From bd14b70679338ad01956496936817d2ca1e6b2fc Mon Sep 17 00:00:00 2001
From: Gabe Kangas <gabek@real-ity.com>
Date: Thu, 20 Jan 2022 20:01:05 -0800
Subject: Fix janky hashtag detection. Closes #1686

---
 activitypub/outbox/outbox.go | 26 ++++++++++++--------------
 utils/utils.go               |  8 +++++++-
 utils/utils_test.go          | 16 +++++++++++++++-
 3 files changed, 34 insertions(+), 16 deletions(-)

diff --git a/activitypub/outbox/outbox.go b/activitypub/outbox/outbox.go
index 53369a7af..cf9c2da46 100644
--- a/activitypub/outbox/outbox.go
+++ b/activitypub/outbox/outbox.go
@@ -112,20 +112,18 @@ func SendPublicMessage(textContent string) error {
 
 	tagProp := streams.NewActivityStreamsTagProperty()
 
-	// Iterate through the post text and find #Hashtags.
-	words := strings.Split(originalContent, " ")
-	for _, word := range words {
-		if strings.HasPrefix(word, "#") {
-			tagWithoutHashtag := strings.TrimPrefix(word, "#")
-
-			// Replace the instances of the tag with a link to the tag page.
-			tagHTML := getHashtagLinkHTMLFromTagString(tagWithoutHashtag)
-			textContent = strings.ReplaceAll(textContent, word, tagHTML)
-
-			// Create Hashtag object for the tag.
-			hashtag := apmodels.MakeHashtag(tagWithoutHashtag)
-			tagProp.AppendTootHashtag(hashtag)
-		}
+	hashtagStrings := utils.GetHashtagsFromText(originalContent)
+
+	for _, hashtag := range hashtagStrings {
+		tagWithoutHashtag := strings.TrimPrefix(hashtag, "#")
+
+		// Replace the instances of the tag with a link to the tag page.
+		tagHTML := getHashtagLinkHTMLFromTagString(tagWithoutHashtag)
+		textContent = strings.ReplaceAll(textContent, hashtag, tagHTML)
+
+		// Create Hashtag object for the tag.
+		hashtag := apmodels.MakeHashtag(tagWithoutHashtag)
+		tagProp.AppendTootHashtag(hashtag)
 	}
 
 	activity, _, note, noteID := createBaseOutboundMessage(textContent)
diff --git a/utils/utils.go b/utils/utils.go
index e4610c97c..e3cb121f4 100644
--- a/utils/utils.go
+++ b/utils/utils.go
@@ -10,6 +10,7 @@ import (
 	"os/exec"
 	"path"
 	"path/filepath"
+	"regexp"
 	"strings"
 
 	"github.com/mssola/user_agent"
@@ -321,10 +322,15 @@ func GetHostnameFromURL(u url.URL) string {
 // GetHostnameFromURLString will return the hostname component from a URL object.
 func GetHostnameFromURLString(s string) string {
 	u, err := url.Parse(s)
-
 	if err != nil {
 		return ""
 	}
 
 	return u.Host
 }
+
+// GetHashtagsFromText returns all the #Hashtags from a string.
+func GetHashtagsFromText(text string) []string {
+	re := regexp.MustCompile(`#[a-zA-Z0-9_]+`)
+	return re.FindAllString(text, -1)
+}
diff --git a/utils/utils_test.go b/utils/utils_test.go
index ababc083a..ea41f48f6 100644
--- a/utils/utils_test.go
+++ b/utils/utils_test.go
@@ -1,6 +1,8 @@
 package utils
 
-import "testing"
+import (
+	"testing"
+)
 
 func TestUserAgent(t *testing.T) {
 	testAgents := []string{
@@ -16,3 +18,15 @@ func TestUserAgent(t *testing.T) {
 		}
 	}
 }
+
+func TestGetHashtagsFromText(t *testing.T) {
+	text := `Some text with a #hashtag goes here.\n\n
+	Another #secondhashtag, goes here.\n\n
+	#thirdhashtag`
+
+	hashtags := GetHashtagsFromText(text)
+
+	if hashtags[0] != "#hashtag" || hashtags[1] != "#secondhashtag" || hashtags[2] != "#thirdhashtag" {
+		t.Error("Incorrect hashtags fetched from text.")
+	}
+}
-- 
cgit v1.2.3