"
+)
+
+type Media struct {
+	URL    string `json:"url"`
+	Alt    string `json:"alt"`
+	Width  int    `json:"width,omitempty"`
+	Height int    `json:"height,omitempty"`
+}
+
+type Note struct {
+	URI            string    `json:"uri"`
+	AuthorDID      string    `json:"authorDID"`
+	AuthorHandle   string    `json:"authorHandle"`
+	AuthorName     string    `json:"authorName"`
+	AuthorNoUnauth bool      `json:"authorNoUnauth"` // author has !no-unauthenticated label
+	AvatarURL      string    `json:"avatarURL"`
+	Text           string    `json:"text"`
+	CreatedAt      time.Time `json:"createdAt"`
+	ReplyTo        *string   `json:"replyTo,omitempty"`
+	Facets         []Facet   `json:"facets,omitempty"`
+	Parents        []Note    `json:"parents,omitempty"`
+	Media          []Media   `json:"media,omitempty"`
+}
+
+type Facet struct {
+	ByteStart int    `json:"byteStart"`
+	ByteEnd   int    `json:"byteEnd"`
+	Type      string `json:"type"`
+	Value     string `json:"value"` // the link destination, whether url, profile or tag
+}
+
+func MissingNote(uri string) *Note {
+	return &Note{
+		URI:       uri,
+		AuthorDID: "MISSING",
+		Text:      "This post could not be fetched. It may have been deleted.",
+		CreatedAt: time.Time{},
+	}
+}
+
+func isChainContained(smallChainPost, bigChainPost Note) bool {
+	if len(smallChainPost.Parents) > len(bigChainPost.Parents) {
+		return false
+	}
+
+	// if the thread is going to differ at any post, it's going to be at the end
+	// thus if the posts match their parents MUST match too
+	// (not that I expect to ever compare a post to itself but this way it won't break if I do. And technically a thread contains itself)
+	if len(smallChainPost.Parents) == len(bigChainPost.Parents) {
+		return smallChainPost.URI == bigChainPost.URI
+	}
+
+	// check if the last post of the smaller chain is the (n+1)-th post of the bigger chain
+	// where n is the number of parents in the smaller reply chain
+	offset := len(bigChainPost.Parents) - len(smallChainPost.Parents)
+	return smallChainPost.URI == bigChainPost.Parents[offset-1].URI
+}
+
+func DeduplicateThreads(posts []Note) []Note {
+	filtered := []Note{}
+	for _, post := range posts {
+		isRedundant := false
+		for _, kept := range filtered {
+			// check if this post's chain is contained in any kept post
+			if isChainContained(post, kept) {
+				isRedundant = true
+				break
+			}
+		}
+		if !isRedundant {
+			filtered = append(filtered, post)
+		}
+	}
+
+	return filtered
+}