From 0362d49da0d09784f9b974718839f5e69df3bae1 Mon Sep 17 00:00:00 2001
From: tobi <31960611+tsmethurst@users.noreply.github.com>
Date: Fri, 15 Mar 2024 18:26:53 +0100
Subject: [PATCH] [bugfix] Parse links that contain non-ascii characters
 (#2762)

---
 internal/text/markdown.go   | 6 +++++-
 internal/text/plain.go      | 6 +++++-
 internal/text/plain_test.go | 7 +++++++
 3 files changed, 17 insertions(+), 2 deletions(-)

diff --git a/internal/text/markdown.go b/internal/text/markdown.go
index 05d0a02f9..50cd6a141 100644
--- a/internal/text/markdown.go
+++ b/internal/text/markdown.go
@@ -24,6 +24,7 @@ import (
 	"codeberg.org/gruf/go-byteutil"
 	"github.com/superseriousbusiness/gotosocial/internal/gtsmodel"
 	"github.com/superseriousbusiness/gotosocial/internal/log"
+	"github.com/superseriousbusiness/gotosocial/internal/regexes"
 	"github.com/yuin/goldmark"
 	"github.com/yuin/goldmark/extension"
 	"github.com/yuin/goldmark/renderer/html"
@@ -61,7 +62,10 @@ func (f *Formatter) FromMarkdown(
 				false, // emojiOnly = false.
 				result,
 			},
-			extension.Linkify, // Turns URLs into links.
+			// Turns URLs into links.
+			extension.NewLinkify(
+				extension.WithLinkifyURLRegexp(regexes.LinkScheme),
+			),
 			extension.Strikethrough,
 		),
 	)
diff --git a/internal/text/plain.go b/internal/text/plain.go
index 1935cec8c..362941773 100644
--- a/internal/text/plain.go
+++ b/internal/text/plain.go
@@ -24,6 +24,7 @@ import (
 	"codeberg.org/gruf/go-byteutil"
 	"github.com/superseriousbusiness/gotosocial/internal/gtsmodel"
 	"github.com/superseriousbusiness/gotosocial/internal/log"
+	"github.com/superseriousbusiness/gotosocial/internal/regexes"
 	"github.com/yuin/goldmark"
 	"github.com/yuin/goldmark/extension"
 	"github.com/yuin/goldmark/parser"
@@ -158,7 +159,10 @@ func (f *Formatter) fromPlain(
 				emojiOnly,
 				result,
 			},
-			extension.Linkify, // Turns URLs into links.
+			// Turns URLs into links.
+			extension.NewLinkify(
+				extension.WithLinkifyURLRegexp(regexes.LinkScheme),
+			),
 		),
 	)
 
diff --git a/internal/text/plain_test.go b/internal/text/plain_test.go
index 43cc588c5..48280bb44 100644
--- a/internal/text/plain_test.go
+++ b/internal/text/plain_test.go
@@ -34,6 +34,8 @@ const (
 	withHTMLExpected           = "<p>&lt;div>blah this should just be html escaped blah&lt;/div></p>"
 	moreComplex                = "Another test @foss_satan@fossbros-anonymous.io\n\n#Hashtag\n\nText\n\n:rainbow:"
 	moreComplexExpected        = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br><br><a href=\"http://localhost:8080/tags/hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a><br><br>Text<br><br>:rainbow:</p>"
+	withUTF8Link               = "here's a link with utf-8 characters in it: https://example.org/söme_url"
+	withUTF8LinkExpected       = "<p>here's a link with utf-8 characters in it: <a href=\"https://example.org/s%C3%B6me_url\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">https://example.org/söme_url</a></p>"
 )
 
 type PlainTestSuite struct {
@@ -70,6 +72,11 @@ func (suite *PlainTestSuite) TestParseMoreComplex() {
 	suite.Equal(moreComplexExpected, formatted.HTML)
 }
 
+func (suite *PlainTestSuite) TestWithUTF8Link() {
+	formatted := suite.FromPlain(withUTF8Link)
+	suite.Equal(withUTF8LinkExpected, formatted.HTML)
+}
+
 func (suite *PlainTestSuite) TestLinkNoMention() {
 	statusText := `here's a link to a post by zork